International Journal of Mental Health and Addiction

, Volume 8, Issue 1, pp 70–81

Limited Cash Flow on Slot Machines: Effects of Prohibition of Note Acceptors on Adolescent Gambling Behaviour

Authors

    • Norwegian Institute for Alcohol and Drug Research
  • Ingeborg Rossow
    • Norwegian Institute for Alcohol and Drug Research
Article

DOI: 10.1007/s11469-009-9196-2

Cite this article as:
Hansen, M. & Rossow, I. Int J Ment Health Addiction (2010) 8: 70. doi:10.1007/s11469-009-9196-2

Abstract

This study addresses the impact of prohibition of note acceptors on gambling behaviour and gambling problems among Norwegian adolescents. Data comprised school surveys at three time points; 2004 and 2005 (before intervention) and 2006 (after intervention). Net samples comprised 20.000 students aged 13–19 years at each data collection. Identical measures of gambling behaviour (gambling frequency and expenditures on slot machines) and indicators of problem gambling (SOGS-RA and Lie/Bet) were assessed at all three time points. No significant changes in gambling behaviour and problem gambling were observed in the period prior to the intervention, whereas slot machine gambling frequency was reduced by 20%, the proportion that gambled frequently on slot machines was reduced by 26%; overall gambling frequency was reduced by 10% and the proportion of problem gamblers (SOGS-RA 4+) was reduced by 20% after the intervention when controlling for potential confounders. It is suggested that these findings can be attributed to the removal of the note acceptors on slot machines.

Keywords

Adolescent gamblingSlot machineProblem gamblingPreventionPrevalenceNote acceptorsNorway

Gambling on slot machines has been characterized as one of the most addictive forms of gambling (Echeburua and Fernandez-Montalvo 1996; Griffiths 1995; Turner and Horbay 2004). In the past decade we have seen an expansion in this form of gambling in all countries that allow gambling, and slot machines often represent the main gambling activity among available gambling devices (Australian Productivity Commission 1999; Norwegian Gaming Authority 2006; Williams and Wood 2004). Slot machines have been described as the “crack-cocaine” of gambling because of the assumed addictive features; short time between stake and payout or loss, rapid event frequency, reinforcing sounds and colours and the risk for loss of control over spendings (Dowling et al. 2005).

It has been argued that adolescents are especially vulnerable regarding the negative consequences of gambling; ‘addictive features’ and easy access makes gambling easy to start and more difficult to end (Derevensky and Gupta 2004; Griffiths 1993, 1995; Hardoon and Derevensky 2002). Adolescence is a vulnerable period from both a cognitive and neurological perspective; the adolescent brain is immature and neurological structures underlying the more complex and reflective cognitions of importance for self-regulation, are not fully developed until the early twenties (Chambers et al. 2003; Metcalfe and Mischel 1999; Mischel and Ayduk 2004). Moreover, adolescence is a period associated with experimentation and novelty seeking; in many countries, the current youth generation grows up in a society where gambling is—more or less—allowed, accepted and available. In line with a notion of increased susceptibility among young people, it is generally found that the prevalence of problem gambling among youths is two to four times higher compared to adults (Gupta and Derevensky 1998; Shaffer and Hall 1996, 2001; Shaffer et al. 1999).

Consequently, the expansion of gambling both nationally and internationally, is a matter of concern, and particularly so with respect to young people (Moodie and Hastings 2008).

What may serve as effective strategies to curb the negative consequences of slot machine gambling is much debated but still unresolved. However, various ways of regulating the gambling market, and particularly so restrictions on availability of gambling, are among the strategies that may have the potential to reduce the extent of excessive gambling and gambling problems (Blaszczynski et al. 2001) The evidence for such an assumption lies first and foremost in studies that have demonstrated that overall amount of gambling and prevalence of gambling problems have increased subsequent to changes in the gambling market that have implied increased availability (see for instance Grun and McKeigue 2000; Room et al. 1999; Turner et al. 1999). Whether there is a symmetry in the availability and problem prevalence association is, however, not clear; hence it is so far not obvious that a decrease in availability (for instance due to public policy regulations of the gambling market) actually will result in a decrease in the prevalence of problem gambling.

In Norway, slot machines have been the most dominant game, both among adults and adolescents (Hansen and Rossow 2008; Lund and Nordlund 2002; Rossow and Hansen 2003). Slot machines have been easily available in grocery stores and shopping centres. The minimum legal age for slot machine gambling is 18 years, however, the enforcement of the age limit has been close to neglible. Survey studies (Lund and Nordlund 2002; Rossow and Hansen 2003) as well as data from a gambling help-line in Norway have also shown that slot machines constitute the dominant game and major problem among problem gamblers.

