The separate effects of self-estimated and actual alcohol intoxication on risk taking: A field experiment.

Abstract

Many risky actions are carried out under the influence of alcohol. However, the effect of alcoholicintoxicationoverthewillingnesstotakerisksiscomplexandstillremainsunclear.Weconductaneconomicfieldexperimentinanatural,drinkingandrisk-takingenvironment to analyze how both actual and self-estimated blood alcohol concentration (BAC) levels influencesubjects'choicesovermonetarylotteries.Ourresultsrevealanegativeimpactof both actual and self-estimated BAC levels on risk-taking. However, for male and young subjects, we find a positive relationship between BAC underestimation (a pattern of estimation error which mainly occurs at high BAC levels) and the willingness to choose riskierlotteries.Ourfindingssuggestthatariskcompensationmechanismisactivatedonly when individuals' own intoxication level is consciously self-perceived to be high. We concludethereforethathumanpropensitytoengageinriskyactivitiesundertheinfluence of alcohol is not due to an enhanced preference for risky choices. In addition to the suggestion in the existing literature that such propensity is due to a weakened ability to perceiverisks,ourresultsindicatethatanimpairedself-perceptionofownintoxicationlevel mayalsobeanimportantfactor.

The influence of groups and alcohol consumption on individual risk-taking

“…However, research conducted in field (natural drinking environments) produce contrasting findings to labbased studies. For example, when tested in natural drinking contexts, intoxication levels (via BAC) are found to be negatively associated with risk-taking behaviour (Lyvers et al, 2015;Proestakis et al, 2013), whereas lab-based studies have often found elevated risk-taking following alcohol consumption (George et al, 2005;Rose et al, 2014) which, heightens with increasing alcohol dose (Lane et al, 2004). Such inconsistencies point to the potential effect of context on either risky behaviour and/or the effects of alcohol.…”

“…Nevertheless, a difficult problem to overcome when measuring risk-taking behaviour in a controlled fashion is the ability to utilise a task with outcomes which participants will perceive as being genuinely negative, thus making their behaviour (at least seem) risky (Dohmen et al, 2011). Tasks in which bonus participant money is gained and lost attempt to address this issue (e.g., Burian et al, 2002;Hopthrow, Abrams, Frings, & Hulbert, 2007;Proestakis et al, 2013).…”

“…However, in contrast, field studies suggest that increased BAC is associated with lower levels of risk-taking behaviour on gambling tasks (Lyvers et al, 2015;Proestakis et al, 2013). This raises questions regarding the extent to which laboratory experiments assess behaviours comparable to those occurring in real world contexts, and may point towards further influencing factors in naturalistic settings, which to date have been absent in lab-based contexts.…”

“…Moreover, it has been noted that care should be taken with the complexity of gambling tasks as participants have previously revealed difficultly in understanding such tasks (Newall, 2017) while intoxicated (Dave, Eckel, Johnson, & Rojas, 2010;Proestakis et al, 2013). To this end, research investigating influence of alcohol on risky gambling behaviour provides some insight into the complexity of risk-taking, and alludes to the importance of establishing measures both understandable, and akin to real-world behaviours to better represent intoxicated gambling 'in the wild'.…”

“…In these environments, alcohol is consumed in social contexts, where the influence of peers may play a contributing role to the observed increase in potentially harmful behaviours. With this in mind, variations between field and laboratory experimental findings (Lane et al, 2004;Lyvers et al, 2015;Proestakis et al, 2013;Richards, Zhang, Mitchell, & de Wit, 1999), may in part be due to contributing influences from drinking contexts (e.g., peers in the social environment).…”