Skip to main content

Sugar tax : some hard questions

A sugar tax, or more correctly, a soda tax, is the talk of the town and the darling of the media these days. Far be it for me to spoil the party, but I’d like to ask some questions of the proponents of a soda tax to try and get the full picture and not just the sloganistic headlines.

#1. Given the extensive data that we have to hand on national patterns of food and nutrient intake, all of which can be readily linked to bodyweight or body mass index (BMI), where can I find the relevant report that shows that, of all the foods that might be taxed to curb obesity, soda is the clear target?

The simple answer is that in the myopic propaganda on soda tax, no such extensive analysis exists that I know of. So, we are not taxing savoury snacks, fast foods, sugar or chocolate confectionery or booze. Just soda. And we are doing so simply because it is now a global fashion. Of the foods just mentioned, how do policy makers know their average contribution to energy intake, the % of the population who consume them and the intake of these foods among consumers only? How do policy makers compare econometric models for the application of a tax on such foods? The answer to both questions is that such data has not been gathered and therefore is beyond policy makers. They must just trust the soda tax lobby.

#2. Does a soda tax have universal or limited benefit?

The answer is that only those who consume sugar-sweetened beverages feel any putative impact of a soda tax.  In Ireland, that means that 60% of the population might benefit. So 40% of the Irish adult population is unaffected. Nobody on the planet can possibly argue that obesity resides solely among those who consume sugar-sweetened beverages. So the excess weight problem of those who are non- consumers of sugary sodas is ignored. Politicians, please note how many votes you are ignoring in your zeal for soda taxes as presently formulated.

#3. What would be a measure of success of a soda tax?

There are many end points that could be chosen to measure the success of such a tax and in this blogger’s view, policy makers have right to know (a) what would constitute a success and (b) within what time frame. The most likely measure to be proposed is a fall in the purchase of sugary sodas. Since sodas account for a fraction of sugar intake (calculated by my students as 8% of added sugars for the average of Irish consumers) one would not expect to see a dramatic fall in energy intake arising for a soda tax. Moreover, all of the biological evidence at hand suggests that at least some, if not most, of this small drop in energy intake would be compensated for by increased energy intake from some other food sources. Since the purpose of the proposed soda taxes is to combat obesity, do policy makers not have the right to ask the question: “Notwithstanding your excellent data on altered purchase patterns of soda post the tax, could you please give me data on (a) energy intake changes and (b) BMI changes”? The short answer is that such data are not likely to be forthcoming because the focus must be on those who, at the time of the introduction of the tax, were consumers of sugary sodas. There is no point in national average values since not everyone is equally affected by these taxes. As regards time frames, the policy-makers are never told when would be a fair time to ask the hard questions of success or otherwise.

#4. Is there some unwritten and unuttered expectation among the pro-soda tax lobby, that this is just the start and other putative contributors to obesity will soon be targeted as taxable candidates?

One would be naïve not to believe that next we’ll have fast food or sugar and chocolate confectionery etc. as targets for taxation. If that is the case, might it be too much to ask for some honesty here and to re-run all the existing econometric models in which all possible future combinations are included? What combinations would be least regressive? What combinations would have the greatest impact on energy intakes taking all reasonable consumptions into account? What might the economic elasticity of demand look like for various combinations of taxable food categories? In this bloggers view, multiple taxation strategies are just around the corner. This follows logically from the view of the high priests of public health nutrition that the regulatory environment is the way forward, based on prior experience with seat belts in cars, to tobacco, to alcohol and so on. For sure we need a regulatory environment but we also need a bottom up community involvement, one that is well resourced financially, is multi stakeholder driven and which takes a medium to long-term perspective. An example of this is EPODE community projects on obesity (

#5. If sugar intakes fall as a result of single or multiple taxes with a sugar focus, what is the likelihood that the % energy from fat will rise and does that not pose a serious question that merits an answer before we go down the avenue of food taxes to tackle obesity?

There is a phenomenon known as the sugar-fat seesaw. It is poo-pooed by the sugar tax lobby as some odd artifact of data analysis or whatever. The hard experimental data say otherwise. If you covertly or otherwise remove sugar calories from the diet, they will be replaced in part or total by fat calories[1]. The energy density of the diet will rise (fat has 2.25 times the calories of sugar on an equal weight basis) and energy intake will increase among the many who are susceptible to such high-energy diets [2],[3].

NOTE: These are just some of the arguments that will appear in my new book, due for release in April: ”Ever seen a fat fox: Human obesity explored”

[1]Markey O et al (2015) Energy compensation following consumption of sugar-reduced products: a randomized controlled trial. Eur J Nutr. 2015 Sep 9. [Epub ahead of print]
[3]  Hall et al (2015) Calorie for Calorie, Dietary Fat Restriction Results in More Body Fat Loss than Carbohydrate Restriction in People with Obesity. Cell Metab. 2015 Sep 1;22(3):427-36


Popular posts from this blog