Nutrition labels will only have a beneficial effect on the population’s health status if they actually affect purchases, and only to the extent that more healthful choices in one product category are not offset by less healthful choices in another category. Such effects should not only be investigated by consumer surveys, where answers are often subject to a social desirability bias, but also by analysing actual sales data. This work package studied the effects of labels on dietary intake.
Consortium partners used a combination of three types of data: sales data on the sales of specific products in supermarkets, product data containing information on product composition and on the label that the product carries, and personal data on characteristics of the individual shoppers. This allowed an analysis to be made on how the introduction of nutrition labels on food products, changes in the format of nutrition labels, and reformulations of products as mirrored in the label information affected product choices, composition of shopping baskets, and overall patterns of dietary intake.