Friday 30 October 2020

The "whole systems approach" to obesity is anti-scientific garbage


Dolly Theis has written a three part article about what the government should do about obesity, as if it's any of their business. Dolly is a well meaning Conservative who has fallen in with a bad crowd as a result of doing a PhD in 'public health'. I was on a Spectator panel with her recently discussing the nanny state, along with Steve Brine and Joanna Williams (you can watch it here). 
Theis's thesis is that successive governments have failed to reduce obesity because of poor implementation and evaluation. She makes the improbable claim that 700 anti-obesity policies have been proposed in Britain in the last thirty years.

[The government] currently proposes obesity policies in a way that does not readily lead to implementation, which is likely to be why it does not then implement its own policies. Could you imagine the same happening in business? No, you couldn’t imagine that because it just wouldn’t happen.

There's a simple explanation for that. Businesses respond to what people want whereas the government responds to fanatical single-issue pressure groups and half-witted academics. The policies they propose are, by and large, unworkable, ineffective and go against what people want. Sometimes they encounter equally clueless politicians, such as the aforementioned Steve Brine, who embrace these bad policies because it makes them feel important. That's where the problems begin.
For example, banning 'junk food' advertising and forcing calorie counts in the out-of-home sector sound like great ideas to illiberal politicians until it is explained to them that there is no legal definition of 'junk food' and that many restaurants change their menus on a daily basis. 
When the costs and unintended consequences of such policies are laid out, politicians have no choice but to do a U-turn or water down them down. They should really think through the implications before they announce them, but they naively think that the likes of Action on Sugar are experts on policy. 

In some cases, policies are reproposed in a laughably short amount of time. For example, Chapter 2 of Childhood obesity: A plan for action was published in 2018 under Theresa May. It contained a number of policies, including a 9pm watershed on unhealthy food and drink advertising, and committed to legislating mandatory calorie labelling in the out of home sector.

Consultations were conducted. Individuals and organisations submitted their evidence, reflections and advice. Then poof! Two years later, instead of having implemented the policies, the Government, now under Boris Johnson, publishes another obesity strategy containing those exact same policies and another consultation process.

They say madness is doing the same thing over and over again and expecting different results. Well, hello…

The Childhood Obesity Strategy is a perfect case in point. David Cameron capitulated to the nanny state imbeciles on a range of tobacco-style regulations for food, but then resigned. Under Theresa May, Fiona Hill got rid of the worst of them, but she then lost her job and May reinstated them in a futile attempt to appease the 'public health' industry. May was then replaced by Johnson who rightly saw the nanny state as unconservative and kicked them into the long grass, but after 'long Covid' messed with his mind, he brought them back. As a result, Whitehall bureaucrats are once again saddled with their absurdities and are trying to limit the damage these daft policies will do. 

Yes, it's messy, but it's politics and the last few years have been unusually turbulent. With Brexit and COVID-19, the government would be quite entitled to bin the whole lot on the grounds that it has better things to do than ban supermarkets from placing sausages at the end of aisles.

Three cheers from anyone not keen on government regulation and legislation. But, hold on! Don’t get too carried away – because government proposes policies in such a way that does not readily lead to high compliance. The result is that sectors don’t do enough or don’t do anything at all, so government is pushed more and more into the regulation and legislation corner.
I assume this is principally a reference to the 'voluntary' food reformulation scheme, one of the most ridiculous projects any government has instigated in my lifetime. I put the speech marks around the word 'voluntary' because, as Dolly makes clear, it came with the threat of regulation if the companies didn't comply. 
The reformulation scheme is the perfect example of a hare-brained idea from gormless 'public health' activists. The government committed to it without giving any serious thought to how companies could take arbitrary quantities of fat, sugar, salt and calories out of food or what they could be replaced with. When it was explained to Public Health England that you simply cannot take sugar out of confectionery, for example, the goalposts were quietly moved to allow companies to shrink their products instead. That's not really 'reformulation'. As Josie Appleton has shown, PHE had no idea about such basic facts as jam having to have a contain amount of sugar to be legally sold as jam. 
The hilarious failure of the sugar reduction scheme, in particular, shows what happens when 'public health' ideologues are given free rein. The idea that it would work if the government explicitly introduced coercion is absurd. You can force companies to produce tasteless rubbish, but you can't make people buy it. You might as well pass a law telling water to run uphill.

[The government] should escalate to suggesting those actions to the responsible actor(s)/sector(s). Governments can name and shame, depending on progress, and state how they will move to more deterrence measures (e.g. taxation, laws, etc) if not enough progress is made.

As a last resort, an actor(s)/sector(s) could be fully incapacitated where action/inaction is deemed harmful.

Behold the progressive new Conservative Party! 

The problem of evaluation is addressed in achieving compliance, but I will make the point again here just in case. Policies should always be evaluated, ideally by an independent body.

I agree. Presumably, therefore, Dolly strongly objects to the sugar levy being evaluated by a bunch of academics who are staunch supporters of the tax, including several whose professional reputations depend on it being seen to be a success and one who believes that God literally told him to bring about a sugar tax in England. She must also object to the smoking ban being evaluated by an anti-smoking crusader (it had no impact on pubs, she reckons) and the salt reduction scheme being evaluated by the chairman of Action on Salt

This is quite obviously a racket. Could it be that if these policies were independently audited, they would be shown to be a waste of time and money?

Government must stop this. How is it supposed to know whether something worked if it is not evaluated properly? We also do not always have high-quality evidence about certain interventions and, in some cases, can only build this by introducing the intervention first.

I don't know about that. We don't licence medicines or perform surgery without evidence of efficacy. We have randomised controlled trials showing that food reformulation doesn't work, plenty of evidence that plain packaging doesn't work and masses of evidence that banning fast food shops near homes and schools doesn't work. The 'public health' industry still pushes ahead with such policies because it uses evidence like a drunk uses a lamppost. It is driven by hunches and dogma, not science. 

For example, to really know what the impact will be of a taxation policy such as the sugar tax, government must first introduce it, and then monitor the various impacts closely over time in order to build high-quality evidence.

Or you could use real world evidence from places that already have sugar taxes, but if you did that you'd have to concede that they don't work either.  

Government must therefore be bold in introducing interventions that have the potential to make it easier for us to live a healthier life, and then build the evidence through high quality evaluations.

The strategy of throwing any old policy into the mix in the hope that some of them will work is known euphemistically in 'public health' as the 'whole systems approach'. It is often illustrated with meaningless graphics and is anti-scientific, illiberal nonsense. Essentially, it gives activists a licence to do whatever they want regardless of the consequences. 
If anti-obesity policies are necessary at all - a question Dolly never really addresses - they should follow the same best practice we expect in other policy areas. Gather the evidence and evaluate the costs and benefits, including - crucially - the costs to wellbeing. The last thing we need is a million 'public health' monkeys bashing away on a million typewriters in the hope of finally coming up with something that works.

Wednesday 28 October 2020

Vulture prohibitionists

The emergence of COVID-19 towards the start of the year was always going to go one of two ways for the 'public health' establishment. 
One possibility was that westerners would see what a genuine public health problem looked like and ask why they were handing over billions of pounds to people who knew little or nothing about infectious disease and who were more interested in political virtue-signalling and trivial lifestyle regulation.
The other possibility was that the lifestyle regulators of 'public health' would watch in awe as governments stripped people of their civil liberties and conclude that they could be more draconian than they thought.
As you would expect, the 'public health' racket has been busy working towards the second of these outcomes. After the initial shock of the pandemic subsided, they started talking about how obesity is the real pandemic and smoking kills more people than COVID-19, etc. Richard Horton, the Marxist loon who edits the Lancet, claimed that COVID-19 wasn't a pandemic at all, rather it was a 'syndemic' and that the real problem was the 'non-communicable diseases' that can lead to underlying health conditions. Conveniently, this meant that Horton et al. could get back to controlling people's lifestyles rather than tackling the virus.
The 'public health' establishment has spent decades trying to conflate the risks from self-regarding actions with the risks from infectious diseases. JK Rowling neatly mocked this stupidity in a single sentence earlier this year.

