Does reducing the serving size of beer influence how much students drink on a night out? A randomised controlled field experiment

Public health researchers have recently suggested that alcohol consumption may be curbed by reducing the default serving size of alcohol (i.e., simultaneously decreasing the size of serving containers and the standard volume served), as this has been shown to influence the overall quantity of alcohol poured, perceptions of portion size, drinking pace, and purchasing/consumption behaviour. In a randomised controlled field-experiment, we examine the effect on alcohol consumption from reducing serving size, i.e., simultaneously reducing the size of beer glasses and the volume of beer served by 20% (from 50 cl to 40 cl) in a student bar. The results show no significant differences in consumption between the treatment and control conditions. Participants who received the smaller serving size drank 102.04 cl of beer per person while participants who received the standard serving size drank 101.42 cl of beer per person (N = 102, p = .95). This outcome appears to be the result of participants served with the smaller serving size drinking significantly more servings of beer than participants served with the standard serving size (p = .02). Thus, the intuitively attractive strategy of decreasing serving sizes to nudge people towards a decrease in consumption of alcohol does not seem to work in as straightforward a manner as sometimes suggested by existing research. To explain this result and its implications, we discuss the strengths and limitations of the experiment and suggest further research avenues.