The effect of food taxes and subsidies on population health and health costs: a modelling study

Tony Blakely,Christine Cleghorn,Anja Mizdrak,Wilma Waterlander,Nhung Nghiem,Boyd Swinburn,Nick Wilson,Cliona Ni Mhurchu

doi:10.1016/s2468-2667(20)30116-x

Abstract

One possible policy response to the burden of diet-related disease is food taxes and subsidies, but the net health gains of these approaches are uncertain because of substitution effects between foods. We estimated the health and cost impacts of various food taxes and subsidies in one high-income country, New Zealand. In this modelling study, we compared the effects in New Zealand of a 20% fruit and vegetable subsidy, of saturated fat, sugar and salt taxes (each set at a level that increased the total food price by the same magnitude of decrease from the fruit and vegetable subsidy), and of an 8% so-called junk food tax (on non-essential, energy-dense food). We modelled the effect of price changes on food purchases, the consequent changes in fruit and vegetable and sugar-sweetened beverage purchasing, nutrient risk factors, and body-mass index, and how these changes affect health status and health expenditure. The pre-intervention intake for 340 food groups was taken from the New Zealand National Nutrition Survey and the post-intervention intake was estimated using price and expenditure elasticities. The resultant changes in dietary risk factors were then propagated through a proportional multistate lifetable (with 17 diet-related diseases) to estimate the changes in health-adjusted life years (HALYs) and health system expenditure over the 2011 New Zealand population's remaining lifespan. Health gains (expressed in HALYs per 1000 people) ranged from 127 (95% uncertainty interval 96-167; undiscounted) for the 8% junk food tax and 212 (102-297) for the fruit and vegetable subsidy, up to 361 (275-474) for the saturated fat tax, 375 (272-508) for the salt tax, and 581 (429-792) for the sugar tax. Health expenditure savings across the remaining lifespan per capita (at a 3% discount rate) ranged from US$492 (334-694) for the junk food tax to $2164 (1472-3122) for the sugar tax. The large magnitude of the health gains and cost savings of these modelled taxes and subsidies suggests that their use warrants serious policy consideration. Health Research Council of New Zealand.

Highlights

Many countries have implemented taxes on sugary beverages,[1] but few have implemented food subsidies or taxes on other products
HALYS are similar to quality-adjusted life-years in that they are calculated prospectively in a simulated population, but they use disability weights from the Global Burden of Disease (GBD) rather than other forms ofutility weights
Our previous direct analyses[7] of the virtual supermarket experiment did not find a statistically significant impact of a fruit and vegetable subsidy on the overall healthiness of foods, but this does not preclude some effect on health-adjusted life years (HALYs), especially given that the price elasticities matrices used in the current study more comprehensively handle substitution and complementary effects

Summary

Introduction

Many countries have implemented taxes on sugary beverages,[1] but few have implemented food subsidies or taxes on other products. Mexico implemented an 8% tax on non-essential, energy-dense foods in 2014, which yielded a 6·0% reduction in purchasing of these foods.[1] In 2011, Denmark implemented a €2·14 per kg tax on saturated fat for products with more than 2·3 g per 100 g of saturated fat. This tax was subsequently repealed in 2013, for the duration of implementation, saturated fat purchases reduced by 4·0%, and deaths attributable to non-communicable disease were estimated to reduce by 0·4%.2. One can use estimated price elasticities, including cross-price elasticities that capture the substitution and complementary effects of a change in price of one food on the purchasing of another food.[5,6]

Methods

Results

Conclusion