The sugar tax straw man

Study not inappropriate for what it was meant for

Aviva Tugendhaft, deputy director of Priceless SA, a unit of the University of the Witwatersrand’s school of public health, contacted City Press in relation to quotes attributed to her in the story below.

The description of the research conducted by her and colleagues as “inappropriate” is only true in relation to the way the soft drink industry’s commissioned research has adopted parts of it, she said.

For the purpose it was meant for, the study is completely appropriate, she said.

“The elasticities we used were certainly appropriate for the modelling we did to show the impact the 20% tax would have on obesity in SA, but these elasticities cannot be used in the way Oxford Economics and Econex have done,” she said.

“The elasticities used in our paper are inappropriate for the type of broader economic modelling in the Oxford Economic and Econex papers.”

Tugendhaft added that the article can be read to imply that her study overstates the potential economic impact of a sugar tax.

“ It should read that the way in which Oxford Economic and Econex have relied on the elasticities for their broader economic modelling results in a significant overstatement of the economic impact,” says Tugendhaft.

This week, Stellenbosch-based consultancy Econex released its research for Bevsa on the tax to coincide with a meeting between the industry, its experts and Treasury.

The Econex study backs up an earlier industry sugar tax research study, conducted by global advisory firm Oxford Economics.

It predicts job losses amounting to 30 000 and a loss to GDP of R3bn, or about 0.1%.

However, a critic of these two studies says that the soft drink industry’s consultants are fixating on an “inappropriate” study – co-authored Aviva Tugendhaft, the deputy director of Priceless SA, a unit of the University of the Witwatersrand’s School of Public Health – to discredit the proposed sugar tax.

This study, rather than the actual Treasury proposal, has been the main focus of industry-commissioned attacks.

The result is a “significant overstatement of the economic impact”, said Tugendhaft.

It also means a wilful understatement of potential health benefits, she told City Press.

Like Oxford Economics, Econex based its findings mainly on one local study by Tugendhaft and several  co-authors, including her Wits University research colleague, Mercy Manyema.

The soft drink industry has fingered the so-called Manyema paper as being the major justification for the sugar tax proposal.

The Manyema study mathematically modelled the hypothetical effect on obesity of a flat 20% tax on sugary drinks, and was cited twice by Treasury in its tax proposal.

“Treasury’s proposal makes mention of many other studies, but it appears industry insiders are keen to attack the work in an effort to undermine the policy,” Tugendhaft said.

Both industry studies used the elasticities from the Manyema study as the basis for estimating economic damage resulting from the sugar tax.

In economics, an elasticity is the relationship between a price change and a resulting change in demand.

According to Tugendhaft, the elasticities in their paper, which the industry has emphasised, are inappropriate for economic modelling.

“They are restricted to limited categories of soft drinks, and therefore do not account for broader substitution and shifting of expenditure to other sectors,” she said.

The effect of this is to overstate economic damage, because you are not really measuring what people buy instead of sugary drinks.

Nicola Theron, managing director of Econex, admitted that these criticisms were valid, but added that “within the time frame, this was the best we could do”.

“Our work was mostly aimed at illustrating the fact that much more research needs to be done before implementing the tax,” she told City Press.

“There may very well be smaller negative impacts – depending on a number of refinements, assumptions or scenarios – but the work must be done. Treasury is busy with its own socioeconomic impact study, and it would be good to see what it comes up with.”

But will it cut sugar?

Apart from uncertain economic predictions, the industry research has tried to demonstrate that the tax is unlikely to achieve its actual aim: to reduce sugar consumption.

“The uncertain health impact should be the focus here,” said Theron.

On this matter, the Manyema paper has also become the centre of attention. It predicted a drop in obesity as the result of a hypothetical 20% sugar tax.

Econex redid the modelling exercise and found that the drop in obesity was statistically insignificant.

That means the effect measured was smaller than the margin of error, making the real outcome as likely to be zero as anything else.

The tax that Treasury proposes is, however, not a flat 20% tax at all.

“Our paper assumed a simple 20% tax rate, while Treasury will implement a levy of R0.0229 per gram [of sugar],” said Tugendhaft.

The effect is to put a far higher tax rate than 20% on drinks with higher sugar content, especially cheaper brands in big bottles – which, as City Press previously reported, would face taxes of 50% or more.

According to Tugendhaft, the Treasury proposal will probably have a bigger effect on sugar consumption than a simple flat tax – by, among other things, incentivising a reformulation by cooldrink makers.

“There is obviously a range of things industry can do,” said Theron.

“We did not want to complicate the analysis by incorporating responses such as reformulation.”

Tugendhaft also criticised the willingness by soft drink industry researchers to point out shortcomings in her and her colleagues’ work only in so far as it served the anti-tax campaign.

“There are definite limitations to the Manyema study,” she said, citing as an example that it looked only at obesity and did not include diabetes and other sugar-related conditions.

“Tellingly, these limitations, however, have not been emphasised by the industry consultants,” she added.