The shameless opportunism of 'public health' wowsers is on full display in this week's Lancet where John Ionnidis and Prabhat Jha suggest exploiting COVID-19 to ban smoking. 
Does the COVID-19 pandemic provide an opportunity to eliminate the tobacco industry?

Ionnidis has made himself unpopular in the scientific community this year by making some fairly wild claims about COVID-19's infection fatality rate (which he thinks is similar to that of bad flu - it isn't). Perhaps he thinks that a bit of prohibitionist tub-thumping will help him get back in their good books.

Tobacco use is the top modifiable global health problem,but the global tobacco market grows 3% annually. Most anti-tobacco measures to date target demand (eg, higher excise taxes). However, the endgame might require reducing supply.
The 'endgame' should require informed adults choosing for themselves whether to smoke or not. That is the only outcome that is acceptable in a free society.
The main counterarguments are financial (eg, economic damage or lost jobs) and defences of personal choice. 
The rest of the article attempts to address the financial arguments but, tellingly, it does not return to the issue of personal choice. 

Most importantly, public health has little experience in enforcing major changes that disrupt markets. 
What?! Disrupting markets is all they know. They explicitly focus on restricting advertising, availability and affordability - three of the main levers of competition. And they have plenty of experience of prohibition, as vapers and drinkers in many countries can tell you.

The ongoing societal response to COVID-19 offers a precedent for drastic action taken to eliminate the tobacco industry.
COVID-19 is a natural experiment: expedient public health considerations have led to decisions being made that have important socioeconomic repercussions. The cumulative disease burden of COVID-19 is large but uncertain. However, if COVID-19 actions were deemed defensible, the risk–benefit ratio for actions to eliminate tobacco is far more favourable.
Smoking is not an infectious disease and 'drastic action' is not required. Individuals can judge the 'risk-benefit ratio' of smoking for themselves.  
Ionnidis and Jha's argument boils down to 'in for a penny, in for a pound'. They are arguing that lockdowns have caused such enormous economic devastation that the problems of tobacco prohibition will seem trivial by comparison. It is the same warped thinking that makes some people think that Britain has already ruined itself with its Covid response so it might as well have a No Deal Brexit for good measure.
Even if all 100 million tobacco-related jobs were lost, this number is still much lower than the number of jobs lost by lockdown measures for COVID-19 worldwide
What an incredibly stupid thing to say.
Until now, only Bhutan has tried banning cigarettes, with mixed effects
That's putting it very delicately. When researchers studies Bhutan's prohibition in 2011, they found...

'... a thriving black market and significant and increasing tobacco smuggling… 23.7% of students had used any tobacco products (not limited to cigarettes) in the last 30 days… tobacco use for adults has not ended or is even close to ending… cigarette prohibition is instrumental in encouraging smuggling and black markets… The results of this study provide an important lesson learned for health practitioners and advocates considering or advocating, albeit gradual, but total cigarette ban as a public policy.'
Bhutan lifted the ban on tobacco in August this year because it was worried that the numerous tobacco smugglers pouring in from other countries would spread the coronavirus. Ionnidis and Jha neglect to mention any of this.
Concerns about smuggling would naturally arise. However, large-scale smuggling can be effectively countered.
Alas, they don't say how this miracle will come about.
They conclude that...

... now that major decisions and actions for health are acceptable under exigency, an unique [sic] opportunity exists to eliminate the tobacco industry.
It's the old fallacy of thinking that what's necessary in times of war will be good in times of peace. The great danger of the COVID-19 crisis is that governments have learned that fearful people will put up with more infringements on their liberty than was thought possible. For prohibitionist vultures, this is an opportunity, as Ionnidis and Jha openly admit. As I wrote back in May, liberal-minded people will have to fight hard to get all our freedoms back when this is all over. While 2020 has been hell for most people, it has been a glimpse of heaven for others.

Monday 26 October 2020

Smoking during a pandemic

ASH are still pushing the idea that a million smokers quit in Britain in the early months of the pandemic. ASH's claims are often unreliable and this one is no different, and the real picture remains hazy. 

We know that tobacco duty revenues rose sharply during lockdown. Even in September, they were still 27% higher than they were in the same period last year. That could be entirely explained by people having less access to tobacco from the black and grey markets, but it doesn't immediately suggest that the number of smokers has rapidly declined.

I'm getting mixed messages. According to Campaign, this year's Stoptober focused on younger smokers because their ranks have been growing during the pandemic...

Some £1m is being spent on this year’s campaign, which is targeting smokers aged 35-60, as well as those in the 18-34 age group, because there has been a rise in smoking among this demographic during the pandemic.

But according to some researchers at Glasgow University, younger people were more likely to quit during lockdown:

Researchers found cigarette smoking also decreased during lockdown, from 15.1 per cent in 2017-19 to 12.1 per cent in April this year, with the decrease most apparent in the younger age groups and among men.

So what's really been happening? The Smoking Toolkit Study is updated every month and has produced some striking results. It shows an increase in the number of people who tried to stop smoking in the last year...

It also shows an increase in the number of people who succeeded in quitting smoking...

Sure enough, it seems that an unusually large proportion of smokers (8.3%) succeeded in quitting in the last twelve months...

The spike is obvious and visible. We don't know at what point in the last twelve months these people quit smoking, but there were 5.7 million smokers in England last year. Even if 8.3% of them had given up in spring as a direct result of COVID-19, it would be 473,000 people, not a million. And that is clearly not a plausible scenario.

The curious thing is that the increased quit rate doesn't seem to have had any real effect on overall smoking prevalence. Although there has been a small step change, it seems to have happened in the second half of 2019, not the first half of 2020.

Nor does there seem to have been any increase in smokers' motivation to quit since the start of the year.

Finally, there has been a notable upswing in the number of young people smoking:

Some of these data are so spiky that I wonder whether the lockdown messed up the methodology somewhat (the researchers were unable to collect any data in March). But, on the face of it, there has been an uptick in the quit rate in the last twelve months, albeit with far fewer than a million people involved, combined with an increase in the number of people who started smoking, especially under the age of 25. In sum, there has been no real change in the smoking rate so far this year.

Friday 23 October 2020

Garbage in, garbage out - food advertising edition

After the year we've had, I suspect a lot of people have had a bellyful of models from 'public health' which reflect nothing more than the assumptions of their authors (see yesterday's post, for example). But let's go once more in the breach.

Ban on junk food ads before the 9pm watershed could stop 160,000 children becoming overweight or obese, study claims

The UK government's ban on junk food ads before the 9pm watershed will stop 160,000 children becoming overweight or obese, according to experts. 

Children would avoid eating about nine calories a day on average, which would reduce childhood obesity figures by nearly 5 per cent, they claim.

Who is claiming this? Step forward Oliver Mytton, who modelled the sugar tax, Emma Boyland, a long time campaigner for food advertising bans, and various colleagues. Not least among them is Russell Viner who has also been campaigning for an ad ban for years and who, in 2018, got £5 million of taxpayers money for an 'obesity policy research unit'

The new model is based on results from a study by none other than Russell Viner which claimed that watching 'junk food' advertising for 4.4 minutes makes kids eat 60 more calories than they otherwise would. The government's impact assessment for the ad ban is also based on this factoid.

