Friday, 20 May 2022

Jamie Oliver addresses a small gathering in London

Contemptible multi-millionaire Jamie Oliver turned up to speak to the staff of his front group Biteback and a handful of journalists in Westminster today in a protest against affordable food. Waving a high-sugar, high-fat, high-calorie luxury dessert, the failed restaurateur demanded the government ban multibuy deals such as '3 for 2' and '50% extra free' on food that contains modest quantities of fat, sugar and salt. The government is doing this anyway, just a year later than over-the-hill chef would like.
I wrote about this issue for Spiked in the week, so I will leave Tom Harwood to summarise the day's events. I particularly enjoyed the last 15 seconds of this.

Wednesday, 18 May 2022

Do nanny state policies work at all?

Carl Phillips and Marewa Glover have produced a very interesting study assessing the efficacy of anti-smoking policies in the USA. With the exception of tax rises, they find very little evidence that any of them have made any measurable difference to smoking rates and that most, if not all, in the fall in smoking rates would have happened without any legislation. 
Why, then, have smoking rates been declining for decades? Their theory is that it began with the "information shock" of public health authorities, such as the Surgeon General, declaring smoking to be a major health risk in the 1960s. This led to an immediate decline in smoking rates which has echoed down the years through inter-generational effects. As they note, the most robust predictor of individual smoking initiation is parental smoking. When parents quit or never start smoking, their offspring are less likely to smoke.  

And so, by their calculations...

Results: About one-third of the observed prevalence decline through 2010 can be attributed solely to fewer parents smoking after the initial education shock. Combining peer-group cessation contagion explains well over one-half of the total historical prevalence reduction. Plausible additional echo effects could explain the entire historical reduction in smoking prevalence. 

Conclusions: Ongoing anti-smoking interventions are credited with ongoing reductions in smoking, but most, or perhaps all that credit really belongs to the initial education and its continuing echoes. Ensuring that people understand the health risks of smoking causes large and ongoing reductions. The effect of all other interventions (other than introducing appealing substitutes) is clearly modest, and quite possibly, approximately zero, after accounting for the echo effects.

I recommend reading the whole study, but the following section should give you the gist of what they mean by echo effects:

We know that choosing to smoke is socially contagious – the more people around someone who smokes, particularly their parents, the more likely they are to start smoking.1 Parental smoking is the most consistent strong predictor of whether a teenager (of a particular age, in a particular population) will start smoking. Smoking prevalence among siblings, peer groups, and the wider community affects uptake via overt and subconscious social signaling. All of these are taken as fact in the scientific literature and in Phillips et al tobacco control politics, where they are cited as motivation or points of leverage for interventions. But one important implication – that a downward shock or trend in smoking prevalence will, by itself, cause further downward trending for more than a generation – is generally ignored. 

Similarly, smoking cessation is a contagious behavior. This is particularly clear for switching to a lower-risk alternative, wherein the person quitting smoking demonstrates to their social contacts that the choice is appealing and educates them about the alternative. However, even if the choice of cessation method is not affected by social-contact education, the demonstration effect of quitting itself is still powerful. Seeing a friend quit smoking takes it from being an abstract possibility to a concrete example of success. In addition, simply having fewer people who smoke in one’s social circles encourages quitting. Each of these, and all of them together, creates a positive feed-forward effect from any smoking reduction. 

Thus, a one-time permanent downward shock in the popularity of smoking – like that caused by initial education about the harms from smoking – causes a long tail of transition to a new lower equilibrium, echoes of the initial shock. If many people quit smoking, then many more who would have started smoking had they come of age earlier will not do so and others will be motivated to quit over time. The subsequent cohorts coming of age not only will experience the effect of the downward shock, but also be subject to less social contagion. There will be a new equilibrium, but it will only be reached slowly, with a substantial portion of the effect taking more than a generation. This will happen with or without any further efforts to discourage smoking. Subsequent interventions could still have effects beyond the secular trend toward a new equilibrium, of course, but it makes no sense to try to quantify those effects without trying to estimate the background effects of the echoes alone.

