Evaluation of a community-based food waste campaign using a national control group.

Communities are increasingly interested in bolstering sustainability by implementing local campaigns to reduce wasted food and divert it from landfills. Evaluation can be challenging, however, as community-engaged interventions may reach all community members, making it difficult to find an appropriate control group. We leverage a recently validated online survey instrument with samples from both the treated community of Upper Arlington, Ohio, USA, and from the United States at large to provide an additional mode for assessing community-based campaign efficacy. We find that the amount of wasted food reported by Upper Arlington households declined by 23% after a multi-modal local implementation of the 'Save More Than Food' campaign while the national sample reported a 29% increase in wasted food over the same period with the 52% net difference between these trends being statistically significant. A contemporaneous curbside audit of Upper Arlington households revealed a 17% reduction in wasted food and a 30% reduction in inedible food scraps where only the latter pre/post campaign reduction was statistically significant and no parallel national curbside audit data was available. There were few significant differences across neighborhoods that received differential intensities of campaign elements, which emphasizes the importance of identifying and conducting parallel measurement in a control group. The inclusion of the parallel national control group survey provided a cost-effective means to improve the accuracy and robustness of local campaign evaluation. We also discuss the campaign's effects on awareness, attitudes, composting behaviors, and non-organic waste rates.