Is it a reliable figure? Reader, it is not. It was arrived at by mixing together a bunch of sketchy and contradictory studies from unrealistic experiments in which (mostly young) children were given unlimited free food and no parental control, as I explained last year. 

If one looks at the studies themselves, rather than rely on Viner et al's interpretation, it becomes clear that the 60 calorie claim is based on some very mixed evidence, dominated by a series of studies from one research group whose members have strong feelings about the subject

The whole thing is a fantasy. As with so much in the modern 'public health' racket, it is a model within a model. The sunshine of reality is never allowed to creep through the curtains.

Viner and colleagues were not in any way deterred from their conclusion by the observable fact that children's 'exposure' to 'junk food' on TV has fallen by 70 per cent in the last 15 years without any apparent impact on calorie consumption or obesity. 

Nor do Viner and his new colleagues let real world evidence distract them in their new study. They begin by admitting that "little is known about the impact of advertising on childhood obesity and overweight". So much for evidence-based policy. Alas, the reader won't know any more by the end of it, because it's all fantasy modelling. They take the fictitious 60 calorie figure and extrapolate that kids will consume 9.1 fewer calories a day if they see one and a half fewer 'junk food' adverts per day under a watershed ban.

Further extrapolations lead them to believe that consuming 9.1 fewer calories a day will lead to childhood obesity falling by 4.6% which sounds implausible but I can't be bothered to look at their methodology because, frankly, what's the point? Interestingly, they use the international definition of childhood obesity rather than the totally unscientific UK definition, meaning that a 4.6% decline in child obesity will mean the rate falling from 8.8% to 8.4%. Add in the number of kids who are 'overweight' (none of whom are fat, in reality) and you get the 160,000 children mentioned in the headline above. 
As if that were not enough guesswork, they then make wild extrapolations about the intangible monetary benefits that will be accrued over the course of these children's entire lifetime based on the assumption that none of them will ever become obese (£7.4 billion apparently). 

They admit that these 'benefits' will be reduced by as much as two-thirds if the ads are shown after 9pm instead (which they will), but that only gives them an opportunity to lobby for the inevitable next step of a round-the-clock ban which seems to be the authors' intention.
Or rather it is their intention now that the government has already capitulated on the 9pm ban. The journal received their manuscript last December when the ban was far from a done deal. Like many models in 'public health', it seems designed to act as a political spur rather than provide scientific illumination. They say that "the modelled scenarios are best understood as explorations of what the policy could achieve rather than predictions of what will happen". Where have I heard that before?   
Nevertheless, they conclude by asserting that...

Our study demonstrates that less-healthy food advertising on television in the UK is making a meaningful contribution to childhood overweight and obesity

No, it really doesn't.

Thursday 22 October 2020

The bottomless incompetence of Public Health England

The UK currently requires travellers entering the UK to go into quarantine for 14 days, except for those coming from countries in the 'travel corridor'. 

Some other countries simply test people for SARS-Cov-2 when they land, so why don't we?

The answer is that Public Health England - for it is them again - reckon that testing upon arrival would only identify seven per cent of cases. 

This is based on a PHE model which has a rather substantial flaw, as a report out today shows.

This 7% assumes that all infected travellers who are symptomatic or detectable with a test on departure do not board flights to the UK and therefore only travellers who become detectable during the course of their flight are included in the 7%.

You read that right. PHE simply assumes that a traveller who is symptomatic or detectable ‘prior to boarding their flight [...] does not make it onto their flight’ because of ‘exit screening or the traveller being too ill to fly’. This shows a faith in the integrity and intelligence of the public that experience suggests is unwarranted. It also assumes perfect testing prior on departure.
While the PHE paper assumes that all individuals intending to fly are infected (see section 4.5 below), only a proportion actually board the plane and travel to the UK. This is because the PHE paper makes the assumption that 100% of the symptomatic as well as the asymptomatic passengers who are detectable at the time of boarding do not fly. As these potential travellers are identified before exiting their country of origin, they are not counted towards the percentage of passengers detected (the 7% figure). However, this part of the infected population would also be detected if they flew to the UK and therefore need to be accounted for in considering the effectiveness of a testing regime. Hence, the paper bases its results on a false premise by assuming, but not accounting for, departure testing

This is not a complex technical issue. It is a basic failure of logic and common sense. The assumptions in the model make it impossible for testing on arrival to appear effective.
Therefore [in the PHE model], only passengers who become symptomatic/detectable during the flight can be detected on arrival. Clearly, this is a very small proportion of all passengers. Detection rates on arrival are therefore low based on the way the model is constructed.
When the data were reanalysed, the number of detectable cases rose from 7% to between 33% and 63%. Not perfect, but since only about a fifth of people who asked to quarantine for 14 days actually do so, it is a very significant improvement. 
There are other problems with the model which can you read about in the report if you're interested, but the take-home point is that...
The PHE paper is based on a theoretical model and is not calibrated to real-world data.

A fitting epitaph for this useless organisation.

The Great Covid Divide

I've written for the Sun about people who have no skin in the game itching for a second lockdown.

Being furloughed during the sunniest spring on record was more agreeable to those who had a large garden than to those who were stuck in a tower block with an abusive partner.

The lockdown greatly exacerbated equalities in housing, employment and personal circumstances.

White collar managers on Zoom meetings were waited on by delivery drivers, shop assistants and other blue collar workers for whom lockdown never happened.

People working in precarious, low-paid jobs bore the brunt, socially and economically, particularly when the schools were closed.



Wednesday 21 October 2020

SCHEER report on vaping pretends it's still 2011

Last month, the EU’s Scientific Committee on Health, Environmental and Emerging Risks (SCHEER) published a preliminary opinion on electronic cigarettes which assesses the risks associated with vaping products and is designed to alert the European Commission to the “potential need for legislative amendments” to the Tobacco Products Directive. Reading it is like taking a step back in time, or being in Australia.

The authors of the SCHEER report appear to be biased against e-cigarettes and harm reduction. The report reheats several arguments, such as the ‘gateway effect’ and the ‘renormalisation’ hypothesis, which are now a decade old and have been contradicted by real world evidence. While it downplays strong evidence showing that e-cigarettes have been a gateway from smoking for millions of people, it amplifies speculation about hypothetical risks. When the authors are unable to find adequate evidence for anti-vaping claims, they quote from organisations which share the same prejudice. Much of the evidence is treated selectively and some of the conclusions made about the strength of evidence are baffling.   

I've written a response to the consultation - you can download a PDF of it here. I made the following points...

Health risks

Harm reduction is a well-established concept in public health. Reduced-risk substitutes for hazardous products are not expected to be totally safe (nothing is totally safe). They are only expected to be significantly safer. The UK’s Royal College of Physicians (2016: 84) has concluded that the long-term health risks from vaping are “unlikely to exceed 5% of those associated with smoked tobacco products, and may well be substantially lower than this figure”. The evidence in the report is consistent with this.

The report would be more useful if it compared the risks of vaping with the risks of smoking, since the vast majority of e-cigarette users are current or former smokers. Instead, it speculates about possible risks from substances in e-cigarette vapour, particularly nicotine. Nicotine is known to have a modest effect on the cardiovascular system similar to caffeine and certain foods, but there is no evidence that this poses a significant health threat to vapers, nor to users of snus or nicotine replacement therapy (NRT). According to a review conducted by Stephens (2018), lifetime cancer risk associated with vaping is less than one per cent of that associated with smoking (this study is mentioned in the report, but that important finding is not.)