In 2003 the Norwegian government made a resolution to prohibit all existing slot machines and establish a state monopoly of slot machines with the purpose to both ensure revenue to idealistic purposes and to prevent and reduce gambling related harm. The Norwegian state was prosecuted by the gambling industry, both in the Norwegian court system and in the EFTA court. The legal decision in this case was not made until March 2007, when the court upheld the Norwegian state’s claim. Meanwhile it was decided to prohibit note acceptors in order to curb an increasing turnover on slot machines. By July 1st 2006 bank note acceptor on all slot-machines in Norway had been removed and the highest stake had been reduced from notes equivalent to 25 Euros to coins equivalent to 2.5 Euro. The turnover dropped with more than 40% after the intervention and this downswing kept stable until the slot machines were removed 1 year later (Norwegian Gaming Authority 2006, 2007).

There is, as noted, very little—if any—evidence of the effectiveness of this kind of market regulation on the prevalence of problem gambling. Yet, some studies on note acceptors on slot machines and excessive and problem gambling are relevant here.

Australian surveys have shown that problem gamblers preferred to use note acceptors while gambling (Australian Productivity Commission 1999), and they have also found a strong relationship between regular and problem gambling and frequent use of note acceptors (McMillen et al. 2004). Furthermore, one study has shown a decrease in expenditures, gambling frequency and size of bets among at-risk and problem gamblers by reducing the bank note denomination (Brodie et al. 2003). Other studies found that problem gamblers used high denomination bills more often than other gamblers (Sharpe et al. 2005) and Blaszczynski and colleagues found that by limiting note acceptors from $100 to $20 denominations, the losses were reduced by 42%.

Although of relevance, these studies have, for various reasons, limited transfer value with respect to possible effects of prohibition of note acceptors on the extent of excessive gambling and problem gambling. First, these studies have addressed the potential impact of changing the size of bank note denominations rather than prohibition of note acceptors, and second, the samples which these studies are based on are with few exceptions collected from certain gambling populations.

The aim of the present study was therefore to assess whether the regulation of the Norwegian slot machine market in terms of the prohibition of the note acceptors on slot machines had any impact on youth gambling. More specifically we have assessed whether gambling frequency, gambling expenditures and problem gambling among young people have decreased in response to the prohibition of the note acceptors on slot machines.

Methods

Participants and Procedures

The present study is based on data from Norwegian school surveys conducted in the same municipalities and schools at three time points; in 2004, 2005 and 2006. The school surveys were conducted in October/November each year as a joint data collection for two different evaluation projects; partly to assess possible effects of alcohol and drug prevention strategies at the local level and partly to assess possible effects of national efforts on regulating the slot machine market. The target population in all three surveys were all students, from junior and senior high school (mainly public schools), in grades 8–13 in altogether 16 municipalities in Norway. In senior high school students were included both from academic and vocational training. In 2004 a written informed parental consent was requested for all students under the age of 18 years, in 2005 and 2006 this was only requested for students in junior high school (grades 8–10) whereas a passive informed parental consent from students below 18 years in senior high school was obtained. The questionnaire was completed at school during one school hour. More information about procedures and data collection is described in detail in Pape et al. (2007). Net samples comprised 20,703 students in 2004 (pre-intervention); 21,295 in 2005 (pre-intervention); and 20,695 in 2006 (post-intervention), and the response rates were 82.7% (2004), 86.7% (2005) and 85.7% (2006), respectively. The respondents were mainly 13–19 years old, average age was 15 years and 50% were girls. Table 1 presents a summary of the demographic data.
Table 1

Demographic Characteristics of All Students (Grades 8–13) Surveyed in 16 Municipalities in Norway

 

2004

2005

2006

n = 20,648

n = 21,260

n = 20,573

Gender:

Males

49.3% (10,203)

50.1% (10,660)

49.6% (10,197)

Females

50.7% (10,481)

49.9% (10,600)

50. 4% (10,374)

Age:

13–15

52.4% (10,847)

49.6% (10,522)

50.5% (10,416)

16–17

31.7% (6,564)

36.5% (7,729)

34.6% (7,136)

18 +

15.9% (3,292)

13.9% (2,956)

15% (3,090)

