Abstract:
This paper investigates whether information costs prevent consumers from fully optimizing their purchase decisions with regards to available nutritional information. We design a market-level experiment in order to analyze the effect of nutritional labels that make information more salient and easier to process. Average treatment effects of this intervention are estimated in a difference-in-differences approach, utilizing store-level scanner data. Our results suggest that consumer purchases are affected by implemented labels. In particular, we find that a shelf label of no trans fat significantly increases sales of targeted products. Consumers also seem to have taste preferences and perceive nutrients differently. Low fat labels significantly reduce quantity sales in some treatments, while correlated low calorie labels do not result in the same effect. Adding the FDA disclaimer reduces sales of treated low fat products even further. Consumers further seem to make only limited inferences about unlabeled products and their nutritional quality. Finally, a decrease in overall category sales for treatment periods and treatment stores might indicate additional quantity adjustments for intended purchases.