You wouldn’t trust an economist to give you a smear test. So is it reasonable to expect those working in health to grasp economics? But still we listen to sugar tax proponents who don’t understand how consumer taxes work, says the NZ Initiative’s Jenesa Jeram.
This is the second in a two-part series presenting both sides of the sugar tax debate. The first, by the Green Party’s Julie Anne Genter, is here.
New Zealand, I don’t know how to tell you this, but we need to lose some weight. And to make matters worse, we’re probably only going to get fatter.
A lot of credible people think a sugar tax will help address this problem. These days, you can hardly read any article on obesity without a public health expert touting a sugar tax as a solution.
Those working on the ground deal with the tragic consequences of obesity and diabetes every day. They are experts in their field. But that doesn’t make them experts in applying economic mechanisms to policy problems.
Despite a growing number of voices in favour of a sugar tax, and an increasing number of countries introducing one, the evidence that a sugar tax will work as intended is unconvincing.
There is still no real world evidence that a sugar tax will reduce obesity, and there are reasons to be doubtful it can.
The case for a sugar tax is a bit like candyfloss. It looks pretty attractive and solid, but it has very little substance. So let’s look at the arguments for a sugar tax:
Of course public health experts understand economics: when the price of a good goes up, people buy less of it
You don’t need a PhD in economics to know that when the price of a good goes up, people buy less of it. No one is denying that general principle.
But economics will also tell us that the magnitude of the tax matters: small increases in price might have negligible effects on purchasing behaviour. For a sugar tax to have the same effect as tobacco taxes, the magnitude would have to be much greater than the 20% tax that is most advocated in New Zealand. The reason why sugar taxes of a 110% magnitude (like tobacco) aren’t advocated in New Zealand is that they would be publicly unpalatable. Yet if that is public health’s end game, they should at least be honest about it from the start.
People also respond differently to the tax based on their ‘demand elasticity’: how responsive they are to price changes depending on their attitude towards that good.
Most importantly, total caloric intake matters: switching from soda to frappuccinos is not a policy victory.
If all this sounds kind of obvious, that’s because it is. I’m not claiming any kind of special knowledge here. Yet they are factors that many public health studies miss.
Some studies used as evidence claim a sugar tax victory if consumers spend less on the product in response to a tax. But even if consumers spend less on a sugary product, they might simply be switching to cheaper brands and keeping their consumption of those products the same.
This is a problem the University of Waikato economist John Gibson has been grappling with in his Marsden-funded work: to what extent purchase choices reflect both the quantity and quality of products. Different food groups (for example, sugar-sweetened beverages) have different brands available, different package sizes and different formulations. Most studies don’t recognise that spending decisions can affect both the quantity and quality purchased. What this means is that you can spend less on soda while consuming more by switching to cheaper brands or different sizes, yet these studies would still claim that as a soda tax win.
After adjusting for this, Gibson finds that “food and drink quantity demand is likely to be much less (about an eighth less) price-responsive than is suggested by many studies relied upon by advocates for SSB and other health-related taxes.”
Switching to untaxed, but equally – or even more – unhealthy goods is also a behaviour that some studies do not fully account for. Some studies try and assume what products can act as substitutes for the taxed products, many of them miss the mark. For example, common substitutes for soda are often assumed to be diet soda, milk, water and juice (in other words, other beverages). Those studies would not assume that people drinking soda for a sugar hit might switch to a candy bar, or if they are drinking soda after a hard day’s work, they might switch to beer.
Any effective tax has to be broad enough to cover all possible substitutes, and of a large enough magnitude to guarantee people will be responsive. But the broader the tax (for example a comprehensive junk food tax or per-gram-of-sugar tax), the more complications policymakers run into in defining what should and shouldn’t be covered.
But some public health studies are based on economic modelling. Economists like models. What is there to dispute?
I’ve said before that there is no real world evidence that a sugar tax will reduce obesity. That is partly because a sugar tax is a fairly new concept, the evidence pool is shallow because it is too early to be drawing conclusions from overseas. In this context, economic modelling can act as second-best evidence.
But economic models are only as strong and reliable as the assumptions they make. As noted above, assumptions about substitution to untaxed but equally unhealthy foods, or downshifting to cheaper brands or container sizes will greatly skew final conclusions.
