Good afternoon, Stephen, and welcome to Commodity Conversations. Could you please tell me a little about yourself, Czarnikow, and CZ App?
I’ve been involved in sugar since joining Czarnikow in 2008. At that time, we operated as a traditional sugar trading house, buying FOB and selling C&F. I was fortunate enough to join a highly regarded sugar analysis team. We had Toby Cohen, who led the team and is now at ASR, and Peter de Klerk, who is now at the ISO. It was a great place to learn.
Over time, Czarnikow has evolved. We have de-risked by moving away from the traditional trade house model. We no longer buy FOB to sell C&F but instead look to build long-term relationships with participants throughout the supply chain. For example, we offer services such as vendor-managed inventory, financing support and procurement assistance, working in partnership with many of our customers. The other significant change over the last 5 to 7 years is that we no longer operate solely in the sugar industry. As a business, we now provide a range of supply chain services across a broad spectrum of goods which are used by the food and beverage sector.
CZ’s analysis team has evolved, too. We strive to be as open as possible, sharing our thoughts online. We make data and analysis available for free on our platform, CZ App. We also offer subscriptions to our sugar analysis to help customers manage their price risk.
Let’s get to the heart of the issue: How do you measure sugar consumption?
It’s the most challenging job we face as sugar analysts. Even on an individual level, people often don’t know how much sugar they consume. If you ask me how much sugar I ate yesterday, I couldn’t tell you. I have no idea how much sugar I consume, and I’m in control of what I eat. This is true for everybody.
It’s partly because sugar appears in foods you wouldn’t expect, but also partly because people are poor at recalling what they eat.
And then, in some cases, when extensive scientific studies ask people to keep food diaries, individuals often under-report. Perhaps people sometimes feel embarrassed about what they’ve eaten over the course of a week.
The problem worsens as you gather more data, which is why a top-down approach is also challenging. You can often see how much sugar a country has produced and how much has been imported or exported, but you can’t really see how much is in a pipeline. Is it a large pipeline with a lot of stock, or a smaller one?
You can’t see how much is being spoiled or lost, or if the sugar quality is deteriorating, from a top-down perspective.
In some countries, you can be fairly accurate. For the UK, consumption is about 1.7 million tonnes, which might be accurate within roughly 100,000 tonnes, but there is no absolute precision in that figure. This makes it challenging to monitor changes in consumption in real-time.
You usually only realise that a change has occurred in consumption after a few years, when either your models break, or the sugar price behaves unexpectedly.
Some statisticians use sugar availability as a proxy for consumption. They consider production, imports, and exports to estimate consumption. However, you don’t know how much sugar is wasted or discarded in yoghurts, bread, or pastries, and similar products.
At CZ, we’ve always made the choice to assume that sugar has been consumed when it’s transformed into a finished good, for example, when it is added to a chocolate bar. This means we don’t need to track finished products across borders. However, it’s clearly not the case that the sugar is then consumed, because food is often thrown away.
People say that 20 to 30% of all food is wasted, with half of this happening at the retail or home level. Do you consider this when estimating consumption, or do you think, well, it’s gone, so it doesn’t matter if it’s wasted or consumed?
The latter. We say, okay, it’s gone. It’s been turned into something which can be eaten, so whether it is eaten or thrown away, that sugar has disappeared. Again, this is for simplicity. I don’t want to spend time calculating food waste rates for different countries.
Is global sugar consumption increasing, and, if so, at what rate?
When I started in 2008, Peter de Klerk told me that there was a rule of thumb that sugar consumption increased by 2% annually, with 1% attributed to population growth and 1% to wealth growth. If sugar prices became extremely high, or sugar becomes unaffordable, or the world goes into recession, then you can reduce that to one per cent.
In years when the sugar price is very low or economic growth is very high, you can raise it to 3%.
That rule of thumb held up well until about 2014, remaining effective throughout the high sugar prices of 2010 and 2011.
From 2014, something changed, coinciding with the introduction of the Mexican sugar tax. It received significant publicity, and it was around that time that greater attention was paid to whether sugar or fat was the dietary villain. Until the early 2010s, fat had been the focus, and everything was low-fat. Then, in the mid-2010s, the emphasis shifted, and sugar began to receive increased scrutiny.
Our statistics show that, from 2014 onwards, global sugar consumption has remained stable on a per-person basis. At best, total consumption has been growing at a rate of 1% per year, in line with population growth; at worst, when prices have been high or during COVID lockdowns, it has been below that rate.
I imagine there are many reasons for that. Are sugar taxes one of them?
Sugar taxes, by themselves, have not caused a significant shift in consumer behaviour.
