Monday, 16 September 2019

Evidence? We don't need no stinking evidence!

The Royal Society of Public Health produced a report last week calling for a ban on fast food shops opening within 400 metres of a school. In effect, this means a near-total ban in most cities.

They also want some form of regulation to clamp down on the shops that already exist...

We are calling for local authorities to introduce planning restrictions on fast food in their local planning schemes; however, it must be noted that this approach is necessarily limited as it can only be used to stop outlets opening, and not to close existing ones. While it is important that comprehensive planning policies are in place so that fast food prevalence near schools cannot increase, it is also vital that councils are able to reduce the current number of outlets near schools (or at least their sales of unhealthy food to school children) – which is already unacceptably high in many areas.

To this end, they suggest banning the sale of food to children and banning shops from opening 'during the post-school hours of the day'.

What could possibly justify such state interference in the out-of-home food market? There must be masses of evidence showing a strong link between fast food outlets near schools and child obesity, right?

Wrong. When I conducted a systematic review last year, I found 74 relevant studies. Of these...

… only fifteen (20%) found a positive association between the proximity and/or density of fast food outlets and obesity/body mass index. Forty-four (60%) found no positive association, of which eleven (15%) found evidence that living near a fast food outlet reduced the risk of putting on weight. Fifteen (20%) produced a mix of positive, negative and (mostly) null results, which, taken together, point to no particular conclusion.

And the evidence that fast food availability causes obesity among children is even weaker... 

Of the 39 studies that looked specifically at children, only six (15%) found a positive association while twenty-six (67%) found no effect. Seven (18%) produced mixed results. Of the studies that found no association, five (13%) found an inverse relationship between fast food outlets and childhood obesity. Two-thirds of the studies found no evidence for the hypothesis that living near fast food outlets increases the risk of childhood obesity and there are nearly as many studies suggesting that it reduces childhood obesity as there are suggesting the opposite.

So how does the RSPH justify its proposals? It says...

A number of studies both at home and abroad have indicated a significant link between repeated exposure to fast food outlets along daily commutes, fast food consumption, and obesity.

'A number' is doing a lot of heavy lifting in that sentence. Does the RSPH summarise the literature? Does it cite any of the seven systematic reviews that have been conducted, all of which concluded that the link between fast food outlet proximity and obesity is extremely weak, at best?

It does not. The RSPH cites just two studies. One of them is a recent study from America which found an association between the body mass index of adult employees (not children) and the number of supermarkets, grocery stores and fast food outlets on their commute route (but not near their workplace).

The other is a study from the UK which showed that adults (not children) who live near lots of takeaway outlets consume fewer calories and had a slightly lower body weight than those who didn't. As the statistician Jeremy Franklin pointed out, the authors had to make some dramatic adjustments to the data to turn those findings on their head, and even this required supermarkets to be defined as 'takeaway food outlets'. If you exclude supermarkets, there is no association under any model.

Not the most compelling evidence then, but even if the studies were stronger it would still be wrong to cherry-pick two studies out of 70-odd.

'Public health' is not an evidence-based enterprise. The RSPH have found a policy they like and they want to do it because they want to do it.

Thursday, 12 September 2019

Public Health England admits that its policies won't reduce childhood obesity

According to the Daily Mail...

Childhood obesity rates in the UK are to continue to spiral over the next five years, a damning report has found.

Continue to spiral? Let's have a look at the existing trend from the Health Survey for England.

If it looks like the rate hasn't risen for fifteen years and is lower than it once was, congratulations, there is nothing wrong with your eyes.

In its latest forecast, Public Health England predicts the number of obese primary school children could jump by up to four per cent by 2024.

More than 34 per cent of 10 and 11 year olds are currently classified as obese, with the new forecast now between 33.4 per cent and 38.1 per cent.

No. That's the number of children who are classed as overweight or obese. The childhood obesity measure has no basis in reality, and the overweight measure is even more absurd. None of the children who are classed as overweight are fat, nor are most of those who are classed as obese.

If you look at the projection, PHE reckons the number of 11 year olds with 'excess weight' will rise from 34.3 per cent to 35.7 per cent, but the confidence interval ranges from 33.2% to 38.1% so it doesn't really tell you anything, even if obesity predictions were based on something more than guesswork, which they aren't, and if the government's measure of childhood obesity reflected childhood obesity rates, which it doesn't.

In any case, obesity forecasts are consistently wrong.

Professor John Newton, the PHE's director of health improvement, admitted it was a 'reminder we need to redouble our efforts on childhood obesity'.

Ah, yes. Doubtless this was the entire purpose of this worthless prediction.

But he claimed it was 'important not to interpret this trend as a sign that what we're doing at the moment isn't working'.

Heaven forfend! What a balancing act Professor Newton is having to perform. It must be tricky trying to scare the public into believing that the situation will get worse having introduced policies that you insist will make the situation better.

