Sweet Truth: an excerpt

I'm on holiday this week so the next four posts will all be excerpts from recent IEA publications, starting with this from Sweet Truth (written with Rob Lyons)...

While Action on Sugar claims that there is underhand lobbying going on, it also trumpets the fact that the health secretary, Jeremy Hunt, had asked them to produce some proposals for tackling obesity (Action on Sugar 2014). Since these proposals are quite typical of this kind of campaign, it is worth looking at them in some depth. With only one exception, they are all predicated on the belief that consumers are unable, rather than unwilling, to make the ‘correct’ choice about what to eat. 

Government limits on the use of sugar

1. ‘Reduce added sugars by 40 per cent by 2020 by reformulating foods (a similar programme to salt)’

At the heart of Action on Sugar’s policy agenda is the assumption that consumers are faced with a lack of real choice in the food market because so much food is ‘spiked’ with sugar. Constrained by an ‘obesogenic environment’ which makes healthy choices difficult, if not impossible, consumers are unable to fulfil their true desire to eat less sugar and fat. Action on Sugar’s solution is to force companies to systematically reduce the amount of sugar in the food supply. (Note the anti-sugar campaigners’ schizophrenic attitude to big corporations when it comes to policy prescriptions, seeing them as a way of bypassing consumers to achieve their ends in this instance.) However, as we have seen, the amount of sugar consumed in Britain has fallen significantly in recent decades and there is no evidence that people are constrained by a lack of choice.

On the contrary, there are many low-fat, low-sugar and low-calorie options on supermarket shelves. The fizzy drink market is a good example of this. No products raise the ire of obesity campaigners more than carbonated beverages and yet there is an abundance of low-calorie options available. Diet Rite was launched in the US in 1958 and was followed in 1963 by Coca-Cola’s equivalent, Tab. Coca-Cola alone now produces Diet Coke, Coke Zero and Coke Life. Diet Coke and Coke Zero contain no sugar and Coke Life is a lower-calorie version of Coke, made with a combination of sugar and natural sweeteners. Nearly all soft drink companies produce similar low-calorie and sugar-free varieties. All are widely advertised and all are available on the same shelves, in the same shops and for the same price as their more sugary cousins. It is very difficult to argue that consumers are nudged, let alone coerced, into buying the high-calorie variants. Indeed, Diet Coke overtook Pepsi in 2010 as the second-most popular soft drink in the US (Riley, 2011) and has retained that position since (Beverage Digest, 2014).

In many other categories, there is already a wide range of products with markedly different sugar contents. For example, in breakfast cereals, Frosties contains 37 grams of sugar per 100 grams, but the otherwise identical Corn Flakes contains just three grams per 100 grams. Shredded Wheat contains no added sugar at all. Since there is a large market for heathier foods, companies heavily promote low-sugar and low-fat products and consumers can easily choose what level of sugar they want in the food and drink they buy. To reformulate brands by diktat with the simple aim of reducing sugar is both unnecessary and would limit choice.

Nevertheless, campaigners have called for mandatory reductions in sugar content, a policy that is explicitly modelled on the salt reduction scheme that was agreed between industry and the government of Tony Blair (MacGregor and Hashem 2014). It is unsurprising that Action on Sugar should propose such a move as it shares its key personnel, website and charity registration number with Consensus Action on Salt and Health, but, as one food industry insider told us, reducing salt content is relatively easy compared to reducing sugar. Salt has a role as a preservative and a flavour enhancer, but it can be cut back to a degree over time without driving too many customers away. Reducing sugar content by such a large amount would be a challenge and the resulting products would be very different. Biscuits, cakes and confectionery with markedly less sugar would lose much in texture and flavour. Sugar-sweetened beverages could swap sugar for an artificial sweetener, but many people dislike the taste of saccharin or aspartame (aka Nutrasweet).

2. ‘Reduce fat in ultra-processed foods, particularly saturated fat – 15 per cent reduction by 2020’

For a group called Action on Sugar, this is an odd demand, and seems directly contrary to the views of some of its leading lights, who have professed that dietary fat is not a problem. It does indicate, however, that it is mass-manufactured food - and the companies who make it - that is the real target.

