Defend Truth


Sugar tax will have negligible impact on public health

Ivo Vegter is a columnist and the author of Extreme Environment, a book on environmental exaggeration and how it harms emerging economies. He writes on this and many other matters, from the perspective of individual liberty and free markets.

While the media has “fact-checked” claims made by the beverage industry in opposition to a proposed tax on sugary drinks, claims made by the government in support of the tax have not been so scrutinised. They should be, because according to research the government itself cites to justify the tax, it will achieve very little in the real world, if anything at all.

“The beverage industry was in full force in Sandton this week, fuming about government’s proposed tax on sugary drinks,” reported Gill Gifford for the Health-e News service in Daily Maverick. “But government is standing firm, saying it has a duty to protect people’s health.”

A question few people appear to have considered is whether a tax on sugary drinks will actually protect people’s health.

Writing in Daily Maverick, Africa Check cherry-picked two of the industry’s claims to “fact-check”. It examined the claims that sugar-sweetened beverages (SSBs) account for only 3% of daily calorie intake in South Africa, and that lower sales volumes could cost the country between 62,000 and 72,000 jobs.

The reason for choosing the latter claim is simple: it smacks of exaggeration, it refers to gross job losses in the non-alcoholic beverage industry rather than net changes in employment resulting from shifting spending patterns, and it can a priori be declared “unproven” because it makes a prediction about the future.

Why the 3% claim was chosen is more of a puzzle. Africa Check rated it “misleading”, even though it does not find it to be substantially wrong. Its own calculations lead it to conclude that SSBs account for 3.2% of an average male’s energy intake and 3.9% of that of a female. But even if you round up, to 4%, the beverage industry argument that SSBs account for only a small fraction of our daily energy intake holds up. The reason why Africa Check finds it misleading is because it is an average which actually represents a range of SSB consumption, so there may indeed be people who drink too many sugary drinks for their own good. This is hardly a grand revelation, nor is stating the average misleading.

The Health-e piece quoted a rather hysterical-sounding health minister, Aaron Motsoaledi, telling PowerFM radio: “In fact, they even said the GDP would be reduced and that the economy would collapse.”

No they didn’t. They warned of job losses, not of a reduction in GDP, or an economic collapse. Motsoaledi went on to refer to “people making mega bucks”, in an attempt to harness popular distrust of profit-making private enterprise. Then he called them “economic hitmen who, to protect their own mega-profits, use the poor to assassinate themselves”. Within a paragraph, his crude exaggeration and name-calling degenerated into incomprehensible nonsense.

This kind of bombastic rhetoric does not give one much confidence that the government has a strong, rational case to support a tax on sugar-sweetened beverages. And upon investigation, it turns out that it doesn’t have much of a case at all. Amid all the industry-bashing, Treasury is simply assumed to be acting in good faith, while industry is assumed to act in bad faith.

So let us, for the sake of this argument, put aside Motsoaledi’s insulting caricature of the private sector as pathological liars trying to protect their mega-profits at the expense of the poor. Let us, instead, consider the government’s proposal purely on the merits cited by government itself.

But before we do that, it is worth noting that according to KPMG, a tax on sugary drinks of 20% (which is what the Treasury proposal works out to in the case of Coca-Cola) could net the government as much as R2.17-billion in taxes. Just like industry has a powerful financial motive to oppose taxes, the government has a powerful financial motive in enacting them. Note also that this revenue will go into the general tax pot, instead of being ring-fenced to combat the public health problems that the government claims to be concerned about. It might be spent on the health sector or on public education campaigns, but it is equally likely to end up in corrupt arms deals or multimillion rand mansions for retired politicians.