When sugar taxes are introduced, the prices of soda and chocolate bars increase. It has a short-term effect on consumer spending, but within a few months, consumers become accustomed to the new price, and their behaviour returns to its previous state. If there’s any effect from the tax on the demand for sugar, it’s short-lived.
Taxes influence consumption when manufacturers reformulate their products to avoid paying the tax. The UK implemented a tiered sugar tax based on the sugar content in a soft drink. Most manufacturers reformulated their soft drinks to reduce sugar and substitute it with alternative sweeteners, leading to around 100,000 tonnes less sugar consumption. The tax led to lower consumption, but not through changes in consumer behaviour.
The media has had an impact, as consumers are now better equipped to understand product labels and the contents of different products than they were 10 or 15 years ago.
In 2010, if I had stopped a consumer in most advanced economies and told them that there was sugar in ketchup or in some bread, most would have been quite surprised. Now, if I stop a consumer in most countries and say, ‘Did you know there’s sugar in ketchup?’ they would probably roll their eyes and say, ‘Yes.’
There has been a rise in awareness of the sugar content in various foods, prompted by media coverage and taxation. It has influenced behaviour.
In 2014, Czarnikow published a report indicating that the per capita consumption in the UK peaked in 1957 at 53 kilograms per person, declined to 48.5 kilograms in 1970, and further dropped to 35 kilograms in 2014. Do you still endorse those figures, and what has changed since then?
I believe those figures are generally accurate. We have been involved in the UK sugar market for over 160 years, which has allowed us to gather extensive data on the country’s sugar industry. Sugar consumption in the 1950s, 1960s, and 1970s was high, mainly through sugar in tea, home-baked cakes, and biscuits, rather than in processed foods. Today, sugar is primarily consumed through processed products.
The decline has persisted since 2014. The UK population is approximately 70 million, with an annual sugar consumption of around 1.7 million tonnes. This results in a consumption of roughly 25 kilograms per person.
In countries such as the United Kingdom and Australia, sugar intake has been falling for several decades and is approaching 25 kilograms per person. Consumption in developing economies is increasing rapidly and should help offset this decline. They are approaching 25 kilos per person from the other side. For example, in India, sugar consumption is rising to the mid-20s.
In general terms, the eight billion people on this planet consume about 180 million tonnes of sugar. This amounts to 22.5 kilos per person annually.
There is still potential for China, India, Indonesia, and Pakistan to increase their consumption, but if you had asked me 20 years ago, I would have said they could reach 30 or 35 kilos on average; now, I believe 25 is a more realistic estimate.
Per-person consumption is declining in Western countries. Is the EU an exception?
The EU27 has a population of 450 million people and a consumption of 15.5 million tonnes. It works out at 34 kilos per person. We may be seeing a decline in Northwest Europe that is being offset by Eastern Europe. Or the figure may reflect the high number of processed products made in the EU, which are then exported. For our statistics, that would count as European consumption, even if those products are consumed elsewhere. However, it is not clear.
What about the US, where you have sugar and HFCS? Is total per-person sweetener demand declining?
According to the US Department of Agriculture, yes. They believe that per capita consumption of caloric sweeteners, such as sugar, HFCS, and other syrups, peaked in 1999 at 157 pounds per year and has since declined to 121 pounds in 2024. Interestingly, through the 1990s, sugar consumption was stable, and the growth in sweetener consumption was primarily due to increased HFCS use. Since 2000, the use of HFCS has declined, but sugar consumption has shown a slight recovery. So American consumers are eating less sweetener, but more of that sweetness comes from sugar. President Trump’s recent announcement that Coca-Cola could use more sugar and less HFCS may continue this trend.
In my Sugar Casino book, I stated that the Brazilian per capita consumption was 62.5 kilos, the highest in the world. Was that figure accurate, and if so, why is it so high?
We currently estimate it at 46 kilos per person, down from 60-65 kilos 20-25 years ago. Brazil has always been among the highest per capita consumers.
This issue highlights our inability to accurately monitor consumption. I find it hard to believe, but that is what the statistics indicate. Other analysts have similar figures.
Per-person sugar consumption continues to decline in the developed world, but could it be that wealthier individuals are eating less sugar and exercising more, while less affluent people are consuming more processed foods, exercising less, and living in food deserts where local shops only stock soda and snacks? Is the average per capita consumption figure hiding significant variation across the population?
It’s possible. I haven’t seen any statistics that break it down by income or similar metrics. However, I can imagine that if you have a higher income and more free time, it’s easier to prepare meals from scratch using ingredients. Therefore, you’re more in control of your sugar intake compared to consumers of processed foods.
Has the slowdown in consumption growth impacted production growth? Are banks and international agencies less inclined to finance sugar production facilities than they were 20 years ago?
Not to my knowledge.