Phillips and Glover stress that it cannot be proved either way whether the bulk of anti-smoking regulation has made a difference to smoking rates. They present a hypothesis and a series of models. But it is an intriguing hypothesis and I have often wondered to what extent the tobacco control lobby has been dining out on a decline in smoking rates that would have happened without them (and for many years did happen without them). 

That would certainly help explain why tobacco-style regulation fails to work when applied to other activities. These policies tend to focus on the Three As - affordability, advertising and availability - but whilst it is Econ 101 to note that higher prices tend to lead to lower consumption, albeit at the expense of consumers, the evidence for the other two As is remarkably thin on the ground. 

Take alcohol. A 2019 systematic review titled 'Do alcohol control policies work?' and written by two members of the South African Medical Research Council concluded that ‘[r]obust and well-reported research synthesis is deficient in the alcohol control field despite the availability of clear methodological guidance.’ The policies examined included restricting alcohol advertising and restricting on- and off-premise outlet density.

With regards to advertising, a Cochrane Review, which is usually considered definitive, found 'a lack of robust evidence for or against recommending the implementation of alcohol advertising restrictions'. 

Even the authors of Alcohol: No Ordinary Commodity, the bible of the secular temperance movement, were only able to make a limp case for advertising bans.

‘Imposing total or partial bans on advertising produce, at best, small effects in the short term on overall consumption in a population, in part because producers and sellers can simply transfer their promotional spending into allowed marketing approaches. The more comprehensive restrictions on exposure (e.g. in France) have not been evaluated… The extent to which effective restrictions would reduce consumption and related harm in younger age groups remains an open question.’

A systematic review published in 2012 tried very hard to find evidence to support orthodox, supply-side anti-alcohol policies. It was written by dyed-in-the-wool 'public health' activists, including Mark Petticrew and Martin McKee, but they really struggled to find what they wanted. 

On advertising, they found seven studies which 'provided inconclusive results for the influence of advertising on alcohol use'. 

There wasn't much evidence and a lot of it was of poor quality, but...

A study rated as ‘strong’ in the quality assessment found no significant association between exterior advertising in areas near schools and adolescent drinking.

The authors nevertheless concluded that...

In general, the findings of this review are consistent with reviews on wider alcohol availability (Popova et al., 2009), which have found that availability has a strong influence on alcohol use.

But this is mere editorialising. The evidence they discuss in the paper doesn't show that at all.

In general, the results of this review are similar to those found in previous reviews (Babor et al., 2003)—studies show mixed results but strongly indicate that greater exposure to advertising is associated with higher levels of alcohol use.

How can mixed results strongly show anything?

They also looked at availability - including licensing hours and outlet density - and again struggled to find evidence to support their priors. They found '21 studies on the influence of availability of alcohol from commercial sources on alcohol use', but, alas...

Overall the findings provided inconclusive results for the influence of availability on alcohol use, although some studies indicated that higher outlet density in a community may be associated with an increase in alcohol use.

With regards outlet density specifically:

For off-premise outlets (such as shops), eight studies found no significant association but there is some indication that a higher density of off-premise outlets may be associated with an increased likelihood of heavy drinking. For on-premise outlets (such as bars and restaurants), results were also mixed but there is some indication that a higher density of on-premise outlets may be associated with an increase in the likelihood of drinking and heavy drinking. 

'Some indication' and 'may be associated' are not phrases to fill policy-makers with confidence and are a far cry from the bald assertions of efficacy you hear from the likes of Alcohol Focus Scotland when they appear on television.

As for local changes to licensing regulations...

Four studies (with four effect estimates) looked at the influence of local licensing changes on alcohol use, which included banning alcohol sales and making changes to the hours, days and volumes of alcohol sales that were licensed. They indicate that licensing restrictions may reduce alcohol use, but the evidence is not very robust.

This, remember, is from a group of people who are absolutely committed to clamping down on the advertising and availability of alcohol, and who are putting the best possible spin on the evidence.