The Treasury paper claims that a tax on SSBs is the most cost-effective health intervention to combat obesity. Tracking down the academic source of this claim – via several previous government reports – leads to this study from 2010. Surprisingly, perhaps, it does not mention sugar at all. It investigates reducing fat intake, increasing dietary fibre intake, and encouraging physical exercise to combat obesity. It does not consider fiscal interventions (such as taxes or subsidies) on individual foodstuffs, and finds that in general they make only 0.5% difference to fat intake as a percentage of total daily energy intake. The only intervention that proved even less effective was mandatory food labelling. Worse, the study found that a tax on fatty foods makes no measurable impact on body mass index, at all.

The reason why fiscal measures are the most cost-effective dietary intervention is not because they are so effective – they clearly aren’t – but because they cost so little to implement. Sugary drinks are an easy and arbitrarily selected target for an empty gesture designed to make it seem like the government is “doing something” about obesity.

The government claims that its tax measure will reduce obesity in adult males by 3.8% and in adult females by 2.4%. It relies for this claim on a paper that purports to predict the impact of a 20% tax on SSBs, written by Mercy Manyema and Karin Hofmann at Wits University.

The Manyema paper constructs a “mathematical model” based on a chain of assumptions, namely that the tax will be passed on to consumers, that this will reduce SSB consumption, which will change the number of calories consumed, which will change the consumer’s dietary energy balance, which will change their body mass index, which ultimately will lead to lower obesity rates. However, these assumptions are, for the most part, unsupported by empirical evidence.

On the contrary. A review of 880 studies linking economic interventions to changes in diet and physical activity, conducted by scientists at the Behaviour and Health Research Unit at the University of Cambridge, has found little evidence that the Manyema paper’s postulated chain of causation actually holds up in the real world.

“Our findings have exposed a complex, limited and largely equivocal evidence base,” reads the review, “suggesting that the public health case for using economic instruments to promote dietary and physical activity behaviour change may be less compelling than some proponents have claimed. This conclusion provides an important counterpoint to what are, in our view, overly optimistic claims made by some authors of individual primary studies and reviews for the use of economic instruments to improve population health behaviour. It implies a need for caution in the development of public health policies intended to alter economic environmental stimuli to incentivise health-enhancing dietary and physical activity behaviour change at population level.”

Basically, the assumptions don’t hold up in the real world, and basing public policy on such ideas is unwise. But let’s assume that the Manyema paper is correct, and against all odds their model predictions come true.

Using obesity statistics from the South African National Health and Nutrition Examination Survey (SANHANES-1), which is the definitive public health survey in this country, we find that 10.6% of adult males are obese, as are 39.2% of adult females. (Why the Treasury paper uses higher figures derived from a global obesity analysis, other than as a propaganda trick to make obesity look worse than it really is, remains a mystery.)

The Manyema projections will reduce obesity in males from 10.6% to 10.1%, and in women from 39.2% to 38.4%. The way the Treasury paper cited these fractions of a percentage point makes the actual benefit look about five times as high as it really is. For a general public that is unlikely to understand the nuances of statistics, this is a lot more misleading than the beverage industry’s claim that sugary drinks account for only 3% of daily energy intake in South Africa.

Moreover, this reduction in obesity is just an average. If the industry’s 3% claim is rejected as “misleading” because an average conceals a broader range of consumption levels, then surely the government’s obesity reduction claims must be viewed equally critically. The number of obese people would, according to the Manyema model, decrease by “over 220,000” people, but the range is between 24,197 and 411,759. An upper limit that is 17 times as high as the lower limit shows incredible uncertainty in the result.

But again, let’s suppose that 220,000 people do benefit, and will no longer be categorised as obese. This represents a tiny 0.4% of South Africa’s population. Taxing everyone to benefit such a small number of people seems remarkably unjust.

Neither the Manyema model nor the Treasury paper account for possible consumption changes from taxed sugary drinks to untaxed sweetened tea, sugar or cocoa. Perhaps the most ironic switch would be increased demand for pure, unsweetened fruit juice, which is excluded from the tax, but contains just as much sugar as the average soft drink.