By 2024, Public Health England intends to have taken 20 per cent of calories out of food. Its target for taking 20 per cent of sugar out of food is supposed to have been reached by next year. By 2024, the sugar levy will have been in place for six years. And yet the impact of all this (and more) on childhood obesity will, according to PHE's own figures, be diddly-squat. Can we have our money back please?

Out of around 556,000 children of primary school-leaving age in the UK, 170,000 are overweight to some degree, figures showed in May last year.

Where are they? Show me them.

Professor Newton also called for tighter advertising laws around promoting junk food to children. 

Of course he did.
'Advertising is the other thing - there is good evidence we need to ensure children and families are not bombarded by advertising for unhealthy foods,' he said. 

Thus, Public Health England continues to sell us a web of lies. Its measure of childhood obesity is obviously, demonstrably, laughably wrong. Its projections are based on nothing. Its anti-obesity policies have not worked in the past and will not work in the future.

Can we shut this money pit down now?

Wednesday, 11 September 2019

The American Vape Scare

The USA is in the grip of a classic moral panic about vaping. It’s been years in the making and is now reaching its crescendo.

The spark came from a spate of health problems caused by black market THC vape cartridges which seem to have been cut with dangerous thickening agents. America’s mendacious anti-vaping lobby has capitalised on these individual tragedies and seems on the verge of banning e-cigarette flavours.

They are beneath contempt, as is that hateful skeleton Mike Bloomberg who has just given £160 million to science-denying bottom-feeders to stamp out vaping.

I have written about this for Spiked. Have a read.

Tuesday, 10 September 2019

Sugar levy cash gets swallowed

Last summer the IEA published a little paper written by me about sin taxes. In it, I mentioned the ringfencing trick in which politicians and campaigners lure voters to sin taxes by promising the money will be hypothecated for feelgood causes.

.. history shows us that tax hypothecation is rare. The public are more likely to support sin taxes if they are told that the revenue will be ring-fenced for health or education, but the money is usually diverted towards general state spending sooner or later (Hoffer and Pellillo 2012; Jacobson and Brownell 2000). Tobacco and alcohol taxes have been used to finance routine government expenditure for centuries and if the USA is any guide, sugary drink tax revenue will be treated no differently (Tanenbaum 2018).

Recent examples of soda tax revenue being syphoned off to the general pot after being promised for cuddly initiatives include Philadelphia and Oakland.

When the UK government promised to use the filthy lucre from the sugar levy to fund a school's breakfast programme and school sports, I thought there was a chance that it would happen. It was such a high profile commitment.

But it seems that I was not cynical enough...

Sajid Javid admits Treasury has swallowed sugar tax cash

The Treasury has pocketed taxes raised through the soft drinks sugar levy that were supposed to support children’s health, new chancellor Sajid Javid has admitted.

This week’s spending review promised an extra £13.4bn in public spending for 2020-21 as he promised to “turn the page on austerity”.

However, the Treasury confirmed that earlier commitments to ringfence the taxes raised by the levy to help tackle the obesity crisis in schools had been dropped.

Forecast to raise £340m in 2020-1, the prospect of the sugar tax disappearing into the Treasury’s coffers appeared was confirmed as the review failed to contain any commitment to ringfence the income for spending on programmes to improve children’s health and food.

And the earth still shifts on its orbit.

It was met with fury by campaigners who accused the government of backtracking on a promise to help children in schools via the tax.


“We’re profoundly disappointed not to get a clear reassurance from the chancellor today that the levy on sugary drinks would continue to deliver a direct dividend for children’s health,” said Children’s Food Campaign co-ordinator Barbara Crowther.

Might as well scrap the tax then, eh?

Monday, 9 September 2019

Soda taxes still not working

A noted anti-obesity campaigner was recently celebrating Scotland introducing more nanny state laws and promised that evidence of their efficacy would be forthcoming. When I asked her about this evidence, she told me that you would have to be literally crazy to expect an anti-obesity policy to actually reduce obesity.

Note that I didn't ask for this tranche of policies to 'solve' obesity, merely that they would reduce it somewhat. She seems to think that they won't, so that makes two of us.

Meanwhile, real world evidence on sugar taxes continues to appear.

Oakland's Sugar-Sweetened Beverage Tax: Impacts on Prices, Purchases and Consumption by Adults and Children

We find that roughly 60 percent of the tax was passed on to consumers in the form of higher prices. There was a slight decrease in the volume of SSBs purchased per shopping trip in Oakland and a small increase in purchases at stores outside of the city, and we find some evidence of increased shopping by Oakland residents at stores outside of the city.
We do not find evidence of substantial changes in the overall consumption of SSBs or of added sugars consumed through beverages for either adults or children after the tax.

Yet another 'public health' win!

