Friday, 19 February 2016

Cancer Research jumps the proverbial

I hope Cancer Research's cancer research is better than their economic research. This is their latest press release...


A 20 per cent tax on sugary drinks could reduce obesity rates in the UK by five per cent* by 2025 - equal to 3.7 million fewer obese people - according to a new report** from Cancer Research UK and the UK Health Forum published today (Friday).

The devil is in the detail of the first footnote...

*Current trends predict that obesity rates will be 34 per cent of the UK population in 2025, which will be reduced to 29 per cent by a sugary drinks tax.

So what we have here is a futuristic counterfactual. CR-UK reckon obesity is going to rise to 34 per cent without a soda tax but will 'only' be 29 per cent with a soda tax. There are some pretty basic problems with this.

Firstly, they must mean 'reduce by five percentage points' not 'reduce by five per cent'. This is a minor criticism, but if you're going to attempt a statistical analysis you should have some grasp of maths.

Secondly, if 3.7 million is five per cent of the population then the population has to be 74 million. The population is currently 65 million, so it is going to have to rise by 9 million in the next nine years for the prediction to be right. This is very unlikely.

Thirdly, it is unlikely in the extreme that obesity rates will rise to 34 per cent by 2025. The adult rate of obesity has been fairly static at around 25 per cent for the last ten years and childhood obesity has been falling. Media-driven obesity predictions have all been wildly and hilariously wrong to date and I will put money on this one—which CR-UK was pushing last month—being no different.

Let's not forget that a third of adults were supposed to be obese by 2012 and 36 per cent of men were supposed to be obese by 2015. The latest prevalence figures are 27 per cent for women and 24 per cent for men.

Fourthly, the obesity prediction CR-UK is referring to relates to adults, not the whole population. CR-UK have simply taken 34 per cent of the presumed total population of 74 million and assumed there will be 25 million obese men, women and children. Even if the prediction was right, this would be wrong. Obesity rates are always much lower amongst children.

Finally and most fundamentally, the idea that doubling VAT on sugar-sweetened drinks is going to prevent or cure 3.7 million cases of obesity is borderline insane. Sugary drinks only supply three per cent of the nation's calorie intake and a twenty per cent tax can be expected to reduce consumption of this relatively minor source of calories by no more than ten per cent.

Briggs et al. (2013) modelled a 20 per cent tax and estimated that it would reduce obesity by 180,000 persons—and Briggs et al. are ardent supporters of soda taxes and used some highly optimistic assumptions. Once brand-switching, store-switching and product-switching are taken into account, the effect would likely be negligible to nil.

The study also predicts that the tax could save the NHS about £10 million in healthcare and social care costs in 2025 alone.

That is also unlikely for the reasons above and the reasons I have discussed before, but even if there was a £10 million annual saving, the tax is going to cost £1,000 million, so that's a £990 million loss to the taxpayer.

This is shambolic, back-of-an-envelope stuff from Cancer Research. Littered with basic errors, it looks like it took about three minutes to write. They know as well as the government that a sugary drink tax is not going to be part of the 'childhood' obesity strategy, but by keeping the media blitz going they can raise expectations and pretend to be disappointed when the state starts to take control of the food supply. As I argued at Spectator Health a few days ago, that is the name of the game now.


The BBC has (obviously) been quick to cover this non-story.

Currently, 29% of people are obese and trends suggest that figure will reach 34% in 2025.

Rather than reverse the obesity epidemic, the forecast predicts the tax would lead to obesity rates levelling off at around 29% - preventing 3.7 million people from becoming obese.

No. Nowhere near. 25% of adults are obese, and that figure has not changed in years. An even lower proportion of people are obese. 29% is what this ridiculous model predicts if there isn't a sugar tax.

The guy from industry—the only voice of opposition in the Beeb's article—doesn't help...

The director general of the British Soft Drinks Association, Gavin Partington, said: "The hypothetical claims made in this modelling study run contrary to real-world evidence.

"In fact, the soft drinks tax in Mexico has reduced average calorie intake by six calories per person per day, with no evidence that it has reduced levels of obesity."

He added that other options such as reducing portion sizes or changing ingredients would be more effective.


No comments: