Driving carbon offset donations: Evaluating the effectiveness of nudging, framing, and Nudge+ techniques

Human-generated carbon emissions have become a major driver of climate change, with increasing e-commerce activities contributing significantly to this issue. One possible way to address this issue is through carbon offset donations. While ‘nudge’ behaviour change techniques are promising, they have also been critiqued. We conducted an exploratory online experiment with four e-commerce checkout prototypes to whether ‘Nudge+’ approaches (which combine reflection and automatic processing) are more effective than default nudge or information provision to promote offset donations. Participants (n=184, members of a UK university) were assigned to one of four groups: Control (no treatment), Default (pre-selected options), Information, and Nudge+ (combined default and information). The findings reveal that Defaults and Nudge+ significantly increase donation rates compared to the Control, but Nudge+ was no more effective than Default alone. Information alone did not significantly impact donations. Positive attitudes toward carbon offsets and higher concern about climate change did not moderate intervention effects. This study highlights the dominance of Type 1 processing while acknowledging the added value of providing reflective opportunities, offering insights for designing effective behavioural interventions to promote sustainable e-commerce practices. Notably, there is no need to inhibit Type 2 reasoning to enhance nudge effectiveness, as individuals predominantly rely on Type 1 processing. However, participants apparently valued the opportunity to engage in Type 2 reasoning, suggesting that a blend of automatic nudges with reflective information can be effective. Further research is needed to generalise these findings across broader populations and settings. • First study to test Nudge+ intervention in the e-commerce sector • Default offset was highly effective at increasing offsetting donations • Nudge+ was no more effective than default alone • Intervention effects were not moderated by prior climate or offsetting attitudes