Spectator report on the effect of the sugar tax needs more context

Yesterday, newspapers reported on research showing that the so-called sugar tax on soft drinks had led to a marked fall in the sugar content of these drinks.

A response written in the Spectator said the policy shouldn’t be seen as a success and that it hadn’t affected the growing trend in obesity. It also claimed that: 

“[I]f we’re looking for successes, the sugar tax can be credited with one achievement.

“Despite recent efforts to spam healthy-living guidance and awareness campaigns across the UK, [Public Health England’s] analysis actually shows a small uptick in sugar consumption in certain product categories: “a 0.5 per cent increase in sugar purchased per person” with people showing a sweet-tooth preference for ice cream and chocolate in particular.”

This needs context.

Firstly it’s worth noting that the sugar tax doesn’t actually apply to these food products. It applies to soft drinks. 

The article doesn’t mention the finding that the sugar content of drinks purchased fell by 21.6% in Great Britain between 2015 and 2018 (or 23.1% per person, given the population increase over that period).

The article could be referring to a “substitution effect”. In this case that would mean that the reduction in soft-drink sugar has meant people getting their fix from other foods.

Neither we, nor the Spectator, can say this for sure because the products analysed by Public Health England (PHE) only covered a small range of foods and drinks. If you wanted to measure the substitution effect of the sugar tax you’d need to do a different piece of analysis. 

Nevertheless the fall in the amount of sugar consumed through soft drinks is bigger than the slight rise in sugar consumption through these foods. Across all these product categories put together, sugar consumption decreased by 3.3% per person in Great Britain between 2015 and 2018.

What is the sugar tax?

The sugar tax is a tax on soft drinks. The tax rate is 18p per litre for soft drinks with 5-8g of sugar per 100ml and 24p per litre for soft drinks with 8g or more sugar per litre.

Pure fruit juices and milk-based drinks are exempt.

The tax was announced in 2016 and introduced in 2018, by which point, many companies had already changed the formulation of their products. For example in 2017, PepsiCo reduced the sugar content of 7up from 11g per 100ml to 7g per 100ml.

What does the evidence show?

In September 2019 PHE published a report on progress towards sugar reduction between 2015 and 2018.

This covered both the changes to the sugar content of soft drinks, and other products in food categories which contribute most to the sugar intakes of children.

PHE found that there had been a 2.9% reduction in the average sugar content in these food categories. But this had been outweighed by an increase in the amount of the goods people bought, so the amount of sugar per person consumed in these foods had increased by 0.5% over the period.

For drinks affected by the sugar tax, PHE found the average sugar content fell by 28.8%, and the amount of sugar in soft drinks sold, per person, had fallen by 23%.

It’s worth noting the limitations with this data.

Firstly, these figures don’t cover the “out of home” sector (such as food and drink consumed in restaurants). PHE did publish some data on this sector, which shows falls in the average sugar content of these products, but no information about the actual amount bought.

Also the “food” sugar levels reported only covered certain categories (including biscuits and ice cream but not, for example, fruit), so we can’t say whether the total sugar consumption of the UK has decreased from this analysis, as it doesn’t cover all foods.

Finally, there are two sets of very similar research on this topic. On the one hand there is the PHE evaluation, the data which is used in this piece. Then there is a recent research paper published in the BMC Medicine journal yesterday, which prompted the most recent news coverage.

The researchers essentially emulated what PHE did with regards to soft drinks, looking at the composition and consumption of soft drinks from 2015 to 2018, but looked at a broader category of soft drinks to cover some not directly affected by the tax (juice and bottled water specifically).

They also looked at the amount of sugar consumed in soft drinks purchased out-of-home (for example pubs and restaurants). 

Due to these differences, the researchers had much higher soft drink sales figures than PHE (due to including more drink types and out-of-home) but found a similar percentage fall in sugar consumption between 2015 and 2018.

Because PHE only looked at changes in sugar content and consumption for drinks covered by the sugar tax, we have used its data throughout this piece and not data from the recently published research paper.