The report notes that the most common respiratory problem associated with vaping is mouth and throat irritation which dissipates over time. It classifies the evidence for this as “strong”. Similarly, it classifies the evidence of systemic effects on the cardiovascular system as “strong”. It is important to remember that the strength of evidence about specific risks does not necessarily imply that the risks are substantial. In this instance, the health risks are quite trivial, especially when compared to smoking.

The report notes rare instances of e-cigarette devices exploding or catching fire. This is an inherent danger with any electrical device and is dealt with through normal product standards regulation. All the incidents of product malfunction in the report involved products that did not comply with the Low Voltage Directive. The report also notes some rare instances of non-fatal poisonings due to people ingesting e-cigarette fluid. It correctly observes that none of the fluids consumed in incidents reported to Safety Gate were compliant with the Tobacco Products Directive (TPD).

Secondhand exposure

The report can find very little evidence that secondhand exposure to e-cigarette vapour may be a health threat. It notes that “despite high levels of carbonyl emissions as reported in several studies above, limited impacts on cardiovascular and/or other health outcomes have been documented”. It adds: “Surprisingly, particularly in relation to cardiovascular and other health effects of passive smoking secondary to electronic cigarettes, the authors found that the complete blood counts of otherwise naïve passive smokers are not affected by such exposures”. Given the low risks associated with active e-cigarette use, it is not clear why the authors find this “surprising”. The doses involved are orders of magnitude smaller, there is no equivalent of sidestream smoke (the smoke that emanates from the lit end of the cigarette), and the nicotine in e-cigarettes is mostly absorbed by the user.

The ability to detect the presence of substances at trace levels is not evidence of harm. Of the four relevant studies cited on pages 51 and 52, one is no more than a description of research yet to be conducted (Shearston et al.), one found no impact of passive (or active) vaping on lung function (Flouris et al.), another looked at carbonyl emissions as described above (Farsilinos and Gillman) and the last found trace levels of nicotine in e-cigarette users’ homes. None of this implies harm to bystanders (Ballbe et al.)

Given the hypothetical nature of the threat and the committee’s acknowledgement that “data on the long-term consequences of passive smoking of electronic cigarettes on human health are lacking”, it is remarkable that the report assesses the evidence of risks to “second-hand exposed persons” as “weak to moderate”. It would be more accurate to describe it as negligible to non-existent. 


The report expresses concern about e-cigarette flavours possibly encouraging initiation by non-smokers and/or minors. Unflavoured e-cigarette fluid is rarely consumed by vapers. ‘Tobacco’ flavour is often classed as unflavoured by anti-vaping activists (and I will follow that convention below), but there is no logical reason for this. ‘Tobacco’ flavour only vaguely resembles the taste of smoked tobacco and is an artificial flavour like any other. Some vapers like it, others do not. To encourage smokers to switch to vaping, it is important to have a wide range of flavours available.

When e-cigarettes first came on the market, they were only available in tobacco flavour. The higher prevalence of tobacco flavour use among older vapers is likely to reflect, in part, the earlier initiation of the first cohort of vapers. Goldensen et al. (2019) cite evidence from the US PATH survey showing that younger people are more likely to use flavoured juice than older people. However, it is important to recognise the majority of vapers of all ages use flavoured (i.e non-tobacco flavour) e-cigarettes.

The report does not mention another finding in Goldensen et al. (2019: 7):

‘Data from observational and qualitative studies suggest that flavored e- cigarettes may aid adult smokers in smoking reduction and cessation efforts. Former smokers cite the wide variety of available flavorings and superior taste of e-cigarettes as factors that aid smoking cessation, and note that restricting the availability of flavorings would make the vaping less enjoyable and reduce the appeal of e-cigarettes.’
A recent study by Yang et al. (2020) found that a ban on e-cigarette flavours in San Francisco led to increased smoking prevalence among 18-24-year-olds. Cigarette sales in the USA have risen in 2020 as a result of federal restrictions on flavours in certain e-cigarette products (Maloney 2020).

The report refers to the Special Eurobarometer 458 survey, but does not mention that it found that “flavours of e-cigarettes were also relatively unimportant as a reason for starting to use them”. This reason was cited by only 12 per cent of EU vapers when asked which factors were important in their decision to start using e-cigarettes. The most common response, mentioned by 61 per cent of vapers, was “to stop or reduce your tobacco consumption”. The report correctly notes that the same survey found that 81 per cent of vapers opposed a ban on flavours, with only 9 per cent in favour.

The section on flavours concludes by repeating claims and recommendations from the European Heart Network, a pressure group that has always opposed tobacco harm reduction and wants e-cigarette flavours prohibited. It makes the unsubstantiated claim that young people who experiment with e-cigarettes are “at substantial risk of becoming regular cigarette smokers”. However, the group admits that: “Whether these young people would have started smoking conventional cigarettes had e-cigarettes not existed is a question that we cannot answer” (European Heart Network 2019: 15). This issue is discussed below.

Gateway claims

The ‘gateway effect’ is a term borrowed from War on Drugs rhetoric designed to encourage the prohibition of low-risk substances on the basis that their consumption inevitably leads to the use of high-risk substances. The concept remains highly controversial. Anti-vaping activists have been claiming that vaping acts as a stepping-stone to combustible tobacco consumption for at least a decade. A number of cross-sectional and longitudinal studies, mostly from the USA, have produced evidence that purports to supports this claim.

The report puts significant weight on a meta-analysis by Soneji et al. (2017) which looked at nine studies, one of which was a conference abstract. The lead author (Samir Soneji) was co-author of a third of the studies reviewed. All the studies found that people who had ever used an e-cigarette were more likely to have later smoked a cigarette. Several studies with similar results have since been published. From this, it is inferred that the individuals who began smoking would not have done so had they not first started vaping.

Since this is impossible to prove (or disprove), the report is wrong to claim that “there is strong evidence that electronic cigarettes are a gateway to smoking for young people”. This mistakes quantity of evidence for quality. Regardless of how many studies find an association between ever-vaping and ever-smoking, they are all flawed in the same way. Meta-analyses then suffer from the problem of ‘garbage in, garbage out’.

The crucial missing variable is personality. The kind of person who is more likely to try an e-cigarette is the kind of person who is more likely to try a cigarette. Conversely, the kind of person who has a strong aversion to cigarettes is more likely to abstain from vaping. These personality traits are extremely difficult to control for and no researcher has yet found an adequate way of doing so. In the absence of sound methodology, ‘gateway’ studies only show that adolescents who are not risk averse will try different things and that the kind of people who would have experimented with cigarettes before e-cigarettes were on the market sometimes try vaping first (Phillips 2015).

Switching from e-cigarettes to cigarettes makes little sense in either financial or health terms, but we cannot rule out the possibility that it has ever happened. At the population level, the important question is whether the relative popularity of vaping among some adolescents has led to a surge in combustible tobacco use. A related question is whether the increased use of e-cigarettes has led to what the report calls the “renormalisation of cigarette smoking [that] could lead to a resurgence of cigarettes smoking” (p. 69). These concerns are now a decade old. The Soneji study suggests that non-smoking vapers are 3 or 4 times more likely to become smokers than abstainers. This would be a large effect, if true. If there was a gateway effect of any significance, it should be apparent in the data. Smoking rates should be rising, or at least falling more slowly, in countries where vaping has become popular.

In reality, e-cigarette use is associated with sharp declines in the smoking rate (including the youth smoking rate). After vaping became popular in 2012, England’s smoking rate fell by 20 per cent in just five years, following five years in which the rate had been almost flat. In the same period, the smoking rate among children halved and is now at the lowest rate on record. Of the people in the UK who have used both cigarettes and e-cigarettes in their lives, 91.8 per cent used cigarettes first while only 0.1 per cent used e-cigarettes first (Office for National Statistics 2020).