The story is much the same when you look at food/obesity. You don't see many randomised controlled trials in the nanny state wing of public health but we have one for the policy of food reformulation. And guess what? It doesn't work. 
Consumption of sugar-reduced products, as part of a blinded dietary exchange for an 8-week period, resulted in a significant reduction in sugar intake. Body weight did not change significantly, which we propose was due to energy compensation.

We also have a RCT for intensive anti-obesity interventions with children over a period of years and they don't seem to work either.

It is often claimed that limiting the number of fast food outlets will reduce obesity, but dozens of studies have looked at the association between proximity to fast food outlets and obesity. The vast majority suggest that there isn't one.

This week saw the publication of a systematic review of food advertising. Again, it was written by fervent interventionists and its lead author is the activist-academic Emma Boyland who is responsible for a fair chunk of the literature herself. She is now not only a professor but also an advisor to thee World Health Organisation.


You won't be surprised that she concludes that governments should restrict food advertising, but it is difficult much of a justification for this in her study.

Evidence on diet and product change was very limited. The certainty of evidence was very low for four outcomes (exposure, power, dietary intake, and product change) and low for two (purchasing and unintended consequences).

Shown in a graphic, the evidence can most charitably be described as 'mixed'.
Even the press release had to come clean:

Their research, published in JAMA Pediatrics, found that food marketing was associated with significant increases in food intake, choice, preference, and purchase requests. However, there was no clear evidence of relationships with purchasing, and little evidence on dental health or body weight outcomes.

If food marketing doesn't have an effect on 'body weight outcomes', there is, of course, no point in restricting it. Obesity and dental health are the only outcomes we're interested in.

None of the studies mentioned above are libertarian hit jobs or industry debunkings. On the contrary, they are written by teammates marking each others homework. The ideological bias and statistical chicanery of many 'public health' researchers will be well known to readers of this blog. If they can't produce persuasive evidence that their policies work, even when reviewed by like-minded friends, we must seriously consider the possibility that none of this stuff does what it is supposed to.

Trying to find common ground on food policy

Sky News have a show presented by Trevor Phillips which tries to find common ground between people who disagree. It is called, reasonably enough, Common Ground and I like the format. It's half an hour long so there is time to get deeper into issues than you get in a standard TV interview. 

Yesterday I was on the show with Thomasina Miers, a restaurateur and food campaigner, to discuss the BOGOF ban and obesity policy in general. She is not a fanatic like Graham 'Mad Dog' MacGregor nor is she a grifter like Aseem Malhotra. She's mostly interested in improving the nutritional quality of school meals, but she also supports things like the BOGOF ban, so we naturally disagreed.

I'm not sure how much common ground we found. Watch it and decide for yourself.

Tuesday, 17 May 2022

Bog off forever

The ban on volume price discounts, such as '3 for 2' deals, for HFSS food has been delayed by a year and will now come into force in October 2023. The advertising ban has also been delayed by a year and will come into force in January 2024. 

Both of them should be ditched for good, as I argue in Spiked today...

The scoundrels of ‘public health’ have persistently claimed that the ban only applies to ‘junk food’. Most of the media have gone along with this conceit, but ‘junk food’ is in the eye of the beholder and has no legal definition. In Britain, the category of HFSS (high in fat, sugar or salt) is used as its nearest equivalent, but this covers far more products than you might expect, including olive oil, raisins and walnuts. 

The absurdity of this came to light when Transport for London introduced a ban on ‘junk food’ advertising and ended up penalising an achingly ethical food-delivery company for showing butter in an advertisement. TfL even had to change its own maps of London to remove offending images of strawberries and cream (Wimbledon) and curries (Brick Lane).

In an effort to avoid a similar embarrassment, the government has decided to define ‘less healthy food’ however the hell it likes. Butter and bacon are no longer on the list, despite both being very high in fat and bacon being high in salt. Forced to name every category of HFSS food that will be covered by the ban individually, the Department of Health has revealed how far-reaching and arbitrary the law will be.

