20% sugar tax: Medical Journal paper flawed

18 August 2015

A much-quoted paper published in the NZ Medical Journal by public health academics from the Universities of Auckland and Otago is riddled with flaws, an independent review has found.

The FGC-commissioned review, by economist Dr Brent Wheeler, was entitled ‘Sweet and Sour: Advocacy for Sugar Taxes Rests on Fallacious Reasoning and Ignores Experience’.

The sugar tax lobby regularly quotes the NZMJ paper entitled ‘Twenty percent tax on fizzy drinks could save lives and generate millions in revenue for health programmes in New Zealand’, which was published last year.

Read the NZ Medical Journal paper here

When it was released it received an uncritical reception. But some major flaws in it were glossed over, and unfortunately the conclusions of the “research” were accepted as fact.

Some of those flaws were:

No distinction between sugar and non-sugar fizzy

The research covered all fizzy drinks regardless of whether they were sugar-sweetened, low calorie or completely without sugar. As Professor Tony Blakely told Radio New Zealand at the time, “the research we did was able to look at all fizzy drinks. Unfortunately we couldn’t divide them into the sugar-lodened [sic] ones and the diet ones…”.

Listen to Tony Blakely’s interview here

That seems to be a fundamental flaw if you are trying to work out how much a sugar tax will impact on the consumption of sugar-sweetened drinks. Calculating a tax and levying it over all fizzy drinks regardless of their energy profile seems a bit silly when full-sugar and zero-sugar fizzy drinks have little in common except the bubbles from carbonation.

The estimated $40 million a year from a 20% tax, which the media focused on, includes taxing all carbonated beverages including diet and zero-sugar drinks. The authors never explain the bizarre logic of including non-sugar drinks in such a tax.

There also seems to be a naïve assumption made that there are no costs of collecting or administering the tax, which would certainly eat into the $40 million.

Claim: A 20% tax “would … avert or postpone 67 deaths from cardiovascular disease, diabetes and diet-related cancers”.

This figure is no more than a guess. It’s virtually plucked out of the air.

New Zealand has a population of 4.5 million. As an intervention (assuming it works) this tax of $40 million across all New Zealanders then costs $597,014 per extended life. It’s an extremely indirect and overall costly way of reaching the 67 to change their lives.

The authors admit being uncertain about the number which had an uncertainty interval of 60 – 73 deaths (could be none!), frankly there’s nothing more than theory and guesswork to indicate that a tax levied at such a low level would have any impact at all, particularly when the estimated reduction in consumption from the tax is 20kj a day – a few sips.

Ongoing quotation of this paper

Flawed papers last a long time as they live on as citations in ones that follow, with each conclusion accepted as truth. This is a problem not just for this paper, but others.

FGC had concerns about the credibility of this paper, but wanted independent on it. It occurred to us that these were health people talking about economics issues, so how correct could they be?

At FGC we are not expert in economics, just as we’re not experts in nutrition – that's why we seek advice and background from people who are experts before we comment. 

So we asked respected economist and consultant Dr Brent Wheeler to review it, and his analysis makes for very interesting reading. It provides an excellent example of the pitfalls of nutrition academics (without a doubt expert in their own area of academia) getting involved in the complexities of economics and tax policy.

Dr Wheeler’s report concludes: “The views … oversimplify several issues, disregard a number of influential factors which negate efficacy of their preferred intervention, are selective in respect of their citing from the heavily ambiguous research in this field, have the potential to create other hazards, run the risk of generating a variety of unintended consequences, and overstate the capacity of taxation to improve health and related welfare.

Read his analysis here

Interestingly enough, in July, Tony Blakely, along with colleagues Professor Cliona Ni Mhurchu, and Associate Professor Nick Wilson, sought to rebut on an Otago University blog what they label “opposition arguments” to food taxes. In doing so they were forced to make a couple of concessions:

  • “… the probabilities of overall health benefit from widespread sugar and fat taxes are not fully certain due to potentially complex substitution effects that require more research.”
  • “Food taxes are regressive and penalise individuals and families on low incomes. There is some truth in this argument from a narrow perspective – especially for people who don’t buy less of the taxed product.”

You can read their blog here.