Health warning labels modify implicit and explicit responses to snack foods through model-free and model-based changes in motivation

Health warning labels (HWLs) show promise in reducing motivation towards energy-dense snack foods. Understanding the underlying mechanisms could optimise their effectiveness. In two studies we compared effects of HWLs and irrelevant aversive labels (IALs) on implicit (approach) and explicit (choice) motivation towards unhealthy snacks. We examined whether labelling effects on motivation arose from the creation of outcome-dependent associations between the food and its health consequences (model-based effects) or from simple, non-specific aversive associations (model-free effects). Both label types reduced motivation towards snack foods but only when the label was physically present. HWLs and IALs showed similar effects on implicit motivation, although HWLs reduced explicit motivation more than IALs. Thus, aversive HWLs affect both model-free and model-based processes, the former through low level associative mechanisms affecting implicit motivation, the latter by emphasizing explicit causal links to health outcomes thereby affecting explicitly motivated choice behaviours.