The report acknowledges that the USA saw a decline in smoking prevalence between 2014-2017 “which coincides with the timeframe of electronic cigarette proliferation in the US” (p. 17). This understates the impact of vaping. Cigarette smoking by American middle school students nearly halved between 2011 and 2019, from 4.3 per cent to 2.3 per cent, and fell by more than half among high school students, from 15.8 per cent to 5.8 per cent. These are historic declines occurring at a time when 4.7 per cent of middle school students and 19.6 per cent of high school students are current (past 30-day) e-cigarette users. Vaping is far more common among US school students than it was in 2011, and whilst this is not a welcome development in itself, it is striking that it has coincided with large declines in the smoking rate. The same is true of American adults who now smoke at the lowest level on record: down from 18.1 per cent in 2012 to 13.7 per cent in 2018.

The US adult smoking rate fell by 4.4 percentage points between 2012 and 2018. England’s adult smoking rate fell by 4.5 percentage points in the same period. By contrast, in Australia, which has a ban on nicotine vaping products and some of the world’s strongest tobacco control policies, the adult smoking rate only fell by 1.8 percentage points between 2013 and 2019.

None of this categorically disproves the existence of a gateway effect (which, as mentioned, would be impossible), but it strongly suggests that if such an effect exists, it is trivial and heavily outweighed by the use of e-cigarettes as a gateway from smoking. Amongst younger people, vaping may well act as a prophylactic against smoking.

The rapid decline of smoking in countries where e-cigarette use is common implies that there has been no ‘renormalisation’ of smoking. There is also empirical evidence from the UK showing that negative attitudes towards smoking among young people have remained strong despite the widespread acceptance of vaping (Brown et al. 2020; Hallingberg et al. 2020).

Smoking cessation

It is now well established that vaping helps people quit smoking. The report does not agree. Why?

Unlike the issues of flavours and the gateway effect, smoking cessation can be studied through randomised control trials (RCTs), the gold standard of scientific evidence. The report acknowledges three RCTs included in the Hartmann-Boyce meta-analysis (Bullen et al. 2013, Carponnetto et al. 2013, Adriaens 2014). All of them showed that smokers were more likely to quit smoking if they used e-cigarettes than if they used a placebo or nicotine replacement therapy. This was particularly impressive since two of the RCTs involved smokers who had no desire to quit at the outset.

The report also mentions more recent RCTs by Hayek et al. (2019) and Walker et al. (2019). The former found smokers to be nearly twice as likely to quit using e-cigarettes than if they used nicotine replacement therapy. The latter found that smokers using nicotine patches plus a nicotine e-cigarette were more likely to quit than those using patches plus a zero-nicotine e-cigarette.  

The evidence from RCTs is consistent with evidence from observational and ecological studies. For example, a study by Zhu et al. (2017), which is not cited in the report, found that the “substantial increase in e-cigarette use among US adult smokers was associated with a statistically significant increase in the smoking cessation rate at the population level.” A study of vape shop customers found that 41 per cent had quit smoking within a year of taking up e-cigarettes (Polosa et al. 2015). A clinical trial using second generation e-cigarettes saw 53 per cent of subjects quit smoking (Pacifici et al. 2015). A recent analysis of 13,057 current and former smokers in 28 EU countries, which is not cited in the report, found that current e-cigarette users were almost five times more likely to have quit smoking in the last two years than non-vapers and more than three times more likely to have quit in the last three to five years (Farsilinos and Barbouni 2020).

Given the large body of evidence showing that e-cigarettes are effective smoking cessation devices, it is strange that the report effectively dismisses the science and relies instead on quotes from the US Surgeon General and the European Heart Network, both of whom have a known prejudice against vaping. The report could easily have quoted eminent academics and health groups who believe that e-cigarettes are effective smoking cessation aids. Either way, it is not obvious what purpose such quotations serve in an independent evidence assessment other than to show whose side the authors are on.  

The report’s conclusion that “there is weak evidence for the support of electronic cigarettes’ effectiveness in helping smokers to quit” is an extraordinary interpretation given that at least five RCTs and numerous observational and ecological studies have provided strong evidence to the contrary. Elsewhere in the report a much weaker set of studies is cited to support the claim that there is “strong evidence” for a gateway effect. Even the almost non-existent evidence on secondhand vaping is described as “weak to moderate”.

This is, at best, inconsistent. When evidence from observational epidemiology shows that some adolescents have as little as one puff on a cigarette having previously tried vaping, it is considered proof of a gateway effect, and yet when evidence from observational epidemiology, national surveys, personal testimonies and RCTs show that large numbers of smokers have switched exclusively to vaping, it is effectively ignored.  

Since the SCHEER report was released, a Cochrane Review has concluded that there is “moderate-certainty evidence that ECs [electronic cigarettes] with nicotine increase quit rates compared to ECs without nicotine and compared to NRT” (Hartmann-Boyce et al. 2020).

Evidence from economics

The report understandably focuses on evidence from the medical and scientific literature, but there are studies from the economics literature that help resolve some of the issues at stake. The observation that e-cigarettes are a substitute for, not a complement to, combustible cigarettes is an important one, not least because it suggests that efforts to suppress e-cigarette use will lead to greater use of traditional cigarettes. This is now well established. Several studies have shown that cigarettes and e-cigarettes are substitute products, both in the USA (Zheng et al. 2017) and in the EU (Stoklosa et al. 2016).

Unsurprisingly, therefore, it has been shown that policies designed to deter e-cigarette use have the unintended consequence of increasing both cigarette consumption and smoking prevalence. Pesko et al. (2020) found that “higher e-cigarette tax rates increase traditional cigarette use” and that an e-cigarette tax of US$1.65 per ml would increase the number of daily smokers by one per cent. Cotti et al. (2020) studied e-cigarette taxes in eight US states and found that a decline in e-cigarette pod sales led to an increase in the sale of traditional cigarettes. Saffer et al. (2019) concluded that a large tax on e-cigarettes in Minnesota prevented 32,400 smokers from quitting. Abouk et al. (2019) found that e-cigarette taxes lead to more women smoking in pregnancy. Friedman (2015) found that banning the sale of e-cigarettes to minors increased the underage smoking rate by 0.9 percentage points.

Economic findings such as these provide indirect evidence that e-cigarettes are used by smokers to quit or cut down their cigarette consumption. And they give us direct evidence of the effects of policies designed to reduce e-cigarette consumption. Interventions that making vaping less attractive to consumers, such as taxes and flavours bans, have been repeatedly shown to increase cigarette consumption and smoking prevalence.


The way in which the report places heavy emphasis on studies which imply negative consequences from vaping while downplaying or ignoring evidence showing benefits suggests a prejudice against tobacco harm reduction from at least some of the authors. There is evidence of bias throughout the document.

For example, the report refuses to use the word ‘vaping’ “because it may imply that the consumption of electronic cigarettes are a ‘healthy' alternative to cigarette smoking and consumers may misperceive risks associated with the use of electronic cigarettes. The SCHEER prefers to use the neutral ‘use (users) of electronic cigarette’” (p. 19). Aside from the fact that ‘vaping’ is a perfectly neutral and accurate term for inhaling e-cigarette vapour, the report elsewhere uses phrases such as “electronic cigarette smoking”, “heavy smokers of electronic cigarettes” and “electronic cigarette smoking behaviour” which are neither neutral nor accurate and wrongly imply that vaping has similar health impacts as smoking.  