Measures

The outcome variables; indicators of gambling behaviour and of problem gambling were assessed by identical measures at the three data collection waves. Two screening measures to assess problem gambling were used; the Lie/Bet Questionnaire (Johnson et al. 1988; Johnson et al. 1997) and the SOGS-RA (Winters et al. 1993). The Lie/Bet Questionnaire is assumed to have high sensitivity (.99) and specificity (.91) (Johnson et al. 1988; Johnson et al. 1997) and to be applicable for adolescents (Götestam et al. 2004; Rossow and Molde 2006). The Lie/Bet score ranges from 0–2, and a score of 1+ as well as a score of two are used to categorize at-risk gamblers. The SOGS-RA is a 12 item gambling screen adapted from the adult version of SOGS-R (Lesieur and Blume 1987) for use among adolescents, and is found to discriminate well between those who gamble regularly and not (Winters et al. 1993), and that past year gambling expenditures and gambling frequency is associated with an increasing SOGS-RA (Poulin 2000, 2002). The SOGS-RA score ranges from 0–12, and a score of 4+ is used to categorize problem gamblers, whereas the corresponding scores for at-risk gamblers are 2–3, while no-problem gamblers are in the 0–1 score category (Winters et al. 1995). These categories were also used in the present study.

Gambling behaviour was assessed in terms of frequency of gambling on six various games in the preceding 12 months and expenditures on slot machine gambling in the preceding week. The six response categories on the six gambling frequency variables ranged from ‘several times a week’ to ‘not gambled last year’. A mid-point value for each category was applied and frequency variables were transformed into semi-continuous variables on number of times gambled last year. On the basis of these six semi-continuous variables a sum score on number of times gambled last year on all games was constructed. In the analyses two variables on gambling frequency were applied; gambling frequency on slot machines and gambling frequency on all games. The students were also asked open ended questions about total expenditures on slot machines during the last week.

No changes occurred in the slot machine market in Norway from 2004–2005. By the end of June 2006 all bank note acceptors were removed on all slot machines. Two school surveys were conducted prior to the removal of note acceptors and one school survey was conducted approximately 4 months after the removal of note acceptors. Thus, changes in the outcome variables from 2005–2006 could indicate a possible impact of the regulation. However, as other factors than regulations of the slot machine market could also contribute to changes in gambling behaviour and prevalence of problem gambling over time and as no control group for the national intervention could be established, data from the first of the 2 years prior to intervention (2004 data) served as control for possible trends in gambling behaviour and prevalence of problem gambling due to other factors.

In 2006 (post-intervention) some additional questions were included to strengthen interpretations of possible intervention impact; whether the students had noticed the removal of note acceptors; whether they gambled more or less on slot machines after the removal of note acceptors; and if so, whether they felt that their changed gambling on slot machines was due to the intervention.

Statistical Analyses

Expenditures and gambling frequencies on slot machines and all games were compared for the years 2006 (after the intervention) and 2005 (before the intervention). Similar pairwise comparisons were conducted for the 2 years prior to the intervention. Differences between means on the continuous variables (number of times gambled last year on slot machines, number of times gambled last year on all games and the amount of money spent on slot machines last week) were tested by analysis of variance (ANOVA) and F-tests whereas differences between proportions (e.g. the proportion that gambled once or more per week on slot machines; the proportion that scored 4 + on SOGS-RA) were tested by Pearson’s Chi-square statistics.

Multi-variate analyses were conducted to assess possible impact of the intervention adjusting for trend in gambling behaviour and problem gambling and minor differences in age and gender distribution in the three data collection waves. Logistic regression models were estimated for the binary outcome variables; the proportion of gambling once per week or more frequently; the proportion not spending money on slot machines last week; and the proportion having spent more than 63 Euro on slot machines last week as well as for the various indicators of ‘At-risk’ and ‘Problem-gambling’ applying various cut-off scores on SOGS-RA and The Lie-Bet Questionnaire. Correspondingly, multivariate linear regression analysis models were estimated for the continuous outcome variables (number of times gambled last year on slot machines, number of times gambled last year on all games and the amount of money spent on slot machines last week). In all models the intervention variable (change from 2005–2006) was entered together with the possible confounding variables age, gender and trend from 2004–2005.

Due to the study design the data comprised information that in part came from the same individuals at two or three time points and thus implied repeated measurements for these individuals. As only a fraction of these same individuals were identifiable over time, additional analyses were conducted on various sub-samples that did not comprise the same individuals at the three time points (e.g. only 8th and 11th graders at each time point).

Results

Compared to 2005, expenditures on slot machines as well as overall gambling frequency, showed a substantial decrease after the intervention. During the 2 year period prior to the intervention, the level of gambling behaviour was more or less stable. For example, average number of times gambled on slot machines last year was stable from 2004–2005; i.e. an average frequency of 15 times per year. In 2006, after the intervention, the average gambling frequency on slot machines had decreased to 12 times per year. The difference between 2005 and 2006 was statistically significant, tested with ANOVA and F-test (Table 2).
Table 2

Indicators of Gambling Behaviour by Survey Year

 

Pre-intervention

Statistical test of difference from 2004–2005

Post-inter-vention

Statistical test of difference from 2005–2006

Estimates from multi-variate models of changes in gambling behaviour from 2005–2006

2004

2005

2006

Adj OR Coeff

Adj regr.

Gambling frequency slot machines:

-number of times gambled last year

15

15

F = 1.98 ns df = 1

12

F = 95.03*** df = 1

 

−3.0***

-proportion who have gambled once or more per week

11.3%

10.6%

χ2 = 5.83* df = 1

8.1%

χ2 = 74.71*** df = 1

0.74***

 

Expenditures:

-amount spent on slot machines (Euro last week/average)a

3.40

3.10

F = 0.86 ns df = 1

2.40

F = 4.31** df = 1

 

−0.7*

-proportion who did not spend on slot machines last week

80%

83%

χ2 = 68.88*** df = 1

88%

χ2 = 198.48*** df = 1

1.50***

 

-proportion having spent > 63 Euro last week

1.3%

1.1%

χ2 = 2.34 ns df = 1

0.8%

χ2 = 9.57** df = 1

0.73**

 

Gambling frequency all games:

-number of times gambled last year

47

50

F = 4.21* df = 1

45

F = 14.81*** df = 1

 

−5.2***

Tests of bi-variate differences in gambling behaviour between the two pre-intervention survey years and between —pre- and post-intervention survey years, estimates from multi-variate models of changes in gambling behaviour from 2005 (pre-intervention) to 2006 (post-intervention).

aCut-off 1125 Euro per week, ns = not significant, * = p < 0.05, ** = p < 0.01, *** = p < 0.001

Multivariate analyses, controlling for age, gender and possible trends from 2004–2005, revealed significantly less gambling behaviour after the intervention. Multivariate logistic regression analyses showed that the adolescents were 26% less likely to gamble weekly or more often on slot machines after the intervention (OR = 0.74; p < .001); 50% more likely not to have gambled at all on slot machines last week (OR = 1.50; p < .001) and 27% less likely to have gambled for more than 63 Euros on slot machines last week (OR = 0.73; p < .01) after the bank note acceptors were prohibited and removed. Multivariate linear regression analyses showed that average number of times gambled last year decreased statistically significantly for slot machines (−3.0) and for all games (−5.2), when controlling for age, gender and a possible trend in the gambling market from 2004–2005 (Table 2).

All indicators of problem gambling showed a significant decrease from 2005–2006; i.e. from before to after removal of note acceptors. In the pre-intervention period (2004–2005) the main pattern was no changes in the prevalence of problem gambling. Although prevalence of at-risk and problem gambling (SOGS-RA 2–3 and 4+) changed in opposite directions from 2004–2005, a combined indicator of at-risk and problem gambling (SOGS-RA 2+) showed a fairly stable prevalence from 2004–2005. A similar stability and decrease was evident when The Lie/Bet Questionnaire was applied as problem gambling indicator.

Multivariate logistic regression models controlling for age, gender and trend from 2004–2005 estimated the associations between the intervention and the problem gambling indicators. The odds ratios varied between 0.83 and 0.71, implying a 17–29% decrease in the prevalence of problem gambling after the intervention. For instance, the prevalence of at-risk gamblers (SOGS 2–3) and problem gamblers (SOGS 4+) was 29% (OR = 0.71; p < .001) and 20% (OR = 0.80; p < .001) lower after the removal of the note acceptors (Table 3).
Table 3

Indicators of At-risk and Problem Gambling before and after the Removal of Bank Note Acceptors, Tests of Bi-variate Differences in At-risk and Problem Gambling between the Two Pre-intervention Survey Years and between —Pre- and Post-intervention Survey Years, Estimates from Multi-variate Models of Changes in Gambling Behaviour from 2005 (Pre-intervention) to 2006 (Post-intervention)

