Abstract
Previous research on pro-environmental behaviours has shown that nudges can be effective in changing behaviour in some settings. However, to be a useful policy tool, nudges should have stable, predictable impacts on targeted behaviours.
In this paper, we use a randomized experiment to test one specific nudge, namely the communication of a descriptive social norm. The paper examines whether the effects of such a nudge on stated preferences for household recycling are indeed stable and predictable in terms of household willingness to pay for waste collection contracts requiring more recycling effort by households.
Three aspects of the social norm are varied: its absolute level, the geographic proximity, and whether relative performance information is included. We find that while respondents generally respond positively to social norm information, the effect is not necessarily monotonic, with the highest social norm levels not necessarily being the most effective, in fact discouraging some respondents from sorting.
In addition, whether and how respondents react to the social norm information strongly depends on their current engagement level, with respondents who sort a lot reacting in a negative manner.