As you might expect, the list includes crisps, pizza and chocolate, but it also includes ‘products made from potato, other vegetables, grain or pulses’; ‘bagged savoury crackers, rice cakes or biscuits’; ‘pitta bread-based snacks, pretzels, poppadoms, salted popcorn, prawn crackers’; ‘ready-to-eat cereals, granola, muesli, porridge oats and other oat-based cereals, bars based on one or more of nuts, seeds or cereal’; ‘croissants, pains au chocolat and similar pastries, crumpets, pancakes, buns, teacakes, scones, waffles, Danish pastries and fruit loaves’; ‘roast potatoes, hash browns, crispy potato slices, potato croquettes’; and ‘fish fingers and fish cakes’.

It also includes shop-bought meal deals of the kind that are popular with millions of people every lunchtime. If any component of the meal deal is deemed to be HFSS – a bit of mayonnaise in the sandwich, some sugar in the drink or a pack of crisps on the side – you will have to buy the items separately at an inflated cost. 

The government’s Impact Assessment expects to see a ‘loss in consumer surplus for consumers who currently make extensive use of price promotions’. That’s one way of putting it. The BOGOF ban should never have made it past the ideas stage and should now be put out of its misery for good.


Sunday, 15 May 2022

Gambling as a 'public health' issue

Martin McKee and Mark Petticrew, two of the most inept 'public health' blowhards in Britain, have bumbled into the gambling debate with a letter in Lancet Psychiatry. McKee has form for talking rubbish about gambling, having been an opponent of the National Lottery way back in 1995. Neither he nor Petticrew have conducted any research into problem gambling, but they recognise that efforts to redefine gambling as a 'public health' issue offers them a new dragon to slay and new opportunities to write opinion pieces about industry websites masquerading as research (which is Petticrew's speciality).

The letter is a bit of a blancmange and it becomes clear straight away that they are out of their depth.

Gambling harms have traditionally been viewed through the lens of psychiatry, psychology, and the neurosciences, with a focus on the individual gambler. This approach reflects an international research agenda that originated with the gambling industry and organisations that it has supported for almost 40 years. The ways in which this literature serves to pathologise people identified as so-called problem gamblers...

The clinical term for 'so-called' problem gambling is pathological gambling. The clue's in the name. It is a psychological problem which can be successfully treated. One of Britain's most prominent anti-gambling campaigners was cured of his compulsive gambling through the use of cognitive behavioural therapy and medication. These things work. 

The approach of the modern, morally bankrupt 'public health' movement is to not get your hands dirty working with individuals but to treat everything as a power struggle with the world of business, and lobby for useless legislation. One of the current fads is for levies on industry which amount to self-serving shakedowns. That's what the anti-smokers want and it is what some anti-gamblers want too. 
The thing is that the gambling industry already provides a lot of money for treatment and prevention, and is prepared to offer more. Such is the moral righteousness of the anti-gambling mob that they would rather tax the general public than accept money from industry. I reported such a case involving the mental health director of the NHS back in February. Some of them seem to have convinced themselves that money from industry is somehow laundered if it comes via a levy instead.

McKee and Petticrew are even opposed to a levy because it would supposedly create a 'conflict of interest'.

Bowden-Jones and colleagues propose a 1% levy on industry earnings to fund independent research. Although this funding would be a clear improvement on the current unsatisfactory system, in which voluntary contributions from the industry are channelled through GambleAware (a charity that both raises industry-derived funds and commissions research), the suggested levy is not a panacea. Linking the available funding to the earnings of the industry would create an obvious conflict of interest, because those involved might hesitate if their actions were to curb profits. We do not understand why public funding for gambling research should not be from government revenues, as is the case with any other threat to health.

This doesn't make much sense. Firstly, a levy would contribute to government revenues. Secondly, by the logic of 'public health' at least, if the industry makes less money (and therefore pays less via the levy) then there must be less gambling-related harm. Thirdly, are they seriously suggesting that researchers will refrain from offering effective solutions because the money will dry up somewhat? That would take a pretty unscrupulous and immoral academic. I believe this is known in psychology as projection.