 

Pre-intervention

Statistical test of difference from 2004 to 2005

Post-intervention

Statistical test of difference from 2005–2006

Adj OR

2004

2005

2006

SOGS-RA:

At- risk and problem gambling (2+)

8.3%

7.8%

χ2 = 3.60* df = 1

5.8%

χ2 = 61.97*** df = 1

0.74***

Problem (4+)

2.4%

2.9%

χ2 = 9.61** df = 1

2.3%

χ2 = 13.54*** df = 1

0.80***

At-Risk (2–3)

5.9%

4.9%

χ2 = 20.00*** df = 1

3.5%

χ2 = 48.37*** df = 1

0.71***

The Lie/Bet Questionnaire:

Lie/Bet (2)

3.7%

3.6%

χ2 = 0.18 ns df = 1

3.0%

χ2 = 12.13*** df = 1

0.83***

Lie/Bet (1+)

14.6%

13.8%

χ2 = 4.61* df = 1

11.5%

χ2 = 49.29*** df = 1

0.81***

ns = not significant, * = p < 0.05, ** = p < 0.01, *** = p < 0.001

All bi-variate and multi-variate analyses presented in Tables 2 and 3 were re-run for various sub-samples comprising only individuals without repeated measurements over the 3 year observation period, as for instance 8th and 11th graders at each time point. For all outcome variables, both on gambling behaviour and indicators of problem gambling, the results from these bi-variate and multi-variate analyses were in line with the results from the analyses of the total samples, - a significant reduction in gambling behaviour and prevalence of problem gambling was observed from 2005 (pre-intervention) to 2006 (post-intervention).

Although it is generally found that boys gamble more often; spend more money on gambling and more often report gambling related problems, as compared to girls (Derevensky and Gupta 2004; Hansen and Rossow 2008), we did not find any significant gender (or age) differences in the reduction of gambling behaviour and prevalence of problem gambling from 2005–2006 (before and after the removal of bank note acceptors). One third of the adolescents who had gambled last year reported to have noticed the removal of the bank note acceptors, and two thirds reported that they had either stopped gambling or gambled less frequently than they did before the intervention. Still, only a small fraction of the adolescents related the changes in their gambling behaviour to the removal of the bank note acceptors (Table 4).
Table 4

Prohibition of Note Acceptors — Self-reported Influence on Gambling Behaviour among Adolescents Who Reported Gambling on Slot Machines Last Year (After Intervention)

 

Number of valid responses (n)

Responses

Have you noticed the removal of note acceptors on slot machines?

10947

Yes:

31.4%

 

No

68.6%

Do you now gamble as much on slot machines as you did before the removal of note acceptors?

9766

Yes:

28.6%

 

No, gamble more:

3.2%

 

No, gamble less:

21.2%

 

Stopped gambling:

47.0%

If you have changed — is it related to the removal of note acceptors?

7420

Yes:

6.7%

 

No:

75.7%

 

Not sure

17.6%

Discussion

This study has examined a possible impact of removal of all note acceptors on slot machines on gambling behaviour and gambling problems among Norwegian adolescents.

Gambling frequency, gambling expenditures and problem gambling among young people from 13–19 year decreased significantly in the time period after the removal of note acceptors. No transition to other forms of gambling was observed after the intervention. The decrease in problem gambling was obvious both for ‘At-risk gamblers’ as well as ‘Problem gamblers’, and the proportion of those who gambled for more than 63 euro per week on slot machines—was reduced by one third from 2005–2006. The decrease was generally of the same magnitude for both boys and girls and in various age groups.

This is the first study where the effect of removing note acceptors from slot machines has been evaluated. Prohibition and removal of all note acceptors on all slot machines has not been carried out in any other country previously and this gave us a unique opportunity to evaluate the effects of a potentially significant measure to regulate the slot machine market and curb extensive slot machine gambling.

Previous research has evaluated the effect of reducing the bank note denomination allowed on slot machines and these studies may indicate whether this kind of intervention may have an impact on gambling behaviour. The results have so far been diverse and the studies have—with few exceptions—been carried out in certain populations of gamblers. However, a general finding of these studies was that a reduction of the bank note denomination influenced the expenditure on slot machine gambling in a downward direction (Australian Productivity Commission 1999; Blaszczynski et al. 2001; Brodie et al. 2003; Sharpe et al. 2005). Moreover ‘At-risk’ and ‘Problem gamblers’ reported a decrease in gambling related problems and ‘excessive gamblers’ decreased their spending to a larger extent than other gamblers (Brodie et al. 2003). In this respect our findings are in line with previous research and point in the direction that restrictions related to note acceptors may curb gambling expenditures and gambling related problems.

