Confounding Bias in the Relationship Between Problem Gambling and Crime

Christopher R Dennison , Jessica G Finkeldey , Gregory C Rocheleau

Although the relationship between problem gambling and criminal behavior has been widely researched, concerns over the causal nature of this association remain. Some argue that problem gambling does not lead to crime; instead, the same background characteristics that predict problem gambling also predict criminal behavior. Yet, studies suggestive of a spurious association often rely on small, non-random, and cross-sectional samples; thus, the extent to which the findings are 4 generalizable to the broader population is unknown. With this in mind, the present study uses data from The National Longitudinal Study of Adolescent to Adult Health and a series of propensity score weighting and matching techniques to examine the role of confounding bias in the relationship between problem gambling and criminal behavior in young adulthood. On the surface, results show a positive and significant relationship between problem gambling and a range of criminal behaviors. However, after statistically balancing differences in several background measures between problem gamblers and non-problem gamblers, such as low self-control, past substance use, and juvenile delinquency, we find no significant relationship between problem gambling and crime. These patterns are consistent across several propensity score weighting and matching algorithms. Our results therefore parallel those in support of the "generality of deviance" framework, whereby a similar set of covariates known to be associated with criminal behavior also predict problem gambling.