But it gets even worse. The Manyema paper finds that if all goes as planned, the tax will reduce average daily energy intake by 36kJ, or 8.6 calories. By comparison, The South African Guidelines for Healthy Eating and Food Guide (sic), published by the Department of Health, specify a daily diet of 10,500kJ (about 2,500 calories) for the average adult male, 8,500kJ (about 2,000 calories) for an adult female, and 6,500kJ (about 1,500 calories) for a child. So a 36kJ reduction would make only a 0.34% difference in daily energy intake for adult males, 0.42% in adult females, and a 0.55% difference in children.

To put this tiny change in perspective, the Manyema paper notes that a 94kJ change in energy intake is needed to effect a 1kg change in body weight. That means that the average weight reduction, if all goes as predicted, will be 383 grams, which is hardly detectable on a bathroom scale. The only reason 220,000 people might no longer be classified as obese is because they will move from fractionally above the official obesity line to fractionally below it. In reality, even supposing that all the assumptions in the Manyema paper hold up, the predicted impact on public health will be negligible.

Instead of taking numbers out of context for propaganda purposes, one must read statistics in context. If one does so, the very model on which the government relies to justify its tax on sugary drinks concludes that it will have an insignificant impact. Other research suggests that even this conclusion is unduly optimistic, and there may not be any measurable effect on public health at all.

Whatever the ultimate impact on the beverage industry and the suppliers and retailers that rely on it, enacting a tax that has no public health benefit is unjustified. It is simply an arbitrary tax grab that exploits poor people who tend to seek out drinks with high energy content. Last I checked, government was supposed to be improving the lives of the poor, not robbing them blind.

Do people even want the government to improve their lives by trying to meddle with their diets? The media’s bloviating on the subject comes from relatively wealthy elites who have the luxury to be fussy about their diets and enjoy the leisure to prepare high-quality meals for themselves and their families. The poor, whom the rich so casually judge, are not so blessed.

The SANHANES-1 study found that 24.9% of South Africans were classified as obese, and another 22.5% are overweight, based on body mass index. But although that makes almost half of all South Africans officially fat, the survey found that only 15.7% of the population report being unhappy with their weight. Clearly, personal beliefs about health, body image and body weight are not the same as those of government officials or the judgemental elite.

The public may well come to resent the nanny state’s interference in a private matter they don’t perceive to be a serious problem, using taxes that achieve hardly any measurable outcome other than milking them for money to pour into Treasury coffers. And whether or not they sympathise with the beverage industry and those who depend on it, this public resentment would be justified. After all, even the government’s own evidence shows that a tax on sugary drinks will not result in any significant public health benefits. DM

Note: This column relies on research commissioned by the South African Institute for Race Relations (IRR) after my previous column on this subject was published. It was used to inform the IRR’S submission to Treasury on its policy paper on an SSB tax. The research and opinions expressed in this column are entirely my own, however.

Clarification: Africa Check claims I misrepresented their finding about the beverage industry’s claim that only 3% of a person’s energy intake is accounted for by SSBs. Although the first stated concern – which I criticised – is that an average represents a range of consumption, Africa Check also noted that children under 15 are not included in the average at all. This is true. As they write, there is no data on SSB consumption by children under 15 in South Africa. Neither the beverage industry nor the government (in relying on the Manyema paper) takes SSB consumption by children into account.

However, extrapolating from the chart Africa Check presented out of the Manyema paper, a second-order approximation would put consumption among children under 15 at 213.5ml per day. As a fraction of daily recommended energy intake, this works out to 5.9%. That means a tax could have a slightly higher effect in children than the 36kJ per day energy intake reduction estimated by Manyema and Hofmann for adults, reducing daily energy intake in children by 60kJ per day.

This would represent 0.7% of the total recommended energy intake in children under 15 and 0.9% of that in children under 10.

These are still very small numbers, so in the absence of empirical data I remain unconvinced that an SSB tax will have a noticeable effect upon public health, even in children, and do not find the claim that SSBs represent only a small fraction of the average South African’s daily energy intake to be “misleading”.


Please peer review 3 community comments before your comment can be posted