20. Problem Gambling, Mental Health, Alcohol and Drug Abuse: Effects on Crime

According to the American Psychiatric Association (APA) “pathological” or “disordered” gambling is an impulse control disorder. The APA provides diagnostic guidelines in its Diagnostic and Statistical Manuals (DSMs III, IV, and now V).

This definition is from The Random House College Dictionary, Revised Edition, New York: Random House, 1982.

National Research Council (1999), Grinols (2004a, 2007), and New Hampshire Gaming Study Commission (2010) discuss the economic impact of gambling and problem gambling. Table 14, p. 48 of the latter identifies 11 categories of costs. Grinols (2004b) considers costs from the perspective of externalities and tax policy. Grinols et al. (2011) attempt to link non-casino visitors to crime, while Grinols and Mustard (2006) focus on the impact of casino location, which presumably leads to an increase in casino visitors, and crime. None of these can directly consider the impact of problem gambling on crime.

Smith et al. (2003, p. 91) note that “investigating officers seldom dig deeply enough to determine whether or not occurrences are gambling-related” and gambling-related crime may be “underestimated because the majority of criminal occurrences are unresolved; that is, a crime is reported, a file prepared with the pertinent documentation, but no arrests, let alone a charge or conviction (p. 46).”

Williams and Hahn (2006). The reader is referred to the detailed end-of-survey initial findings and summary contained in Williams et al. (2015).

“This particular area was chosen because one of the original goals of the study was to assess the social and economic impacts of a new gambling venue, the Quinte Exhibition and Raceway – II (QER-II), that was to be built in Belleville in the spring or summer of 2007.The existing horse racing track at the Quinte Exhibition and Raceway (QER) in Belleville was to close and a new horse race track with slot machines (QER-II) would be built.Horse racing did end at the existing QER after the 2006 season, but a new QER-II was never built.When the QER-II did not get built, the present study’s exclusive focus became the natural course and etiology of gambling and problem gambling (Williams 2015, p. 27).”

“The purpose of the At Risk sample was to oversample people at risk of developing gambling problems to better ensure there were a sufficient number of people in the cohort who developed gambling-related problems during the course of the study (Williams 2015, p. 31).”

An observation refers to the vector of survey data collected on a single individual for a single year of the 5-year panel study.

The survey instrument states: “The information you provide will be stored as a computer data file. This file is confidential, and can only be seen by the six members of the international Research Team (headed by Dr. Robert Williams & Mr. Robert Hann).The local Research Team in Belleville has access to your contact information, but not to anything else. All personally identifying information will be erased from the data file once the study is finished.Also, only group results will be reported when the study is published (Williams et al. 2015, p. 179).”

The importance of confidentiality was explained by Williams and Hann (2006): “The strict confidentiality of the information provided will be emphasized. Participants’ data will be automatically converted to an SPSS file with only the Principal Investigators having access to that file. Participants will also be asked to identify two friends/relatives who would be in the best position to verify information the participant provided, as among the 4000 participants a small percentage of randomly selected collaterals may be contacted to corroborate this information. Although it is unlikely that we will actually contact collaterals, the possibility of independent corroboration significantly improves validity of self-report (Babor et al. 1987; Roese and Jamieson 1993; Williams and Nowatzki 2005).”

The reported categoriesof illegal activity (to each item, respondents answered yes or no) are as follows. In each case, we also list in square brackets the distinct number of individuals having performed this illegal activity in one or more of the sample years: “assault (illegal activity done in the past 12 months [46]); sexual assault (illegal activity done in the past 12 months [7]); robbery/mugging (illegal activity done in the past 12 months [5]); break & enter (illegal activity done in the past 12 months [13]); theft (illegal activity done in the past 12 months [46]); shoplifting (illegal activity done in the past 12 months [53]); fraud or embezzlement (illegal activity done in the past 12 months [15]); drug trafficking(illegal activity done in the past 12 months [52]); arson (illegal activity done in the past 12 months [3]); causing a disturbance (illegal activity done in the past 12 months [78]); vandalism (illegal activity done in the past 12 months [26]); impaired driving (illegal activity done in the past 12 months [203]); dangerous operation of a vehicle (illegal activity done in the past 12 months [121]); other (illegal activity done in the past 12 months [122]).”

These numbers can be compared to those of Clark and Walker (2009, p. 122), who report for their sample that “roughly 19% (1169 individuals) of the sample admits to having committed a crime over the past year (Serious crime, our dependent variable).”

Blaszczynski and Steel (1998), Welte et al. (2004), Potenza (2008).Twenty-five to sixty-three percent of pathological gamblers satisfy lifetime substance abuse criteria (Shaffer and Korn 2002).

We also considered income. It was statistically insignificant and made virtually no difference in the results. Because the other demographic variables are exogenous and income is not, we exclude income in the reported regressions.

To avoid linear dependence, one group is dropped.

The panel data is arranged by individual by year. Regressions were run using Stata command xtprobit with robust standard errors, vce(robust).