To measure the impact of a reimbursement-based consumer subsidy on vegetable expenditures, consumption and waste.

Two-arm randomized controlled trial; two-week baseline observation period, three-week intervention period. Participants’ vegetable expenditures, consumption and waste were monitored using receipts collection and through an FFQ. During the intervention period, the treatment group received reimbursement of up to 50 US dollars ($) for purchased vegetables.

Participants were solicited from Palo Alto, CA, USA using materials advertising a ‘consumer behavior study’ and a small participation incentive. To prevent selection bias, solicitation materials did not describe the specific behaviour being evaluated.

One hundred and fifty potential participants responded to the solicitations and 144 participants enrolled in the study; 138 participants completed all five weekly surveys.

Accounting for the control group (n 69) and the two-week baseline period, the intervention significantly impacted the treatment group’s (n 69) vegetable expenditures (+$8·16 (sd 2·67)/week, P<0·01), but not vegetable consumption (+1·3 (sd 1·2) servings/week, P=0·28) or waste (−0·23 (sd 1·2) servings/week, P=0·60).

The consumer subsidy significantly increased participants’ vegetable expenditures, but not consumption or waste, suggesting that this type of subsidy might not have the effects anticipated. Reimbursement-based consumer subsidies may therefore not be as useful a policy tool for impacting vegetable consumption as earlier studies have suggested. Moreover, moderation analysis revealed that the subsidy’s effect on participants’ vegetable expenditures was significant only in men. Additional research should seek to determine how far reaching gender-specific effects are in this context. Further research should also examine the effect of a similar consumer subsidy on high-risk populations and explore to what extent increases in participants’ expenditures are due to the purchase of more expensive vegetables, purchasing of vegetables during the study period that were consumed outside the study period, or a shift from restaurant vegetable consumption to grocery vegetable consumption.