The fact remains that choice is not a problem for buyers of food in twenty-first century Britain. Indeed, there is so much choice that some critics complain of being overwhelmed by the 40,000 products that sit on supermarket shelves (Hastings 2013: 37). The food industry has responded to people’s concern with a vast range of low fat and low sugar products in accordance with the scientific consensus of the day. The idea that the government should force manufacturers to reduce sugar and fat content in individual products by an arbitrary percentage is not a response to a lack of choice in the market, rather it is a response to the fact that many consumers exercise their choice by rejecting the low calorie options. 

Advertising bans

3. ‘Cease all forms of marketing of ultra-processed, unhealthy foods and drinks to children’

There is an enormous amount of economic evidence showing that advertisements for established products increase demand for specific brands but do not increase demand for the entire product category (Bagwell 2007, Schmalensee 2008). An advertisement for the butter-like spread Clover, for example, might increase sales of Clover, but is unlikely to increase sales of fatty spreads overall. Despite the evidence that advertising reflects, rather than controls, primary demand, public health campaigners continue to regard advertising as a powerful corporate weapon of coercion which, if curtailed, would lead to consumers abandoning tobacco, alcohol and high-calorie foods. The mere fact that all these products were consumed on a massive scale long before the advent of advertising does not shake their conviction, nor does the conspicuous failure of advertising bans to reduce the consumption of these products in recent decades (Duffy 1995, Qi 2008). Health campaigners are also largely indifferent to the benefits that advertising provides, such as funding the media, incentivising high quality standards and allowing new entrants to break into the market (Harris and Seldon 2014).

In Britain, a broadcasting ban on advertising of foods high in salt, fat and sugar in programmes with a large audience of children has been in force since 2007. The ban appears to have been a significant factor in the closure of ITV's children's television production department. According to one report at the time, 'it is estimated that round 70 per cent of the cost of children’s programmes is funded from advertising and a significant proportion of that comes from food advertising' (Thomas, 2006). Yet the ban has been so ineffective that there have been calls to extend it to any programming before the 9pm 'watershed' (Wallop, 2010). Such a move would be unlikely to fare much better. Children have plenty of direct access to 'junk' food without the need to advertise it, through shops on the way to and from school, for example. In any event, most of the products that campaigners rail against - such as breakfast cereals and ready meals - are bought by parents, not children. (It is because parents frequently watch television with their children that advertisers pick these slots. Campaigners claim that food companies are attempting to harness ‘pester power’ but this cannot explain why payday loans, cleaning fluids, laundry detergent and other adult-oriented products are also advertised during children’s programmes.)

Economist’s assumptions about consumer sovereignty and rationality do not necessarily apply to children (Cawley 2011: 132), but it is not difficult to see how policies ostensibly aimed at ‘protecting’ children from ‘exposure’ to advertising could restrict communication between businesses and adult consumers. Whereas the UK’s existing ban targets programmes that are predominantly viewed by children, a watershed ban targets programmes that are predominantly, though not solely, viewed by adults. It would effectively confine the promotion of a vast swathe of food and drink products, including cheese, bacon, butter, cakes and biscuits, to a few hours late at night. Any kind of broadcast ban would likely reduce the quality and quantity of television programmes and would limit the right of producers to tell the world about their products.

Anti-sugar campaigners have almost as dim a view of consumers as they have of corporations. It is assumed that we are clueless and unwitting fools who accept marketing at face value - a case of 'monkey see, monkey do', a frequent and false claim made against advertising generally. Therefore, the 'experts' - like Lustig and Malhotra - must step in to save us from ourselves. No doubt, campaigners would point to the ban on tobacco advertising and say that the end - the possibility of better health - justifies the means. In our view, free speech - including truthful and honest advertising - should be an absolute in a free society. Most people would agree that alcohol, which is acutely poisonous and can create a potentially dangerous loss of physical control, should not be sold to minors. Cigarettes are a risky product with long-term health implications, so preventing their sale to children is also widely agreed to be sensible. But preventing the advertising of legal food products generally considered to be safe is, in effect, the government choosing what can and cannot be said. There is a commonly used name for that: censorship.