More substantively, we’re all metabolically different, meaning our bodies process foods in different ways due to genetic and lifestyle differences. A marathon runner will have different caloric needs compared to those of us with desk jobs. That’s why real world effects matter. Many of the studies used as evidence for a sugar tax have not adjusted for differences in how bodies react to changes in diet and exercise. Assuming a tax is desirable, some of us will be under-taxed and others will be over-taxed. In a tongue-in-cheek opinion piece, Catherine Rampell has suggested that because of these differences, it might be more efficient to tax obesity itself, rather than taxing inputs. For those who find such a suggestion repulsive because it is regressive, fat-shaming or unfair to punish people for their genes, then a sugar tax would do exactly those things.
But even the World Health Organisation thinks a sugar tax will work
You can’t win a policy debate without slaughtering a few sacred cows. And when it comes to public health, the WHO remains the lead cow.
It probably takes a brave or stupid person to take on the health expertise of the WHO. Perhaps the WHO is banking on this, because its latest report contained less meat than a petrol station mince pie.
That wouldn’t matter so much if the report were simply a summary of the sugar tax landscape so far. But to be touted as ‘evidence’ is plain dishonest when public health groups all over the world are using the report to justify sugar taxes in their own countries.
Many of the bold claims of the report are uncited, or the WHO adopts Inception-like referencing by citing the draft version of its own paper, which appears unavailable to the public. Now, citing your own prior work is fine, but citing a publicly unavailable working draft of the same paper is not so fine. An unavailable draft version of this op-ed proves how bad this is – but don’t ask me for a copy.
The WHO then point to other countries who have tried a sugar tax or other food taxes. While the number of governments implementing a food tax is growing, the WHO acknowledges that many countries focus on the economic benefits rather than the health benefits. But the WHO does not go on to talk about the health outcomes these countries have achieved, mainly because it doesn’t know.
Many of the countries the WHO cite have yet to undergo formal evaluations, while in other countries the policy has since been abolished. Other countries have shown reductions in purchases or reductions in the consumption of goods that are taxed, but as I’ve argued, that’s not the same as reducing obesity.
Other countries are doing it, New Zealand is behind the eight ball
While we talk about sugar taxes as a general concept, these taxes have been applied in so many different ways in different countries that the term is beginning to lose meaning.
Even if there were a growing pool of evidence that taxes are improving health outcomes overseas (there isn’t yet), policy analysts still need to think about whether different countries’ approaches will work for New Zealand.
Nevertheless, let’s talk about Mexico as it is considered a raging success by many New Zealand sugar tax advocates. The most recent study, from the British Medical Journal (2016) showed only a tiny effect from a sugar tax. The largest decrease was among the poorest households, and was the equivalent of one sugar cube, per person, per day. While it is true a decrease is still a decrease, the magnitude of the effect needs to be balanced against the total sugar that population will be consuming, and whether it will make a dent in obesity.
Even then, there is reason to doubt the validity of the Mexico study’s findings given problems with the econometric methods (for a technical takedown of the study, see Eric Crampton’s analysis here).
Even if doesn’t reduce obesity, can it save kids from traumatising dental surgery?
Like many of you, I have found the recent stories about children needing serious dental surgery because of fizzy drinks disturbing and disheartening. But will a sugar tax really fix things? Will an incremental increase in the price of sugary drinks be enough to change parents’ minds about putting soda in sippy bottles, rather than education and appeals about their own child’s wellbeing? It is highly doubtful that price is the primary factor driving such decisions.
I’m not trying to parent-shame here, but think a bit about the motivations behind such behaviour. In a country where tap water is free in most places (I acknowledge there might be some problems with potable water in some rural areas), a simple price increase does not seem like a convincing way to tackle this kind of problem.
Aren’t economists all about addressing externalities? Obesity is costing the taxpayer!
Just as economists shouldn’t perform smear tests, the term ‘externalities’ gets butchered a lot by people who aren’t economists (actually, it gets butchered by some economists too). For a primer on what is and isn’t the kind of externality that matters for policy, check out The New Zealand Initiative’s report). The short story is that ‘costs to taxpayers’ isn’t the kind of externality that causes real problems on its own. It is akin to worrying that the neighbours’ having a kid imposes an externality on you, who will have to pay for their education through the tax system?