Friday, 6 September 2019

The Social Market Foundation's alcohol tax report

In a report that was "kindly supported" by the Institute of Alcohol Studies (neé UK Temperance Alliance), the Social Market Foundation calls for a reform of Britain's alcohol duty system. They rightly point out that the system is incoherent and irrational. I said likewise in 2017 when I proposed taxing all units of alcohol at the same rate to make the system genuinely Pigouvian. This implied a flat rate of 9p per unit and a large decline in the amount raked in by the government.

The SMF are broadly right about what is wrong.

While a relatively weak 6% ABV bottle of wine faces duty of 50 pence per unit of alcohol, as of the time of writing a 6% ABV cider faces duty of just 7 pence per unit of alcohol.

This is because the UK has traditionally imported its wine while producing its own cider, but that is not a valid reason.

Given that alcohol consumption generates externalities – impacting government, families and other individuals as well as the direct consumers of alcohol – the overall rate of alcohol taxation should be reflective of the size of these externalities. 

It would be easy to misread this as saying that the impact on 'direct consumers of alcohol' is an externality. Obviously it isn't, but assuming that's not what they mean, they are right. 

Brexit could open up possibilities to rationalise alcohol taxation in the UK.

Very true.

Given the political nature of duty changes, the UK has been lumbered with a system of alcohol taxation ill-suited for meeting key objectives of government: raising tax revenue, protecting jobs or improving public health and other social outcomes. Rather than examining the evidence base and reflecting on the alcohol duty regime as a whole, the system has been tweaked according to political whims over time.

Also very true.

Arguments often used to justify duty freezes and favourable treatment for certain beverages are deeply flawed. For example, jobs-based arguments used to justify cider and spirits duty freezes ignore the fact that cider accounts for a very small number of jobs in the economy

Indeed. Economists are generally dismissive of jobs-based arguments. A reduction of employment in one area tends to lead to a growth in another.

But the SMF is also wrong about quite a lot too...

Jobs-based arguments also ignore work lost through excessive alcohol consumption. Analysis by Public Health England found that, in 2015, there were 167,000 working years of life lost due to alcohol consumption – 16% of all working years lost in that year.

That's not a jobs argument. People dying doesn't cause unemployment, just as people being born doesn't reduce unemployment.

Arguments related to the regressive nature of excise duties are also flawed. Firstly, alcohol taxation does not appear to be particularly regressive, given relatively high rates of non-drinking among lower income households.

Only if you look at people as groups. If you look at individual drinkers, it is clearly regressive.

Secondly, regressivity alone is not a strong argument against alcohol duty.

That's a matter of opinion.

Studies suggest that the total costs of alcohol to UK society could stand at between 1.3% and 2.5% of GDP.

This is a statistic that first appeared in Public Health England's woeful report on alcohol which quickly turned into the claim that alcohol misuse costs society 'up to £52bn a year, far more than previously thought'. This is based on looking at a range of international estimates which do indeed range from 1.3% to 2.5% and picking the biggest one.

But guess where the cost is estimated at 1.3%. That's right: the UK. So it's not necessary to use a range of estimates from around the world to estimate the UK figure when one of those figures is the UK itself, is it? This is a small but typical example of the 'public health' lobby lying with statistics.

Moreover, that estimate of 1.3% (£21 billion) is nearly 20 years old and is mostly made up of costs that are not external.

Hazardous and harmful drinkers account for a staggering 78% of alcohol consumed in England. 

This is based on the discredited lie that it is hazardous to drink more than 14 units a week.

There is also growing evidence suggesting that even moderate rates of alcohol consumption are associated with higher risks of cancer.

That evidence is very weak.

The SMF have five main policy recommendations:

1. Introducing a duty strength escalator, to focus alcohol duty on the higher strength products disproportionately consumed by heavy drinkers, and create stronger incentives to produce lower strength products.

If you have a flat tax on each unit of alcohol, stronger drinks are taxed more almost by definition. That should be sufficient. If, however, it could be shown that certain drinks create greater externalities, a case could be made for the kind of system proposed by the SMF. There is some evidence for this, but it is far from clear that taxing those drinks more would not simply displace consumption, with the effect that another form of alcohol becomes associated with the same externalities.

2. Levelling the playing field across same-strength products. Products of the same strength should face the same rate of duty and duty should be a function of the pure alcohol content of drinks, rather than the volume of the final product. This would help simplify the alcohol duty system.

.. We recommend taxing alcohol on a consistent basis, according to the pure alcohol content of the beverage.

So do I.

3. Allowing pubs to claim back a proportion of alcohol duty through a new “Pub Relief”. This would focus alcohol duty on the off-trade, which is particularly reliant on sales to hazardous and harmful drinkers.

.. We assume that our proposed “Pub Duty Relief” rate is set at 50% - meaning that duty rates in the on-trade are half as much, per unit of alcohol, than on the off-trade.