4. ‘Disassociate physical activity with obesity via banning junk food sports sponsorships’

Food campaigners get particularly agitated about the sponsorship of major events such as the Olympics by fast food and soft drink companies. Yet while such advertising may be useful as a means of promoting one brand versus another, there is little evidence to suggest that total sales of fast food are increased by sponsorship of sporting events. In the case of the Olympics, the Olympic Park at London 2012 itself was free of such advertising. Sporting sponsorship may qualify under the remit of 'corporate social responsibility' - companies trying to look good by 'giving something back' to the community. But to suggest that sponsorship automatically leads to sales is to reiterate the view that consumers are dim-wits and flies in the face of a large body of economic evidence.

Moreover, if we are simply vessels for whatever combination of images are displayed to us, surely all this sporting sponsorship would be encouraging the consumers of Coca-Cola, McDonalds and the rest to put on their training shoes or hit the gym? Clearly, that is not happening.

Restricting availability

5. ‘Limit the availability of ultra-processed foods and sweetened soft drinks as well as reducing portion size’

As with mandatory reductions of sugar content in food, restrictions on the sale of food and drink products are an attack on choice. Never mind the problem with defining an 'ultra-processed' food, it suggests a return to the 1970s, when the availability of takeaway food was limited and most shops closed by 7pm. It fits with a policy among some councils of banning takeaway food shops in the vicinity of schools, too. Returning to an earlier example, Waltham Forest council in east London has reportedly turned down 83 per cent of applications to open fast food shops in the past five years (O’Brien, 2014). Such an approach stifles competition, favours incumbents, and distorts the market by preventing supply from meeting demand. This is likely to result in higher prices and poorer quality.

The most high-profile attempt to limit soda portion sizes was in New York. In September 2012, the city's health board approved a plan by the then mayor, Michael Bloomberg, to limit servings in restaurants, cinemas and sports venues to 16 ounces (473ml), though there would be no limit on how many portions a customer could buy, nor would there be any limit on refills. However, the ban would not have applied to convenience and grocery stores. As a result of the uneven application of the ban, the courts ruled the health board had gone beyond its powers (Grynbaum, 2014).

'Nudge' theorists would argue - probably correctly - that even if a customer wanted to order a 32-ounce or even 64-ounce portion - as in 7-Eleven's much maligned 'Double Gulp' - they would be much less likely to do so if they needed to order more than one portion. Yet such a ban would have a limited impact. In the UK, for example, the largest serving of soft drinks at McDonalds is 500ml - barely more than Bloomberg's suggested maximum.

Sin taxes

6. ‘Incentivise healthier food and discourage drinking of soft drinks by planning to introduce a sugar tax’

More than anything else, taxes are the most widely suggested idea for how to reduce sugar consumption. Basic economics suggests that if you make something more expensive, demand for it will decline. In November 2014, the Californian city of Berkeley became the first to vote for a large soda tax - in this case, of one cent per fluid ounce. However, whether imposing taxes is effective, fair or efficient is another matter.

As an aside, it could be argued that a ‘junk food’ tax, including on fizzy drinks, is already in place, in effect, in the UK. VAT is not charged on raw meat and fish, vegetables, fruit, cereals, nuts and pulses, bread, and a variety of other foods. But VAT at the standard rate of 20 per cent is charged on alcoholic drinks, confectionery, crisps and savoury snacks, hot food, sports drinks, hot takeaways, ice cream, soft drinks and mineral water. While the fit between ‘healthy’ and ‘unhealthy’ foods and VAT-rating is not perfect (cakes are zero-rated, mineral water is standard-rated), by and large the foods that are supposed to be bad for us are 20 per cent more expensive than they would otherwise be thanks to the tax system. While HMRC does not keep separate figures for these categories, a rough calculation would suggest the income from 20 per cent VAT on snacks (sales in 2013: £3.2 billion (SNACMA, 2014)), confectionery (£5 billion (IBIS, 2015)) and added-sugar soft drinks (£5 billion (BDSA, 2014)) is in the order of £2.6 billion per year.

