Food tax policies are good at raising revenues; reducing health inequalities should involve food supply policies.

Smed and Robertson raise a burning issue as most countries are considering the relevance and the cost-effectiveness of a tax policy on unhealthy food. They highlight the uncertainties of a food price policy in Denmark, ranging from misunderstanding of consumer responses to uncontrolled reactions of private firms, and complicated by the lack of an evaluation. Clearly, more studies are needed to assess the health efficiency of a food tax. Moreover, we think that the current debate on food tax policies has not paid enough attention to the role of supply and to the major issue of health inequalities.

Taxes are made for revenues, health is only an uncontrollable side effect. It seems that the main—or even only—advantage of a food tax policy relies on the huge amounts of money it may generate in a context of restricted budgets and expanding health costs. For that sole purpose and without any nutritional aims nor use for health policy funding, taxes on soft drinks and some snack foods were introduced in most US states for over 50 years. In line with this position, a tax on soft drinks was introduced this year in France, with the aim of raising revenues to lower agricultural labour costs. With more focus on nutritional purposes by supposedly directing revenues to health costs, and by anticipating changes in consumer behaviour, Hungary (1) and Denmark (2) provide the first large-scale experiment fields for taxes on unhealthy foods.

As suggested by Smed and Robertson regarding the Danish experience, while food taxes are good at generating revenues, favourably influencing consumer diet and heath behaviour through food prices is a bold attempt. Both sufficient sensitivity to prices and healthy food substitutions are required to improve food choices with a food policy, and none of them were demonstrated. In addition to the limitations of a food tax policy already raised by Smed and Robertson, other concrete difficulties must be acknowledged such as the high variability of responses to a variation in food prices, depending on the population and the level of aggregation of taxed products. Moreover, to get a real change in behaviour and impact the diet, the policy should concern a wide range of foods, including foods consumed away from home. The main justification for a food tax is to modify the relative price hierarchy of foods to facilitate the fulfilment of dietary recommendations. Indeed, the current structure of food prices does not favour healthy food choices because foods with good nutrient profile are generally more expensive than less healthy foods, and the price gap is huge: 1 calorie from fruit and vegetables is 5 times more expensive than a calorie from other foods(3). Therefore tax levels must be sufficiently high (20 to 50%) to counteract the unfavourable food price hierarchy, but this will more strongly affect the poorest. Worsening social inequalities through a nutritional tax is certain because of a higher budgetary share of food among low income consumers. Consequently, unhealthy food taxes meet the solid prospect of generating resources with a disproportionate burden on lower socioeconomic consumers, and as a probable side effect, possible gains or losses in health. In addition, it is not because a Danish study found that poorer consumers were more responsive to prices (4), that the evidence is strong enough to generalize such conclusions. Very few works studies took into account different levels of income and/or education in the sensitivity of consumption or health indicators to food prices. The regressive impact of food taxes is often acknowledged in the literature, but the inequality issue embedded in taxation is not sufficiently raised up. The hypothetical higher reduction of disease among poor people linked to their higher morbidity claimed by Mytton et al.(5) to justify the tax can be reached more adequately through other policies. Though worsening (see last figures of OBEPI surveys on obesity in France), inequality has not been a major issue; it should be a key factor when designing health policies (6).

How can we integrate the reduction of health inequalities in the design of food policies? Targeting the lower income population, which is more at risk in terms of morbidity and unhealthy food choices, through specific policies is possible (subsidies in form of coupons and lower food prices can be implemented in specific circuits such as social groceries; targeted information and education programs), but it can be also reached with general policies.

Emphasis has been placed mainly on the responsibility of consumers. Let us share it with the food industry and involve the supply side. Recent evidence from both France (7) and the US (8) shows that the food offer matters, with significantly higher obesity rates found among shoppers in supermarkets selling low-cost foods (even after adjustment for many individual and contextual variables). Generalizing the implementation of food taxes will add to a trend of increasing food prices, pushing the poorest to look for cheaper options, or to cut off healthy foods budget. It will encourage the segmentation of the market between branded goods and low-cost substitutes. Yet, we have insufficient evidence to affirm that there is no link between the price of a food and its nutritional quality, within a given category of food products(9;10). This may well depend on the level of processing of the products. We therefore advocate for more studies on the relationship between price and food quality, conducted with similar methods and definitions. Public intervention should then focus on strong incentives for the development of good quality low cost products, in order to make sure that a varied range of products with good nutritional quality is accessible on the market. This can be achieved through voluntary agreements of private firms, which have begun in UK and France with encouraging results. Large scale distributors may be engaged in nutritional improvement of generic brands representing an important volume of sales (http://www.sante.gouv.fr/les-chartes-d-engagements-de-progres-nutritionn...). To speed up favourable food reformulation, why not use more coercive means, through regulation for instance? Due to the wide market it reaches, added to its capacity to target mostly lower socioeconomic groups, the intervention on suppliers seems to be an efficient tool that has to be added to specific programs on the consumer side. Let us prefer a policy with a large scale impact and a positive effect on inequalities, to a regressive nutritional tax.

1. Holt E. Hungary to introduce broad range of fat taxes. Lancet 2011;378:755.
2. Smed S. Financial penalties on foods: the fat tax in Denmark. Nutr Bull 2012;37:142-7.
3. Darmon N, Darmon M, Maillot M, Drewnowski A. A nutrient density standard for vegetables and fruits: nutrients per calorie and nutrients per unit cost. J.Am.Diet.Assoc. 2005;105:1881-7.
4. Smed S, Jensen JD, Denver S. Socio-economic characteristics and the effect of taxation as a health policy instrument. Food Policy 2007;32:624-39.
5. Mytton OT, Clarke D, Rayner M. Taxing unhealthy food and drinks to improve health. BMJ 2012;344:e2931.
6. Marmot M, Allen J, Bell R, Bloomer E, Goldblatt P. WHO European review of social determinants of health and the health divide. Lancet 2012;380:1011-29.
7. Chaix B, Bean K, Daniel M, Zenk SN, Kestens Y, Charreire H et al. Associations of supermarket characteristics with weight status and body fat: a multilevel analysis of individuals within supermarkets (RECORD study). PLoS.One. 2012;7:e32908.
8. Drewnowski A, Aggarwal A, Hurvitz PM, Monsivais P, Moudon AV. Obesity and supermarket access: proximity or price? Am J Public Health 2012;102:e74-e80.
9. Cooper S,.Nelson M. 'Economy' line foods from four supermarkets and brand name equivalents: a comparison of their nutrient contents and costs. J.Hum.Nutr.Diet. 2003;16:339-47.
10. Darmon N, Caillavet F, Joly C, Maillot M, Drewnowski A. Low-cost foods: how do they compare with their brand name equivalents? A French study. Public Health Nutr 2009;12:808-15.

Competing interests: None declared