Default options, incentives and food choices: evidence from elementary-school children

To examine whether requiring children to place fruits and vegetables on their lunch trays increases consumption of these items.

Observational study that exploited naturally occurring variation between two school districts and a pre–post observational study at schools that changed their lunch policy mid-year.

Fifteen elementary schools from two school districts, one requiring students to place a fruit or vegetable on their tray and one that does not. In addition, three schools that implemented a default option part way through the school year.

Students at eighteen elementary schools (41 374 child-day observations) across the two experiments.

Requiring that fruits and vegetables be placed on each child's tray increased the fraction of children who ate a serving of fruits or vegetables by 8 percentage points (P < 0·01) but led to an extra 0·7 servings being thrown away per lunch served (P < 0·01). The default option approach cost $US 1·72 to get one additional child to eat one serving of fruits and vegetables for 1 d. However, when default options were combined with a small rewards programme the efficacy of both interventions increased.

A default option, as a stand-alone programme, had only a limited impact on fruit and vegetable consumption but was much less cost-effective than other approaches. Schools requiring children to take fruits and vegetables with their lunch might consider adopting additional interventions to ensure that the additional items served do not end up being thrown away.