No one loves pubs more than me, but this sounds like the kind of tweaking of the system according to political whims that the SMF complain about at the start of their report.

To be fair, they come up with a superficially plausible justification for what seems like rent-seeking...

...heavier drinkers are more likely to consume alcohol on the off-trade. Therefore, in line with our proposed framework of focusing alcohol duty where harms are most generated, we suggest explicitly favouring on-trade consumption of alcohol in the tax system.

This could be right but there are reasons to doubt it. Whilst the majority of health harms are linked to off-trade alcohol (most alcohol is sold on the off-trade, after all), those harms are to the drinker and are thus not externalities (for the most part, at least). It is less clear that the external harms are disproportionately associated with the off-trade. Most alcohol-related violence takes place in and around the on-trade. Drink driving is also, I suspect, more closely linked to the on-trade than the off-trade.

In any case, if this is an attempt to 'support the pub sector', as the BBC claims, the SMF's own figures show that it won't work:

The impact of switching to the duty regime described above would be to cut overall alcohol consumption by 5.4%, compared to the current regime. This would come from a 1.9% decline in on-trade alcohol consumption and a 6.8% decline in off-trade consumption.

On-trade consumption of alcohol falls, in the model, despite declines in on-trade prices. This reflects cross elasticities of demand for alcohol products. Increased prices in the off-trade can reduce demand in the on-trade as well.

I have explained one of the reasons for this here.

4. Explicitly linking alcohol duty to the social costs of alcohol, rather than treating it as a cash cow. At the very least, alcohol duty should cover the health, crime and welfare costs to government and wider society (the “externalities” associated with alcohol consumption). 

It should cover the net externalities, yes.

At this point, regrettably, the influence of the 'Institute of Alcohol Studies' becomes apparent...

Paternalistic arguments, which consider the ability of individuals to make “bad” and regrettable lifestyle choices (such as those that undermine their job prospects), could justify a higher tax take.

Coercive paternalism has no place in a free society and there is no point producing a lengthy report about alcohol taxation if you're going to throw out your economic analysis and resort to saying 'but muh paternalism'.

.. We do not take a strong position on the degree of paternalism that should be embedded in the alcohol duty system in this report – ultimately, we believe this is a matter for politicians and the electorate. 

You work for a think tank, for God's sake. Get off the fence!

Our key argument is that tax take from alcohol duty should be focused on alcohol-related costs, whether that be the externalities of alcohol consumption or a broader measure covering private costs drinkers.

Drinkers already pay the private costs. If you pretend those costs are externalities and add them to the tax, they end up paying twice. There is no ethical or economic justification for that.

5. Regularising the uprating of alcohol duty, with inflation or earnings uprating being the “norm”. This would help depoliticise the setting of alcohol duty.

I'm comfortable with automatically increasing the tax by the rate of inflation once we've reformed the system. In practice, that means significantly reducing the rate of tax charged on most, but not all, drinks.

Thursday, 5 September 2019

Next up: the snack tax

With crushing inevitability, the first UK modelling study promoting a tax on food has been published in the BMJ. And guess what? The researchers reckon that it will work. In fact, they think it will work even better than taxes on sugary drinks - which have never worked anywhere. An accompanying editorial in the journal gives the policy its blessing.

The BMJ's modelling study promoting sugary drink taxes has now been largely forgotten, despite being highly influential at the time, but it promised a fall in obesity of 1.3 per percentage points. Today's study predicts a decline of 2.7 percentage points.

One of the authors told the Guardian:

Dr Pauline Scheelbeek from the London School of Hygiene and Tropical Medicine, the lead author of the study, said a snack tax could cut obesity in the UK population from about 28% to about 25%. “That is, on a population level, a huge impact,” she said.

'Public health' has a wretched record of health-economic modelling. It repeatedly promises outcomes that are not delivered and then uses retrospective modelling to make it look as if they were (eg. Mexican sugar tax, English smoking ban).

Models are only as good as the assumptions fed into them by the researchers. If the researchers tell the model that the policy will work, all the model can do is estimate how well it will work to the nearest decimal point.

The modellers of this new study don't look at how much their tax on confectionery, biscuits, and cake will cost consumers, but it would clearly run into many millions of pounds and hit people on low incomes the hardest. The researchers brush this off with the usual rhetoric about the health of the poor benefiting the most, despite there being no evidence that this is how sin taxes operate in practice.

I would bet a pound to a pork scratching that if the tax proposed has any effect on obesity it will be far, far smaller than is reported in this model. But let's imagine that it is correct. Let's imagine that it really does reduce the adult obesity rate from 28% to 25%. Do you think the nanny statists will congratulate themselves on a job well done and leave the public alone? Or do you think they will devise new ways to tax and control us?

You don't need a wonky predictive model to work that one out.