The Finnish government has recently decided that the “candy tax” currently applied to sugar-rich confectionaries, ice-cream and soft drinks would be eliminated on January 1, 2017. The decision, driven by concerns from the European Commission over the distorting effect of the tax, has re-opened the debate over the desirability of replacing the candy tax with a tax based on the sugar content of each product, i.e., a sugar tax.
That debate has a long history within but also outside of Finland. The incidence of obesity and associated costs of diet-related chronic diseases keep rising in high-income countries, while traditional informational measures seem ineffective to avert that trend. The idea of taxing unhealthy foods, and potentially subsidizing healthy ones, has therefore gained momentum world-wide.
Fat and sugar taxes certainly have an intuitive appeal: due to the law of demand, which is known by economists to hold in all but very particular circumstances, as the price of an unhealthy product rises, quantity demanded falls and, all else constant, diet quality and health improve. As is often the case with food issues, however, this apparent simplicity may be misleading and there are, in fact, multiple ways by which well-intended taxes may be ineffective or even have negative unintended consequences.
The first issue relates to the substitutions that individuals make as a result of the tax. Due to their energy requirements, if people eat less of something, they automatically eat more of something else, but that something else, and its healthiness, are unfortunately unknown a priori. Thus, taxed candies may be replaced by crisps rather than fruits, with uncertain health benefits. I invite those thinking that this is a far-fetched argument to check the academic literature on the subject, and the numerous “perverse effects” that it uncovers. As an example, an early simulation exercise led by the British Heart Foundation found that, paradoxically, “taxing ‘less healthy’ foods could increase cardio-vascular and cancer deaths by 35–1300 yearly”, in that case due to an unanticipated negative impact of the tax on intakes of fruits and vegetables.
The second issue relates to the magnitude of the adjustment and associated equity effects. Typically demand for food is inelastic, meaning that it responds to changes in prices in a less-than-proportional manner. As a consequence, it is widely accepted that “sin taxes” applied to food should be large – at least 20% – in order to have any effect on diet quality and health. But high tax rates on foods are also economically regressive: they hurt the poor the most, because they spend the highest share of their income on the taxed goods. In addition, high tax rates inevitably trigger stark opposition from various interest groups, which raises questions about their political feasibility. The removal of the short-lived Danish fat tax demonstrates that this is not a purely academic argument.
The third issue has to do with the difficulty of implementation and associated costs. As far as sugar is concerned, distinguishing between naturally-present sugar (e.g., in fruits) from added sugar (e.g., in juice from concentrate), the treatment of imported goods, and the identification of the goods subject to the tax create difficulties that are costly to address. As an illustration of the potential bureaucratic headaches that the practical implementation of targeted food taxes may generate, I refer the reader to the famous English court case seeking to establish whether a Jaffa Cake is best described, for tax purposes, as a biscuit or a cake. Fascinating stuff, and, yes, as the name indicates, it is a cake! Anybody who has taken the ferry from Helsinki to Tallinn will also understand that domestic taxes distort international trade and tend to generate black markets.
Even leaving out the minefields of political philosophy and nutritional epidemiology from this discussion, there are many additional issues, for instance related to the transmission of a sugar tax to consumers along an imperfectly competitive food chain, or whether environmental aspects should be taken into account as well. From a pragmatic point of view, however, it is worth reflecting on the latest evaluation of the Danish fat tax which concludes: “The fat tax was associated with marginal changes in population risk of IHD [ischemic heart disease]. One estimate suggests an increased population risk of IHD by 0.2% and the other estimate suggests that the risk of IHD decreased by 0.3%.” Strangely, the authors interpret their results as follows: “Policymakers must therefore be more ambitious in relation to food taxes”. A simpler conclusion would be that this nutrient tax failed as a public health measure.
Of course, one could also point to more favourable pieces of evidence from the literature. The bottom line, however, is that it is far from obvious that a sugar tax, in spite of its intuitive appeal, would improve public health and raise social welfare in Finland. No amount of conceptual debate will provide definite answers either. What is needed is a careful empirical examination of how food consumption and diet quality respond to price signals. I am happy to report that Luke works actively to fill this knowledge gap as part of the EU-wide project on sustainable diets (Era-Net SUSDIET). Results of the analysis of tax scenarios are planned for next year – stay tuned!