So what do they propose instead? Insofar as they have a plan, it involves leaving problem gamblers to their fate while 'public health' grandees make pompous speeches about the 'commercial determinants of health' which is code for nutty anti-capitalist activism.

We call for a transformational change in how we conceptualise gambling harms, based on a public health framework that moves away from the current individualistic focus on so-called problem gamblers, takes seriously the upstream drivers of harm (eg, harmful business practices, products, and policies), and prioritises prevention of all forms of gambling harms, with funding mechanisms that are consistent with these goals.

It would be bad news for everybody if these ideologues get their way, but they seem to have some support, such as this chap...

Nothing much to worry about there, you might think. Just some bloke on Twitter who wrongly believes that treatment doesn't work. 

But the bloke who thinks treatment doesn't work is the Clinical Lead & Consultant Psychologist for the NHS Northern Gambling Service. Presumably he'll be handing in his resignation in the morning.

Friday, 6 May 2022

Prohibition doesn't work - South African edition

Mark Petticrew and a few of his headbanging colleagues have written a little diatribe for the BMJ. The gist of it is that arguments and evidence they don't like are a form of 'pollution' and should be controlled as such. Or something like that. I don't recommend reading it as it is largely gibberish, but I was amused by this sentence...

Public health research and advocates were also framed as “nanny staters” or “prohibitionists”6 engaging in “class warfare.”7
The citation for the supposedly unfair claim that 'public health advocates' are prohibitionists is this article from Tobacco Control which literally calls for a total ban on the sale of cigarettes (the authors argue that if they call it 'abolition' rather than prohibition, they will throw the public off the scent).

Being prohibitionists, a number of people in the 'public health' racket got rather excited about the ban on tobacco and alcohol sales in South Africa during the 2020-21 lockdowns. They obviously saw it as a test case for their long term goals. 

The consequences of alcohol prohibition in South Africa were exactly what you might expect. Tobacco prohibition worked out in much the same way, as a new study has found:

South Africa temporarily banned the sale of tobacco as part of its COVID-19 response. Despite the ban, the sale of cigarettes did not cease; rather, it caused major disruption to the cigarette market. The ban inadvertently benefited manufacturers who were previously disproportionately involved in illicit activities; these manufacturers increased their market share even after the ban was lifted. The ban may have further entrenched South Africa's already large illicit market. Our results show that there are unintended consequences associated with a temporary ban on the sale of cigarettes.
You don't say!

Thursday, 5 May 2022

World Health Organisation admits its anti-obesity policies don't work

From The Times...

Britain is on track to be the fattest nation in Europe in a decade, with the rise in obesity driven by takeaway services such as Deliveroo and sedentary lifestyles, a World Health Organisation expert has warned.

By the early 2030s, 37 per cent of British men and women are expected to be obese, a report from global health chiefs says. The present figure is 28 per cent.

Regular readers will be aware that obesity predictions from the likes of the WHO are worthless. As I wrote in 2015... 
In 2006, a Department of Health report predicted that 28 per cent of women and 33 per cent of men would be obese by 2010. Although a prediction for just four years in advance sounds a modest task, 2010 came and went with obesity rates of 26 per cent for both sexes.

In 2007, the well-regarded Foresight report predicted that 'by 2015, 36% of males and 28% of females will be obese'. Figures for 2015 have not yet been published but the most recent data for England show rates of 26 per cent and 24 per cent respectively.

Undeterred by these failed efforts at clairvoyance, The Lancet published a report in 2011 which predicted that nearly half of all British men and 43 per cent of British women would be obese by 2030. It would take an extraordinary surge in prevalence for this to come to pass. Current rates are below The Lancet's lowest confidence interval and appear unlikely to catch up.