On the question of effectiveness, soft drink buyers appear to be relatively 'price inelastic', that is an increase in price translates to a reduction in sales, but the percentage decline in sales is less than the percentage rise in price. The economics literature suggests that soft drinks have a price elasticity of 0.79, meaning that a price rise of 10 per cent should reduce consumption by 7.9 per cent (Andreyeva et al. 2010). The confidence interval of this 0.79 figures is very wide, however, and price-sensitive shoppers could avoid a price rise in a variety of ways. Anti-sugar campaigners hope that shoppers will switch to drinking water or, at the very least, switch to sugar-free versions of popular drinks from the much-maligned Coca-Cola and Pepsi. But as mentioned earlier, many dislike the taste of artificial sweeteners (and some fear that they are carcinogenic). Alternatively, price-sensitive shoppers could avoid a price hike by buying larger sizes with lower unit costs or switching to cheaper brands. For example, a two-litre bottle of Tesco own-brand cola is about one-third of the price of Coke. Indeed, ‘private label’ brands (own-brands) took almost as much in sales in supermarkets and stores (£1.51 billion) in 2013 as Coke-branded (£1.19 billion) and Pepsi-branded (£362 million) drinks combined, suggesting that the volumes of own-brand being sold in stores far outstrip the two biggest brands (Britvic, 2014). On the other hand, those with a strong attachment to a particular brand and flavour will simply pay the higher price. Moreover, given the high price of their products in comparison to the cost of production, there is plenty of room for the big beverage brands to cut prices in order to maintain sales. That might hurt their bottom line, but it would reduce the impact of a soda tax on sugar consumption.

There are also practical questions about how such a tax would apply to lower-sugar brands such as Coke Life. If the aim is to encourage reductions in sugar content, applying a flat-rate tax to all sugary drinks would discourage drinks manufacturers from producing lower-sugar alternatives. And how would fruit juices be dealt with when they are still part of the 'five-a-day' message, according to the NHS Change4Life website, but frequently contain as much sugar as fizzy soda?

Which brings us to the question of effectiveness: how much effect does a tax on one supposedly causative element in the problem of obesity actually have? There are numerous confounding factors. For example, Ketan Patel, a doctoral student at Northwestern University in Chicago, argues that the effect of a soda tax on obesity would be limited by the fact that obese people already strongly prefer sugar-free drinks. If calorie consumption from soft drinks falls due to a soda tax, there is also the problem of calorie-substitution - people eating or drinking other products with the effect of offsetting any gain in calorie reduction due to the tax.

A review of the effect of soda taxes by Greek researchers concluded: 'The effectiveness of a taxation policy to curb obesity is doubtful and available evidence in most studies is not very straightforward due to the multiple complexities in consumer behavior and the underlying substitution effects. There is a need to investigate in-depth the potential underlying mechanisms and the relationship between price-increase policies, obesity, and public health outcomes' (Maniadakis, 2013). An American review concluded: 'The limited existing evidence suggests that small taxes or subsidies are not likely to produce significant changes in BMI or obesity prevalence but that non-trivial pricing interventions may have some measurable effects on Americans' weight outcomes, particularly for children and adolescents, low-socioeconomic status populations, and those most at risk for overweight' (Powell, 2009).

Other studies find some positive health benefits from such taxes, but only at fairly high rates and usually only in theory (computer models produce most of the evidence for soda taxes). Taxes that have been implemented in the real world seem to have been too low to have any impact. A 20 per cent tax rate might have some effect, but it would be a difficult sell, politically. Moreover, as with so many sin taxes, the idea really only becomes attractive to politicians in desperate need of revenue. So in 2010, the unfortunately named mayor of Philadelphia, Michael Nutter, proposed a soda tax to help fill the city's $120million budget deficit, promising that 'some' of the money might go to health programmes (NBC Philadelphia, 2010).