On page 20 of the report, it says: “This Opinion is restricted to the terms of references given by the European Commission. It covers electronic cigarette products complying with the TPD”. And yet the report makes frequent references to the US version of Juul which is illegal in the EU (although it refuses to refer to it by name), and its section on poisonings and injuries almost exclusively involves products that do not comply with EU regulations. Since illegal products are beyond the committee’s purview, it is not clear what purpose this serves beyond spreading alarm.

On several occasions, the authors resort to quoting external organisations when they are unable to find scientific evidence to support anti-vaping claims. On page 46, for example, having admitted that the “health impacts of electronic cigarette’s use are still difficult to be established due to the lack of long-term data from epidemiological studies or clinical trials”, the report quotes the World Health Organisation saying, in 2016, that e-cigarettes “are harmful to health and are not safe”. It then gratuitously lists some anti-vaping policies that the WHO would like to implement. Unless the WHO has some evidence that has been withheld from SCHEER, it is not clear what relevance its opinions have.

Similarly, on page 60, the report quotes another WHO document written by anonymous authors which claimed that “long-term use is expected to increase the risk of chronic obstructive pulmonary disease, lung cancer, and possibly cardiovascular disease as well as some other diseases also associated with smoking”. None of this is supported by the evidence in the SCHEER report. The only reference for these claims in the WHO document is Britton et al. (2016) which, in turn, cites Nicotine Without Smoke, a report from the Royal College of Physicians (2016). The relevant section of the RCP report states that long-term e-cigarette use “would be expected, from first principles, to increase the risk of lung cancer, COPD, cardiovascular disease and other diseases caused by smoking, but at much lower levels of risk” (my emphasis). The RCP report further notes that “there is very little evidence that short-term use of e-cigarettes causes any appreciable harm to users or to others” and concludes: “The risks attributable to long-term inhalation of nicotine in isolation from tobacco smoke, and of the propylene glycol, glycerine and other components unique to e-cigarettes, are also uncertain but likely to be low. The health harm to long-term users of e-cigarettes is therefore likely to be marginally greater than for those who use conventional NRT.” The SCHEER report, like the WHO document it quotes, ignores this important context.


E-cigarettes have been on the market for over a decade. They have always split opinion, and they continue to face resistance from some activist-academics, particularly in California and Australia, but hard data from countries in which vaping has become mainstream has allayed the early concerns. By reviving concepts such as ‘renormalisation’ and the ‘gateway effect’, as well as by casting doubt on vaping’s role in smoking cessation, the SCHEER report is a step backwards.   

The authors appear to be biased against e-cigarettes and harm reduction. While the report downplays strong evidence showing that e-cigarettes have been a gateway from smoking for millions of people, it amplifies speculation about hypothetical risks. When the authors are unable to find adequate evidence for anti-vaping claims, they quote from organisations that share the same prejudice against e-cigarettes. Much of the evidence is treated selectively and some of the conclusions made about the strength of evidence are baffling.   

By misclassifying weak evidence as moderate or strong (and vice versa), the report sends a misleading message to the European Commission and to member states. Attention should be paid to evidence from the economics literature, which is not included in the report, but which helps resolve some of the questions faced by policy-makers.


References are only listed below if they are not listed in the SCHEER report.

Abouk, R., Adams, S., Feng, B., MacLean, J. and Pesko, M. (2019) The Effect of E-Cigarette Taxes on Pre-pregnancy and Prenatal Smoking. NBER Working Paper No. 26126.

Britton, J., Arnott, D., McNeill, A. and Hopkinson, N. (2016) Nicotine without smoke—putting electronic cigarettes in context. British Medical Journal 353: i1745.

Brown, R. et al. (2020) A qualitative study of e-cigarette emergence and the potential for renormalisation of smoking in UK youth. International Journal of Drug Policy 75: 102598.

Cotti, C., Courtemanche, C., Maclean, J., Nesson, E., Pesko, M. and Tefft, N. (2020) The Effects of E-Cigarette Taxes on E-Cigarette Prices and Tobacco Product Sales: Evidence from Retail Panel Data. NBER Working Paper No. 26724.

Caponnetta, P., Auditore, R., Russo. C, Cappello, G. and Polosa, R. (2013) Impact of an electronic cigarette on smoking reduction and cessation in schizophrenic smokers: a prospective 12-month pilot study. International Journal of Environmental Research and Public Health 10(2): 446-61.

European Heart Network (2019) European Electronic cigarettes and cardiovascular disease – an update from the European Heart Network.

Farsilinos, K. and Barbouni, A. (2020) Association between electronic cigarette use and smoking cessation in the European Union in 2017: analysis of a representative sample of 13 057 Europeans from 28 countries. Tobacco Control doi: 10.1136/tobaccocontrol-2019-055190.

Friedman, A. (2015) How does electronic cigarette access affect adolescent smoking? Journal of Health Economics 44: 300-308.

Hallingberg, B. et al. (2020) Have e-cigarettes renormalised or displaced youth smoking? Results of a segmented regression analysis of repeated cross sectional survey data in England, Scotland and Wales. Tobacco Control 29: 207-216.

Hartmann-Boyce, J., McRobbie, H., Lindson, N., Bullen, C., Begh, R., Theodoulou, A., Notley, C., Rigotti, N.A., Turner, T., Butler, A.R. and Hajek, P. (2020) Electronic cigarettes for smoking cessation (Review). Cochrane Database of Systematic Reviews 10 (CD010216).

Maloney, J. (2020) Cigarette Smoking Makes Comeback During Coronavirus Pandemic. Wall Street Journal 28 July:

Office for National Statistics (2020) E-cigarette use in Great Britain (Table 6):

Pacifici, R., Pichini, S., Graziano, S., Pellegrini, M., Massaro, G. and Beatrice, F. (2015) Successful Nicotine Intake in Medical Assisted Use of E-Cigarettes: A Pilot Study. International Journal of Environmental Research and Public Health 12(7): 7638-46.

Pesko, M., Courtemanche, C. and MacLean, J. (2020) The effects of traditional cigarette and e-cigarette tax rates on adult tobacco product use. Journal of Risk and Uncertainty 60: 229-258. 

Phillips, C. (2015) Gateway Effects: Why the Cited Evidence Does Not Support Their Existence for Low-Risk Tobacco Products (and What Evidence Would) International Journal of Environmental Research and Public Health 12(5): 5439-5464.

Polosa, R., Caponnetto, P., Cibella, F. and Le-Houezec, J. (2015) Quit and smoking reduction rates in vape shop consumers: a prospective 12-month survey. International Journal of Environmental Research and Public Health 12(4): 3428-38.

Royal College of Physicians (2016) Nicotine without Smoke. London: RCP.  

Saffer, H., Dench, D., Grossman, M., and Dave, D. (2020) E-Cigarettes and Adult Smoking: Evidence from Minnesota. Journal of Risk and Uncertainty. 60: 207-28.

Stoklosa, M., Drope, J. and Chaloupka, F. (2016) Prices and E-Cigarette Demand: Evidence From the European Union. Nicotine & Tobacco Research 18(10): 1973-80.

Yang, Y., Lindblom, E. N., Salloum, R. G., & Ward, K. D. (2020). The impact of a comprehensive tobacco product flavor ban in San Francisco among young adults. Addictive Behaviors Reports, 11: 100273.

Zheng, Y., Zhen, C., Dench, D. and Nonnemake, J. (2017) U.S. Demand for Tobacco Products in a System Framework. Health Economics 26(8): 1067-86.

