Targeted Beverage Taxes Influence Food and Beverage Purchases among Households with Preschool Children.

BACKGROUND: How beverage taxes might influence purchases of foods and beverages among households with preschool children is unclear. Thus, we examined the relation between beverage taxes and food and beverage purchases among US households with a child 2-5 y of age.

OBJECTIVES: We examined how a potential tax on sugar-sweetened beverages (SSBs), or SSBs and >1% fat and/or high-sugar milk, would influence household food and beverage purchases among US households with a preschool child. We aimed to identify the lowest tax rate associated with meaningful changes in purchases.

METHODS: We used household food and beverage purchase data from households with a single child who participated in the 2009-2012 Nielsen Homescan Panel. A 2-part, multilevel panel model was used to examine the relation between beverage prices and food and beverage purchases. Logistic regression was used in the first part of the model to estimate the probability of a food/beverage being purchased, whereas the second part of the model used log-linear regression to estimate predicted changes in purchases among reporting households. Estimates from both parts were combined, and bootstrapping was performed to obtain corrected SEs. In separate models, prices of SSBs, or SSBs and >1% and/or high-sugar milk, were perturbed by +10%, +15%, and +20%. Predicted changes in food and beverage purchases were compared across models.

RESULTS: Price increases of 10%, 15%, and 20% on SSBs were associated with fewer purchases of juice drinks, whereas price increases of 10%, 15%, and 20% simulated on both SSBs plus >1% fat and/or high-sugar milk (combined tax) were associated with fewer kilocalories purchased from >1% fat, low-sugar milk, and meat, poultry, fish, and mixed meat dishes.

CONCLUSIONS: Our study provides further evidence that a tax on beverages high in sugar and/or fat may be associated with favorable changes in beverage purchases among US households with a preschool child.