But on a broader note, the ‘costs to taxpayers’ argument doesn’t really stack up anyway, unless we are considering the net costs. That is, the costs of a disease combined with any fiscal savings. The Institute of Economic Affairs in the United Kingdom recently calculated that the net costs of obesity are much, much lower than the direct medical costs of disease because of fiscal savings in foregone superannuation.
Now, not everyone will feel comfortable talking about the fiscal ‘savings’ incurred from disease. It can seem pretty tasteless. But exaggerating the burden obesity places on the public purse compared to those who live long and healthy is surely pretty tasteless in itself.
Even if there’s no current evidence, surely there’s no harm in trying, right?
Despite the lack of evidence for a sugar tax, there are still some sugar tax advocates who will argue there is no harm in trying. Policy experimentation is the only way we will know for sure whether the policy is effective.
But here’s the rub: there is harm in trying.
Even sugar tax advocates will acknowledge that those who consume the most sugar goods, and are most likely obese, are also disproportionately poor. But they argue the tax will not be regressive because the poor are most likely to be responsive to price increases, and any revenue gathered from the tax could be redistributed to health and social programmes to benefit poorer communities.
Once again, great liberties are taken with the economic term ‘regressive’. There’s no way of getting around it, a regressive tax is one which takes a larger proportion from lower income households than higher income households. Arguing that the poor will be more responsive to a tax assumes that these people really are willing to change behaviour. If the problem is that poorer people cannot afford healthy food, then surely a more targeted policy response would be raising their incomes to ensure they can? If policymakers believe increases in income won’t affect purchasing decisions, then the problem is not of price alone.
A sugar tax is also poorly targeted as not all poor people are obese. For those who cannot afford many material luxuries, food can become an affordable pleasure. In busy households facing multiple stresses, some families may purchase unhealthy but convenient foods because healthy food preparation is just one pressure too much.
Imposing a regressive tax on poorer people ‘for their own good’ is only likely to be beneficial if the targeted population see it that way, otherwise it is deeply punitive.
Another way advocates try and get around the ‘regressive’ accusation is by arguing that tax revenue can be redistributed into health-promoting programmes. Economists call this a hypothecated tax, and it makes very little sense in the area of public health. If the purpose of the tax is to discourage behaviour, then the tax will not raise a lot of money for these programmes. If these health-promoting programmes are as valuable and laudable as experts claim they are, then why not make the case for funding them out of the existing government budget? While hypothecated taxes might be more appealing to politicians and voters, they make very little sense from the perspective of improving health outcomes.
While a sugar tax might not be effective on its own, it is part of a suite of measures to improve health outcomes
This argument, frankly, avoids all policy accountability. If a sugar tax is indeed introduced as part of a suite of measures, then it is even harder to attribute effectiveness even if obesity were to decrease. Some would say that attributing effectiveness doesn’t matter, as long as obesity has decreased then their job is done.
But as above, there is harm in trying. Making poor people poorer is part of that. But sugar taxes are also an incredibly blunt instrument: they will affect the entire population, obese or not obese. The current government has already acknowledged that it is spending billions of taxpayers’ dollars without knowing what works. Imagine taxpayers giving more money to government under the guise that it will be helping to improve outcomes, without any reassurance that their contribution is doing so. Measuring outcomes of individual policies matters.
So you’re saying do nothing?
Of course, a rejection of sugar taxes is not a reason to dismiss concerns about obesity, diabetes and children undergoing traumatic dental surgery. For those who are concerned about these things though, there is an even greater need to get the evidence right. Otherwise, sugar taxes could be a red herring – a distraction from finding more targeted and effective policy solutions. That’s at best. At worst, they can actually cause harm to those populations a sugar tax is meant to help.
Jenesa Jeram is a policy analyst at The New Zealand Initiative. Her previous report ‘The Health of the State’ has looked at lifestyle regulations and the public health studies that form the evidence base.
To see the companies and organisations who make up the NZ Initiative’s membership base, click here.
For the other side of the argument, read Julie Anne Genter’s post advocating for a sugar tax here.
The Society section is sponsored by AUT. As a contemporary university we’re focused on providing exceptional learning experiences, developing impactful research and forging strong industry partnerships. Start your university journey with us today.
The Bulletin is The Spinoff’s acclaimed daily digest of New Zealand’s most important stories, delivered directly to your inbox each morning.