Zhu, S., Zhuang, Y., Wong, S., Cummins, S. and Tedeschi, G. (2017) E-cigarette use and associated changes in population smoking cessation: evidence from US current population surveys. British Medical Journal 358: j3262.

Tuesday 20 October 2020

The Great Barrington conspiracy

Some people have got it into their heads that the Great Barrington Declaration (calling for a sort of herd immunity approach to Covid) is a plot by shady vested interests. Alas, they can't identify the vested interests and it's difficult to see who could benefit from it.

I've written about it for The Critic. No paywall, read it all.

Friday 16 October 2020

E-cigarettes: the Cochrane Review

A new Cochrane Review of e-cigarettes was published this week. Cochrane Reviews focus on randomised controlled trials, which are generally seen as the gold standard of scientific evidence. You don't see many of them in 'tobacco control', but they are possible when it comes to smoking cessation.

The previous Cochrane Review could only find two high quality RCTs. The new review found four, with nicotine replacement therapy and/or nicotine-free e-cigarettes used by the control group. It concludes:

There was moderate-certainty evidence, again limited by imprecision, that quit rates were higher in people randomized to nicotine EC than to non-nicotine EC (RR 1.71, 95% CI 1.00 to 2.92; I2 = 0%; 3 studies, 802 participants). In absolute terms, this might again lead to an additional four successful quitters per 100 (95% CI 0 to 12).

The element of randomisation almost certainly leads the the real world efficacy of vaping in cessation being downplayed. These studies show that e-cigarettes help people quit smoking even when they don't intend to quit smoking at the outset. For people who actively want to quit smoking, e-cigarettes are even more effective. I'll be writing more about this next week. 

Wednesday 14 October 2020

Lockdown junkies

I've written for the Telegraph about the knee-jerk demands for another lockdown...

I have no doubt that quarantining the entire population is an effective way of reducing the number of infections. Amputating your leg is an effective way of getting rid of a verruca. The question is whether a second national lockdown (sorry, circuit breaker) is a proportionate and necessary response that will do more good than harm. Lockdowns have never been a feature of the NHS’s pandemic response plans, and were never seriously considered until the Wuhan police started welding people’s doors shut. 

Remember when lockdown was a last resort to prevent hospitals being overrun? Protecting the NHS was the only reason given when Boris Johnson appeared on television on March 23. Most of us signed up to that. Since then, we have vastly increased our testing capacity, hired 25,000 contact tracers, banned gatherings of more than six people, built the Nightingale hospitals, discovered effective treatments, made masks mandatory, closed bars and restaurants at 10pm and introduced local lockdowns. Social distancing is now ingrained in our way of life and working from home has been normalised. 

This is what living with an endemic virus means. We always knew there would be a resurgence as winter approached, but the measures we’ve taken are slowing the spread appreciably. On March 20, there were 1,500 people in hospital with Covid-19. Nineteen days later there were nearly 20,000. On September 24, the number of people in hospital with Covid-19 reached 1,500 again. Nineteen days later (yesterday), there were 4,367. In the whole south of England, including London, there were only 681.


Tuesday 13 October 2020

Public Health England's sugar reduction fiasco

When it should have been preparing for outbreaks of infectious disease, Public Health England's big 'world leading' initiative was getting the food industry to reduce the sugar, fat, salt and calorie content of food by arbitrary amounts so the British public would lose weight without even knowing they were on a diet. 
Known as 'health by stealth', the scheme assumes that people blindly consume the same diet every week without paying attention to flavour. It is a textbook example of simplistic thinking by out-of-touch bureaucrats. As Adam Smith said of the 'man of system'...

He seems to imagine that he can arrange the different members of a great society with as much ease as the hand arranges the different pieces upon a chess-board. He does not consider that the pieces upon the chess-board have no other principle of motion besides that which the hand impresses upon them; but that, in the great chess-board of human society, every single piece has a principle of motion of its own, altogether different from that which the legislature might chuse to impress upon it.
PHE buys into the dogma of the economically illiterate 'public health' lobby who think that consumers do whatever industry wants them to. And so it set food companies a target of reducing sugar in most processed products by 2020.
The abolition of PHE was announced in July but the quango will stagger on like a zombie until the spring. Last week, it published a 'progress report' on the sugar programme. There has not been much progress.
The sales weighted averages show that:

overall there was a 3.0% reduction in total sugar per 100g in products sold between baseline (2015) and year 3 (2019)

• there were larger reductions for some specific product categories (yogurts and fromage frais down 12.9%, and breakfast cereals down 13.3% compared with the 2015 baseline)

• there was a reduction of 6.4% for ice creams lollies and sorbets and 5.6% for sweet spreads and sauces compared with 2015

• there were reductions of 4.8% for cakes and 5.6% for morning goods, compared with their baseline of 2017

• there were much smaller reductions for 3 other categories: biscuits (1.6%), chocolate confectionery (0.4%) and sweet confectionery (0.1%)

• there was a small increase in the puddings category

This is obviously a long way from the 20% reduction that was hoped for, but we can see that some categories have seen non-trivial reductions in sugar content.
You can lead a horse to water but you can't make him drink. These reductions don't mean anything if consumers change their behaviour. This is where things get even worse for the PHE bureaucrats.

For retailers and manufacturers, it can be seen that:

• overall there has been an increase from 723,103 tonnes of sugar sold at baseline to 741,966 tonnes in year 3 which represents an increase of 2.6%

• as the population of Great Britain increased during this period the increase in sugar sold represents no change in sugar purchased per person from the food product categories included in the programme

• the largest increases in tonnes of sugar sold were 16.3% for chocolate confectionery, 7.2% for sweet confectionery, 6.1% for sweet spreads and sauces, and 5.7% for biscuits

• the largest decreases were 15.9% for yogurts and fromage frais, 13.9% for breakfast cereals and 4.0% for puddings


Furthermore, if we look at the volume of food sold...

• overall there has been an increase from 2,804,089 tonnes of products sold at baseline to 2,900,197 tonnes in year 3 which represents an increase of 3.4%

• as the population of Great Britain increased during this period, the increase in tonnes of products sold represents a 0.8% increase per person from food product categories included in the programme

• sales decreased in 3 categories: breakfast cereals (down 0.5%), puddings (down 3.0%), and yogurts and fromage frais (down 3.3%)

• there were increases in sales in other categories including chocolate confectionery (up 16.3%), sweet spreads and sauces (up 12.0%), ice cream, lollies and sorbets (up 8.0%), sweet confectionery (up 7.3%) and biscuits (up 6.8%)

Note that the two categories which saw the biggest fall in sugar per 100g (breakfast cereals and yogurts/fromage frais) saw a decline in sales. Coincidence? Probably not.

Meanwhile, in the out of home sector (cafés, restaurants etc.) things look just as bad from PHE's perspective.

For businesses in the eating out of home sector:

• overall there has been hardly any change in the simple average sugar content from 24.6g per 100g in 2017 to 24.5g/100g in 2019

• the largest decreases were 17.1% for breakfast cereals, 6.8% for cakes, and 3.9% for biscuits

• there was an increase for chocolate confectionery of 10.7%

PHE concludes by putting a brave on it...
There has been progress in some, but not all, food categories. Sustained progress in sugar reduction has been seen for breakfast cereals (down 13.3%) and yogurt and fromage frais (down 12.9%). However, as described previously, these reductions are not being fully realised in the programme overall. This is due to a reduction in the proportion of total sales from these lower sugar categories and increases in sales in higher sugar categories such as chocolate confectionery (which is reporting hardly any change in total sugar per 100g).

Overall these changes have resulted in more sugar from these products now appearing in shopping baskets than was the case in 2015.