Even if such a tax were to have positive health benefits, they would come at a high price. Small reductions in the incidence of obesity and, perhaps, cardiovascular disease would be paid for by large aggregate costs across society. Moreover, the tax would be thoroughly regressive. Firstly, because it would apply no matter what a person's income was, with the poor paying the same tax as the rich. Secondly, because sugary drinks are more popular among the poor, they would actually pay more tax than the well-off, not just proportionally, but in absolute terms.

From an economic perspective, the key question is whether such taxes will be efficient. The best economic justification for a ‘sin tax’ comes when the use of a product creates negative externalities which have to be paid for by others. In the field of public health, this typically means publicly funded healthcare costs. But the tax itself is an externality for consumers of the product. It reduces their consumer surplus. Both the costs and the benefits must be weighed before proceeding.

In Britain, obesity is said to cost the NHS £5 billion a year in healthcare costs. This is widely portrayed as an unnecessary burden on taxpayers, particularly on those who eat healthily and exercise. A tax of 20 per cent on sugary drinks has been proposed as a way of reducing this burden. One of the proponents of the tax, the Children’s Food Campaign, estimates that it would reduce healthcare costs by £15 million a year, but their own figures show that the tax itself would relieve taxpayers of £1,000 million a year (Boseley 2015, Children’s Food Campaign 2013). The cost of the tax would therefore vastly exceed the savings.

As a means of cutting healthcare costs, sugar and soda taxes would be highly inefficient, and as a way of relieving the burden on taxpayers would be counter-productive. It could be argued that a sugary drink tax would at least result in sugar fiends shouldering more of the healthcare costs, but that would not occur in the example above. The Children’s Food Campaign does not anticipate any of the revenue being put towards healthcare costs, nor does it expect tax cuts in other areas; instead, it proposes the creation of a £1 billion per annum ‘Children’s Future Fund’ to provide ‘education and skills’ (ibid.). The costs of healthcare would therefore be distributed between fat and thin as before.

But there is an even more fundamental problem with obesity-related taxes. As van Baal and others have shown, the lifetime healthcare costs of obese people are actually lower than average due to their shorter lifespans (van Baal et al. 2008, Grootjans-van Kampen 2014). Add to this the financial savings to the state that come from a reduced pensions bill as a result of premature mortality and it becomes clear that obesity is not a drain on government resources. On the contrary, it is probably cost-saving. This, of course, is an economic argument, not a moral one, but it illustrates the hazards of presenting moral arguments in economic terms.

Transferring regulatory power to quangos

7. ‘Remove responsibility for nutrition from the Department of Health and return it back to an independent agency’

This policy illustrates another element of the modern campaigner's mindset - a desire to distance decision-making over things like lifestyle choices from democratic control. This technocratic, anti-democratic tendency has been rife in political circles for some time. When the Conservative-Liberal Democrat coalition government came to office in 2010, it promised a 'bonfire of the quangos' (the flame soon went out). While campaigners have often found willing listeners in Whitehall's civil service and even among ministers, elected politicians still have to answer to voters. Taking power away from departments and giving it to quangos means that accountability is no longer there.

More ambitious public health campaigners have called for powers to be pushed even further away from the hands of democratic governments. In November 2014, several hundred health groups and individuals wrote a letter to the World Health Organisation - which is itself an unaccountable NGO - requesting a binding international treaty to tackle obesity (Long et al. 2014). This proposed ‘Global Convention to Protect and Promote Healthy Diets’ included many of the anti-market policies discussed above, including ‘restrictions on marketing to children’, ‘compositional limits on the saturated fat, added sugar and sodium content of food’, ‘removal of artificial trans fats’,  ‘restaurant calorie labelling’, ‘fiscal measures and financial incentives’ and ‘public health impact assessments in trade and investment policies’. Since national governments are capable of introducing any of these measures of their own volition, a global treaty would serve no other purpose than to impose the views of the Western public health elite on politicians now and in the future.