Taxing and labeling unhealthy food: Making public policies more effective

Background: Taxing and labeling unhealthy food and drinks have been proposed to shift food choices towards healthier items and address obesity rates. Preliminary evidence suggests such methods could be minimally effective depending on how they are structured.

Methods: We constructed a snack shop in a large, urban public hospital. Customers could purchase healthy and less healthy beverages and snacks from October - December 2011. We examined experimental conditions testing relative influence of price changes via a 30% tax on less healthy items; labels highlighting less healthy products; and labels indicating the product is taxed because it is less healthy. The five experimental conditions were: baseline (A); tax only (B); “less healthy” labels (C); tax plus “less healthy” labels (D); and tax plus “less healthy” labels with details of tax highlighted on the label (E). We recorded every purchase: 2,151 unique sales transactions.

Results: At baseline, 47% of items sold were not healthy. Analysis is underway and will be complete before the meeting. We will examine sales data to indicate differences in food purchases, beverage purchases, composition of transactions, total price per transaction, calories, and fat. We will be able to determine whether taxes affected sales, signage alone impacted sales, and whether tax salience improved the effect.

Conclusions: This experiment allows us to see results not only of taxing less healthy items, but of highlighting the taxes in different methods suggested by behavioral economics research. We will tease out whether taxes, labeling and/or tax salience could shift purchases towards healthier items.