However, global sugar production has remained relatively stable for nearly 15 years. We have been producing about 175 million tonnes of sugar, give or take 15 million tonnes each year. The highest was 189 million in the 2017/18 crop year, and the lowest was 160 million. During this period, fluctuations have occurred in yields, acreage, and other factors; however, the overall trend has remained relatively consistent. Production has increased in Brazil, while it has decreased elsewhere.
Meanwhile, the world’s population has grown by a billion people, and global trade has expanded.
To what extent is ethanol responsible for that?
It has made a difference in some countries. India, for example, can now redirect 5 or 6 million tonnes of sucrose into ethanol production instead of sugar. However, Brazilian cane production has remained steady at around 600 million tonnes for a decade.
You might also argue that having 175 million tonnes of sugar production 15 years ago was simply too high. We faced overcapacity, and it took 15 years to eliminate it.
GLP-1 drugs appear to not only suppress appetite but also shift dietary preferences towards high-fibre, low-sugar foods. How will they influence sugar consumption?
GLP-1 drugs diminish sensations of hunger and curb impulsive human behaviour. One study showed that people on Ozempic were 30% less likely to become addicted to nicotine. Another found that individuals on GLP-1 drugs have a lower risk of developing alcohol use disorder. From a sugar perspective, this means people binge eat less and prefer foods that are lower in calories.
It doesn’t seem to work only through metabolism. It also appears that these drugs involve the brain’s reward centres.
Weight loss can have a puritanical streak. Some argue you’re approaching weight loss incorrectly if you depend on GLP-1 drugs. They contend that one should eat less and exercise more, rather than relying solely on medication.
However, for at least the last 40 years, the standard advice for overweight people has been to eat less and exercise more. It is good advice, but just because it is good advice doesn’t make it easy to follow. Obesity has increased significantly over time. Yet, we now finally have a drug that seems to overcome the willpower barrier.
With patents expiring in some countries, it might be more economical for health authorities to provide free GLP-1 drugs to obese individuals rather than covering the costs associated with obesity-related health complications. I also observed that one company is developing a pill instead of an injection. I expect people will favour pills over injections, which could boost their usage.
We will probably reach a stage where the drugs become significantly cheaper, and you will be able to take them in tablet form.
The UK’s National Health Service estimates that obesity costs £11.4 billion each year. However, determining the exact figure is challenging due to related health conditions. Being obese doesn’t necessarily mean you will become ill, but it does raise the risk of developing type 2 diabetes, heart failure, joint problems, and other health issues.
There are an estimated 10 million obese adults in the UK, so providing these drugs free of charge would likely become more cost-effective as GLP-1 prices fall. I wonder if any government will implement this, but the cost-benefit analysis suggests they should. It’s worth considering, at least, whether patents should be bought out and these drugs made widely accessible.
Have you lowered your sugar consumption forecasts because of GLP-1 drugs?
We’ve made a small allowance for some countries, but there are still many aspects of GLP-1 drugs that we remain unaware of.
We estimate that around 5% of Americans are on GLP-1 drugs at any given time. However, this figure is fluid, as people start and stop them. We also know that the US is the world’s largest drug market, making up about 50% of global drug sales.
However, obesity is not just a problem for affluent countries. There are other countries, such as Malaysia, for example, which have high rates of obesity and type 2 diabetes.
Predicting the adoption of these drugs is difficult. Will we see a future where 10% of most populations are on them at any point in time? It’s possible. However, it’s equally likely that they are not widely used for some reason.
I haven’t seen any solid statistics indicating that when someone is on a GLP-1 drug, they consume a certain percentage less sugar. That research doesn’t exist. All we know now is that there will be an increased uptake of GLP-1 drugs in the future, and they might have some negative impact on sugar consumption; that’s about it.
However, we can’t ignore it. We must make some allowance. But all I know is that whatever allowance we make will be wrong. So, it’s just an indication now, a guide to what we think is happening.
We have reduced our US sugar consumption figures by 1% in 2024 and by 2.5% in 2025.
We will adopt a similar approach for Western European countries and Australia from 2025. If GLP-1 generics are produced in India and China at considerably lower costs within the next five years, we can then begin implementing reductions in consumption more widely across the globe.
A final consideration is that there are individuals who are superusers, highly dependent on processed foods. For example, 20% of consumers account for around 65% of American ice cream and chocolate sales. If you consume a lot of ice cream and chocolate, you may be overweight and more likely to be on GLP-1 drugs. If five per cent of the US population takes a GLP-1 drug, they could represent more than 5% of US sugar consumption. It’s not a direct one-to-one relationship. Even with a relatively small number of GLP-1 users, their impact on overall consumption can still be substantial.
Thank you, Stephen, for your time and input.
© Commodity Conversations® 2025