It has been found that coinless machines can speed up the game with 15% (Palmeri 2003) and as absence of note acceptors implies that gamblers will have to break up gambling sessions in order to exchange bank notes into coins, this may therefore reduce the likelihood of rapid and continuous gambling sessions. It is possible that such breaks facilitate considerations on whether gambling should be continued or not and thereby contribute to a less continuous play.

It is possible that interruptions in the gambling sessions can be a trigger to increased consciousness and self-regulation. Mischel and Ayduk (2004) differentiated between two types of cognitive processing; a “cool system” that is slower and more cognitively complex generating more rational, reflective and strategic behaviour, and a “hot system” that enables quick and emotional processing. The hot system is essentially an automatic system, governed by virtually reflective stimulus-response reactions, consisting of relatively few representations. People with gambling problems often describe their slot machine gambling as an automatized kind of behaviour. An interruption in the game can contribute to a “cool down” period and the gamblers’ opportunity to make a new choice on whether to continue the gambling or not, is strengthened. During gambling, losses and chasing the losses are often associated with stress. Metcalfe and Mischel (1999) noted that an increased stress level results in an increasingly dysfunctional “cool system”, leaving the “hot system” to dominate processing. In a gambling context, the need for exchange into coins can result in a change from “hot state” to “cool state”, resulting in a more rationally and less emotionally driven cognition which makes it less demanding to break off the gambling session. Because adolescent neurodevelopment occurs in brain regions associated with impulsivity and addiction, adolescent novelty seeking can partly be explained by immature brain development (Chambers et al. 2003). Therefore an understanding of how self-regulation among young gamblers can be facilitated is of high relevance.

No other interventions in the gambling market occurred concurrent with the prohibition of note acceptors during the data collection period. However, it appears that there has been a change in the norm climate in the Norwegian society. This has been expressed in several ways; firstly by an increased focus and concern by clinicians, relatives to gamblers and pathological gamblers and secondly actions taken by politicians to curb the negative consequences of slot machine gambling. It seems likely that norm climate and attitudes towards slot machine gambling has turned more negative in part due to media focus on gambling problems. If so, it is possible slot machine gambling has become less socially acceptable during this period which again may have reinforced the intervention effect.

Ladouceur et al. (2005) suggested that a public health policy on slot machines requires measures that maximise the protection of slot machine players against excessive gambling. The present study suggests that removing note acceptors may serve as one such public policy measure. A few previous studies have suggested that a total consumption model of gambling seems applicable as there appears to be a clear positive association between overall amount of gambling (total consumption) and prevalence of problem gambling (Hansen and Rossow 2008; Lund 2008). The findings of the present study are also in line with a total consumption model of gambling as we observed a decrease in the prevalence of problem gambling concurrent with a decrease in gambling frequency and expenditures after the intervention.

Some study limitations and strengths should be mentioned. There is a relatively short time span from the intervention to the data collection at T3. Four months is a short period to evaluate changes and it is reasonable to ask how much change can be expected in 4 months. Further, we only have cross-sectional data and thus we do not know who has reduced gambling behaviour after the intervention. Nevertheless, the observed decrease in gambling frequency, expenditures and prevalence of problem gambling was significant, and it is difficult to see any alternative explanations to this decrease other than the removal of note acceptors on slot machines. The stability in observed gambling behaviour in the 2 years prior to the intervention as well as lack of any other significant changes in the gambling market during the observation period, suggest that it is reasonable to attribute these findings to the removal of the note acceptors on slot machines. A considerable strength of this study is the large samples and the high response rate at all three points of time of the data collection.

The motivation to regulate the Norwegian slot machine market was to limit and prevent gambling related problems—especially among adolescents. This study suggests that regulatory measures may have an effect—not only on the total expenditures on slot machines—but also on the extent of gambling related problems and excessive gambling. Research in this field is scarce and replication of the findings is warranted. A challenge for further studies in this area will therefore be to assess whether and to what extent the removal of note acceptors on slot machines may have had a similar effect in the Norwegian adult population, and furthermore, whether this kind of intervention would have a similar effect in other countries and other kinds of gambling markets.

Conflict of interests

None.

Copyright information

© Springer Science+Business Media, LLC 2009