Social bond theory has been subjected to countless empirical tests since the publication of Hirschi’s Causes of Delinquency. Despite significant variation in data and methodology, there has been consistent support that weak attachment, a low stake in conformity and weak beliefs in the common values of society increase the risk of deviance and delinquency (Sampson and Laub 1993; Gottfredson 2006; Akers and Sellers 2008). While the control mechanism of involvement has also received significant attention in sociological and criminological investigations, results are less consistent. This has been especially true regarding the role of involvement in sport by young men and women. The current study aimed to undertake a more comprehensive examination of how involvement in sport is associated with delinquency.
Our results offer a nuanced story of the relationship between participation in sports and delinquency. Although moderate participation in sports is associated with a greater risk for beer and spirit consumption compared with individuals who engage in no sports whatsoever, an increase in the hours engaged in sports reduces these risks. Our results suggest that the risk of hard liquor consumption is actually significantly less among youth who spend at least four hours a day engaged in sports compared with non-sports-involved youth. Additionally, marijuana use is less likely among young men engaged in at least three hours or more of sport relative to non-sports-involved young men. This finding adds an important nuance to previously reported associations between sports participation and drugs and alcohol. Namely that while some sports engagement may be associated with more drugs and alcohol, more engagement in sport activities does not necessarily mean more risk. Rather there is a turning point where more engagement actually means a reduced risk.
This is likely explained via a number of mechanisms. First, increased time engaged in sport reduces the number of hours within which drinking and alcohol can take place, in line with classic defences of the social control capacity of involvement (Hirschi 1969). Second, the four-hour turning point identified by this study likely demarcates casual sports enthusiasts from more serious athletes who may weigh the performance costs of alcohol and drug use. Importantly, however, the relationship between the probability of alcohol consumption and sports involvement varied by international context. The protective effects of sports involvement on the probability of alcohol consumption were particularly evident among youth in Anglo-Saxon countries.
The other significant finding of this study regards the oft-cited violence–sports relationship. While more time in sports is not necessarily associated with a greater risk of alcohol and drug use, individuals who engage in any more than two hours of sports a day have significantly greater odds of recent violence compared with students who do not engage in any sports. This relationship held while controlling for individual beliefs in the legitimacy of violence, suggesting that the relationship between sports and violence cannot be explained by the moral reasoning of individuals who self-select in to sport (Nucci and Young-Shim 2005). Supplementary analyses (not displayed, but available from author) confirmed this was true whether we considered minor or more serious forms of violence such as fighting or carrying a weapon. Results also suggest that this is generally true for both males and females, contrary to prior findings suggesting that violence and aggression is more often associated with male sports participants (Miller et al. 2006; O’Brien et al. 2012) and that sports involvement may actually be a protective factor for young women (Booth et al. 2008). Furthermore, surprising gender differences were evident in some country clusters. Perhaps most noticeably, young women involved in more sport had significantly greater predicted probabilities of violence in post-socialist countries compared with young women who did not engage in sports, whereas as young men in post-socialist countries did not experience an increase in the predicted probability of violence.
While results confirm that the sports–violence link generally traverses international political contexts, an exception was found among students in Northern European countries. Although students from this region had among the lowest predicted probabilities for violence overall, their risk did not increase significantly with sports participation, unlike students from every other region. These findings add further weight to arguments for considering gender differences in the importance of social bonds (Belknap and Holsinger 2006; Booth et al. 2008) as well considering national and political context in the application of criminological theory. As a result, this study echoes previous criticism of social bond theory for its inattention to how macro-social changes and changing politi