Leaving the WHO's prognostication skills to one side, the claim that UK obesity rates will rise from 28% to 37% in the next decade does not say much for the WHO's anti-obesity policies. By 2030, the tax on sugary drinks - which the WHO portrays as one of the 'best buys' for tackling obesity - will have been in place for 12 years. The food reformulation scheme - which is apparently so effective that it is being rolled across Europe by the WHO - will have been in place for 14 years. Heavy restrictions on the promotion and sale of so-called junk food will have been in place for 8 years. Mandatory calorie labelling will also have been in place for 8 years. A ban on advertising food that is deemed high in fat, sugar or salt on TV before 9pm and online any time will have been in place for 7 years.

The WHO would doubtless say that the government should do even more, but this not nothing, is it? Once the last of these policies is rolled out next year, Britain will have the most far-reaching anti-obesity legislation anywhere in the world. If this stuff worked, you would expect obesity rates to fall by at least a few percentage points. Instead, the WHO reckons they will rise by a third and Britain will overtake Turkey and Malta to become the fattest country in the WHO Europe region.

Hardly a glowing endorsement of their own policies, is it?

The impact of minimum pricing

I've summarised the results of my recent research into minimum pricing for Cap-X...

From an economist’s perspective, the unusual natural experiment of minimum pricing offers a chance to see how human beings respond to floor prices. The sale of wine stayed more or less unchanged, but the sale of fortified wine rose by 25%. Much of this came from a surge in the sale of the notorious tonic wine Buckfast which was never sold below 50p per unit to begin with. 

Spirits sales fell overall, but the sale of whisky rose by 11%. The Scotch Whisky Association, who delayed the introduction of minimum pricing through a series of legal challenges, must be wondering why they bothered.

Another interesting finding from our research is that consumers often shifted to significantly more expensive drinks after minimum pricing began. Perhaps this is not so surprising. Minimum pricing effectively wiped out the bottom end of the market and pushed consumers towards the mid-range. It encouraged many consumers to experiment with mainstream brands that they might not otherwise have bought, such as Famous Grouse whisky or Gordon’s gin. If you’re going to spend more money on alcohol, why not buy a brand you recognise? Mid-range brands have never competed purely on price and do not cluster around the 50p per unit price point. 

On a different note, I've written about Ronnie O'Sullivan for Spiked.

Wednesday, 4 May 2022

A swift half with Martin Durkin

The latest Swift Half with Snowdon features the excellent Martin Durkin. He makes all the best documentaries. Check it out.

Tuesday, 3 May 2022

The cost of minimum pricing


When minimum unit pricing was introduced in Scotland in May 2018 it became illegal to sell alcohol for less than 50p per unit. One of the SNP’s flagship policies, it aimed to reduce alcohol-related harms, including death and crime, by raising the price of the cheap, off-trade alcohol that is often associated with harmful drinking. Lacking the power to raise alcohol duty itself, the Scottish government turned to minimum pricing as a way of using the price mechanism to reduce consumption. It was assumed that reducing consumption would lead to a decline in the associated harms.

Even if it worked, this would clearly come at a cost to many drinkers. How much it would cost was contested, with advocates of the policy insistent that moderate drinkers would barely be affected. Although Public Health Scotland is currently evaluating minimum pricing, the financial impact on consumers has not yet been investigated. 

And so, in a study published by the Institute of Economic Affairs over the weekend, John Duffy, Mark Tovey and myself used off-trade sales figures to estimate the cost of minimum pricing to Scottish consumers in the four years since it was implemented. We believe it exceeds a quarter of a billion pounds. Our midpoint estimate is £270 million. That works out at £59.39 per adult or £71.12 per drinker.

Minimum pricing is currently being evaluated by MESAS for Public Health Scotland. So far, they have struggled to find any benefits from the policy. Looking at the data from the past four years, we were also unable to find any discernible impact on crime, absenteeism, unemployment, hospitalisations or deaths, all of which were projected to improve after the policy was implemented. Admittedly, some of the projected benefits were so small that they would be difficult to identify in aggregate data, but all of the key indicators have remained unchanged or worsened since minimum pricing began.

Read the full report here.