You have to raise your hat to the British public. Faced with a sugar tax and a food reformulation scheme, they responded by eating more chocolate, ice cream and biscuits. PHE are very proud that soft drinks have less sugar in them, on average, since the sugar tax was introduced, but I suspect that the overall impact on sugar and calorie consumption - and therefore obesity - will be approximately zero.
PHE says it expects the forthcoming advertising ban to rid the nation of its sweet tooth (it won't) and there is a suggestion that the quango will continue lobbying for ineffective nanny state policies until it is finally put out of its misery.

Transparent monitoring of the sugar reduction programme, and further expert advice on the potential levers to address excess sugar consumption, will continue to be provided to government.

I doubt there is a single person in the country who associates Public Health England with 'expert advice' anymore. Good riddance!

The surprisingly controversial benefits of moderate drinking

I've written a long read for Spectator USA about moderate drinking, the J-Curve and alcohol guidelines.

The benefits of moderate drinking are an inconvenience to those who want to send a clear and simple message to the public (‘don’t drink’) as well as to those who want to introduce restrictive, tobacco-style legislation for alcohol. If alcohol can be part of a healthy lifestyle, it is difficult to demonize. Thus, for the new breed of temperance campaigner, it would be better if there were ‘no safe level’ of alcohol consumption.
Everything you need to know in one place. Have a read.

Monday 12 October 2020

Pubs and coronavirus - what does the evidence say?

Chris Whitty, the Chief Medical Officer, was lobbying MPs to close pubs in the North of England with a dodgy dossier that supposedly showed that a third of COVID-19 infections took place in pubs. Tellingly, he hasn't made this evidence public, but screenshots taken of one of his slides suggests that it is tenuous post hoc ergo propter hoc stuff. I wrote the following about the evidence on pubs and the coronavirus on Friday. You can download it as a briefing here.


The COVID-19 pandemic requires governments to balance health risks against social and economic wellbeing. The hospitality industry is Britain’s third biggest employer and has an annual turnover of £130 billion. It was effectively shut down for over three months during the lockdown from March 2020 at great expense to the industry and to HM Treasury. This briefing discusses the arguments for a further shutdown and the likely unintended consequences.

1. Flawed claims about infections in the hospitality sector

Scotland’s Chief Medical Officer justified closing pubs in many areas on the basis of data from NHS Test and Protect showing that 20-25% of infected individuals report having been in a hospitality venue (pub, restaurant, cafe, etc.) in the past week. However, he admits that this is not evidence that the individual was infected in a hospitality venue, nor that they infected others in the venue. Without knowing how many non-infected people visit pubs and restaurants each week, we cannot know whether visiting these venues makes it more or less likely that a person will catch the virus.

The most recent COVID-19 surveillance report for England found that the most common activity of people contacted by NHS Test and Trace in the week ending 27 September was ‘shopping’ (13.3%) followed by ‘eating out’ (13%). A smaller number of infected individuals reported engaging in an ‘activity event’ which includes ‘hospitality’ but also includes ‘arts entertainment or recreation’, ‘community and charity activities’, ‘public events and mass gatherings’, ‘teaching and education’, ‘transport’ and much more. The way the figures are presented makes it impossible to tell how many infected individuals visited pubs, but even if we assume that most of the ‘eating out’ and ‘hospitality’ took place in pubs, the number is likely to be below 20 per cent. Again, we do not know how this compares to the general population. It may be that pubgoers are less likely to be infected. 

According to the Telegraph, Chris Whitty has been lobbying MPs for pub closures on the basis that a larger proportion of infected people under the age of 30 report going to the pub as compared with older people. This only shows that young people are more likely to go to the pub in the current circumstances and is not evidence of widespread transmission in the hospitality sector.

More usefully, NHS Test and Trace has figures showing where infected people have had ‘close, recent contact [with other people] and places they have visited’. As the table below shows, the most common exposure, by far, is in the home. The hospitality sector is classified as ‘leisure/community’, a broad category that also includes ‘eating out, attending events and celebrations, exercising, worship, arts, entertainment or recreation, community activities and attending play groups or organised trips’. Despite the wide range of activities included in ‘leisure/community’, only around five per cent of individuals with the virus report having had close contact with other people in those settings. The amount of close contact in pubs must be even smaller.

2. Local restrictions on pubs have failed to reduce the infection rate

Analysis by the Labour Party shows that local ‘lockdowns’ have failed to reduce the spread of infection (Iacobucci 2020). The partial exception is Leicester where the infection rate fell initially but has since risen again (see table below from the British Medical Journal). 

Leicester and Bolton are of particular interest since they were both forced to close their pubs. Pubs reopened in Leicester on 3 August when there were 27 new cases per day. The infection rate then fell steadily to 13 cases per day in late August, but then began rising sharply in September and is currently at over 80 per day. The rise in cases does not correlate with the opening of the hospitality sector.

In Bolton, pubs were closed from 8 September when there were 88 new cases per day. Shutting down the hospitality sector has not reduced the infection rate which now stands at 109 cases per day.

Nationally, there was no surge in infections after pubs reopened on 4 July. The number of new cases reported remained below 1,000 a day in England throughout July and most of August and only began rising significantly in September. In Liverpool, Manchester and Newcastle, cases did not begin to spike until the second half of September, corresponding with the start of the new university year.

3. The 10pm ‘curfew’ failed to reduce the infection rate and may have made it worse 
The new closing time of 10pm for the hospitality sector has been accompanied by a further rise in the number infections, from an average of around 6,000 in England on the day it was introduced (24 September) to over 11,000 today.

The ‘curfew’ has led to impromptu mass gatherings in town centres and unnecessary crowding on public transport (see photo - taken at 10.10pm on 2 October on the London Underground). There is substantial anecdotal evidence that it has led to legal and illegal house parties as drinkers look for alternative venues to socialise. Informal gatherings of this kind lack the social distancing and other protections provided by the hospitality sector. It is likely that the ‘curfew’ has undermined respect for the law and led to increased transmission of the virus. 

4. The answer may lie in proper enforcement of existing control and mitigation measures in the hospitality sector 

When pubs reopened in early July, they were required to put in place comprehensive protocols around social distancing and other protective measures. Rather than fall back on lockdowns and other stringent restrictions, it may be time to revisit these measures, strengthen them based on lessons learnt, and ensure consistent enforcement.

Interventions that can stem the spread of COVID-19 in public venues are now better understood and have been tested. Face masks and shields, designated areas and physical distancing of tables, limits on the number of patrons and party sizes, and proper procedures for cleanliness and service have all proven their effectiveness. Heated outdoor areas can extend the season of open-air entertainment, and continued contact tracing can further mitigate potential infection. As noted by the Scotland’s Chief Medical Officer, ‘high compliance with all restrictions adopted and mitigating measures put in place will give us our best chance of suppressing the virus without having to implement the most stringent restrictions.’


There is very little evidence to suggest that pubs have been a significant factor in the recent rise in COVID-19 cases in the UK. Changes in the infection rate do not correlate with the reopening of pubs in early July, nor do they correlate with local pub closures in Leicester and Bolton. The 10pm closing time seems to have had no positive impact and has likely made the situation worse.

The hospitality sector already has to comply with increasingly rigid regulations, including social distancing, contact tracing and mandatory table service. Indeed, 85 per cent of pubgoers think their local is complying with, or exceeding, government guidelines. There is no reason to believe that they will not continue to do so.

A substantial majority of new infections appear to be taking place in private households. It is likely that further restrictions on the pub sector will lead to a further increase in illegal and unregulated private gatherings in the home where transmission of the virus is easier.