Does the range of ecolabels on offer alter their impact on meal selection? An online randomised control trial
Biggs E., Garnett EE., Pechey R.
Background: Environmentally sustainable meals might be more likely to be selected when ecolabels indicating environmental impact are present, but the effectiveness of ecolabels may depend on the range of meal options available (e.g. their comparative environmental impact). Methods: We ran a proof-of-principle study to test whether changing the range of ecolabel values (A-E: lowest – highest impact) shown in hypothetical meal selection scenarios affected sustainable meal choices. In a randomised controlled trial, all participants were presented with five themed meal selection scenarios, which contained three realistic meal options (e.g. a vegetarian, chicken, or beef burger) from which they selected their preferred option. We randomised a sample of the UK population (n = 2646) to a control (no ecolabels present) or intervention (ecolabels present) group. Participants in the intervention group were randomised to see different combinations of ecolabels assigned to each meal, in each scenario. For example, one participant could see ecolabels ‘A’, ‘C’ and ‘E’ assigned to meals in one scenario, whilst another may see ‘B’, ‘C’, ‘D’. Results: We found that whilst ecolabels discouraged selection of the least sustainable meals in comparison to control (no ecolabels present), changing the range of ecolabels on offer had little effect on sustainable meal selection. The only exception was that participants who saw ‘C’ ecolabels uniformly assigned to each meal had significantly lower odds of choosing a sustainable item than participants who saw the widest range of ecolabels (‘A’, ‘C’ and ‘E’) (OR for selecting the lowest impact product (vs. medium or highest impact products): 0.71, CI: 0.58–0.88, p = 0.001). Conclusions: We found that the widest range of ecolabels is more effective than the narrowest, with limited evidence on the impact of intermediate ranges. Trial registration: Registered 6/6/2023 on the Open Science Framework (https://osf.io/69xct/).