« Previous | Contents |
Listen
Footnotes
1. For the SOGS scores, the British survey used a cut off of 5 positive responses to determine 'problem' rather than 'pathological' gambling, in line with many other recent studies. It used a cut off of 3 for the DSM-IV scores to determine problem gambling, but for purposes of clarity did not incorporate the additional cut off of 5 or more to determine pathological gambling (Orford et al 2003: 158). Because of this, and in order to allow straightforward comparison with other countries, the results of both screens have been listed under the 'pathological' gambling heading in the Table. In a move which makes the measurement and comparison of gambling problems ever more complex, many countries now use a score of 5 to diagnose problem rather than specifically pathological gambling. Hence, it is the figures in the 'pathological gambling' column which should be looked at most closely in this table.
2. i. e.: the heaviest gamblers would be more likely to be regular attendees and so more likely to be available for interview, possibly skewing the overall results
3. New South Wales has almost 103,000 machines in total, of which only 1500 are located in casinos. Victoria has almost 30,000; of which 2500 are in casinos. Western Australia has under 1500 in total, all located in its casino (SA Centre for Economic Studies 2005).
4. In Australia, EGMs are associated with 65-80% of all problem gamblers receiving counselling, and in New Zealand, 90% of new counselling clients report non casino EGMs as their primary problem (Productivity Commission 1999; Abbott 2001).
5. Using a short, five item version of the SOGS, on which respondents report two or more problem items (rather than five in the longer version) to be classed as having problems. The authors also averaged out all the positive responses, and reported an increase in these from 0.131 to 0.198 - smaller figures, but which still represent a significant increase of round one half.
6. In the year before the Lottery, 40% of households gambled, spending £1.45 per week (0.5% of income), while in the year after, this had risen to 75%, spending £3.81 or 1.5% of household income per week.
7. New Zealand is the only country where national longitudinal surveys have been undertaken using the same measurement instrument, the revised version of the SOGS ( SOGS-R).
8. It was found that weekly gambling participation also decreased (from 48% in 1991, to 40% in 1999). This reduction was a consequence of fewer people participating weekly or more often in continuous forms including electronic gaming machines, track betting and instant lottery tickets.
9. Some have pointed to the fact that prevalence rates that are relatively consistent across jurisdictions and stable over time may obscure the movement of individuals in and out of disordered behaviour. In other words, the suggestion is that some people develop gambling problems, while others recover from them in an ongoing basis in a way that cancel each other out. For example, an eleven year longitudinal study with four assessment periods found that the overall prevalence of problem gambling remained steady at about 2-3%, although different groups of individuals contributed to these rates during the study (Slutske et al 2003; Shaffer et al 2004)
10. In the U.S., between 1990 - 2000 the population of counties with casinos grew on average 5% faster than those without. More densely populated counties (over 479,000 residents) grew about 8% faster (Baxandall and Sacerdote 2005).
11. In addition, the Indiana study also utilised estimates by the National Opinion Research Centre [ NORC] (which has conducted analyses for the U.S. National Gambling Commission). The estimates of social costs based on Grinols' methodologies were approximately double those calculated by NORC.
12. 10. Scores for problem gambling increased as income declined, from 0.2% or 0.3% as calculated by SOGS and DSM-IV respectively, for households earning more than £32,000 per year, to 1.5% and 1.0% for those earning less than £15,000 per year.
13. The exception to this appears to be casino gambling. The British prevalence study a slight correlation between casino table games and higher levels of educational attainment, indicating again the slightly different demographic status of casino gamblers. (Sproston, Erens & Orford 2000). Numbers were too small to determine relations between problematic casino playing and education, however.
14. The researchers noted that a 'clustering effect' existed and appeared to be based on the fact that the social groups that problem gamblers come from - disproportionately poor and ethnic minorities - also tend to reside in the same neighbourhoods. However, they also the controlled statistically for socio-economic status, which meant that the effects of 'neighbourhood disadvantage' could be viewed as a separate variable.
15. The survey did not cover Northern Ireland.
16. The 1999 survey study team developed and pre-tested their own DSM-IV based screen. The 1999 prevalence study team also set their own threshold score on each measurement tool before analysis of the data (the main reason for this focused on findings from the literature review which criticised the SOGS cut off as being too low). The cut off adopted by the research team for SOGS was a score of 5 or more which is higher than the original SOGS threshold of 3 or 4. The cut off for the DSM-IV did not change.
17. Data downloaded from the Scottish Neighbourhood Statistics website http://www.sns.gov.uk
18. http://www.scotland.gov.uk/library5/government/glsimd-00.asp
http://www.scotland.gov.uk/Publications/2005/09/2792129/21335
19. N.B. Long term unemployment figures here are expressed as a percentage of the unemployed. This means that if the unemployment rate is lower, the number of long-term unemployed will be lower, even for a given long-term unemployment rate. Long-term unemployment is defined here as being out of work for over twelve months.
« Previous | Contents |