Deconstructing supermarket intervention effects on fruit and vegetable consumption in areas of limited retail access: Evidence from the Seacroft study

Concern about diet and access to healthy foodstuffs is felt worldwide. The introduction of large retailers, with low prices and wide product ranges, to poor access areas has been seen as a solution. We apply quantile regression to data related to one such opening, the Seacroft Intervention Study in the United Kingdom, allowing consideration at different levels of the fruit and vegetable consumption distribution. For residents with easy access to the new store, captured using Ordnance Survey Integrated Transport Network for improved representation of shoppers' journeys, a significant average increase of half a portion per day was found, increasing to 0.7 portions or more for households with no car access. However moving away from the average effects considered in the literature thus far, shopping at the new store is significant only for those at the top end of the distribution and, importantly, not for those whose diets were previously poor. Attitudes to healthy eating, relative cost of fruit and vegetables, and deprivation are shown to be key factors at lower intake levels. Access remains a significant determinant of consumption. Hence we urge caution in accepting the conclusion that new supermarkets can benefit all, and suggest that policy makers should consider more targeted measures to help improve the worst diets.