A Spoonful of Sugar: What Postwar Rationing Can (and Can’t) Tell Health Policy

While she was pregnant with our son in 2018, my wife and I agreed to a no-sugar diet. We also draped her in a lead vest and a blanket that blocked low-frequency electromagnetic fields (basically a Faraday cage) whenever we flew. She wore a mask on the plane, in case row 32 was harboring any viruses. My wife tolerated it well. This wasn’t evidence-based by any formal definition; we were following Pascal’s Wager. Namely, take the precaution if the cost is low and the potential harm lasts a lifetime; don’t wait for the study.

Fast forward eight years. The balance sheet is mixed. The Faraday blanket was dumb back then and still is. The mask looks a lot smarter than it did in 2018. And the sugar bet was validated (mostly) by a paper in Science in 2024. Tadeja Gracner, Claire Boone, and Paul Gertler noticed that Britain rationed sugar from January 1940 to September 1953. When they stopped rationing, the British did what you would think they would after 13 years without sugar: they binged. The authors used government food surveys to show that purchases of sugar per adult doubled from about 40 grams per day to about 80 grams per day over the course of a few months. They compared individuals in the UK Biobank born before the end of the rationing period with those born after and found a third less diabetes and a fifth less hypertension in the former group. The effects began prenatally and increased at 6 months when they began consuming other foods. The effects were greatest for those that were rationed to age two.

Since then, more than a dozen additional publications using the same cutoff point have shown reductions in cardiovascular disease, multiple types of cancer, dementia, rates of aging and death, and anxiety and depression.^1-3 I take issue with some of the numbers. Toddlers also consumed less butter and bacon during the ration period, and some findings will likely lose significance as more studies are conducted. However, these findings align with what we know about sugar, and we can litigate methodological quibbles elsewhere. As a policy question, I’m curious about what the intervention actually was.

Nobody was instructed to eat less sugar during the ration. There were no pamphlets given to the mothers of Great Britain about the putative health benefits of lowering their infants’ sugar intake. They simply gave out less sugar. Under rationing, people consumed about as much sugar as we currently recommend they consume (around 40 grams of added sugar per day or slightly less than 10% of their caloric intake). Moreover, everyone consumed about the same amount regardless of class.⁴ That was the policy that decreased diabetes by a third. Today, nobody would intentionally implement such a stringent policy in a democratic society. But it is important to understand what was being measured in these papers. Because we currently have the exact opposite policy in place.

Our current policy is recommendation. We recommend that children under two do not consume added sugars according to WHO guidance and the Dietary Guidelines for Americans. These are good recommendations backed by evidence (e.g., on dental caries, early childhood obesity, and taste formation). But they rely on the individual to make the same Pascal’s bet I described earlier. And, as we know, Pascal’s bet has a problem. There are too many gods. If you bet on everything that might benefit your child at a low price, you end up putting a lead vest on your child when you fly. Only some people can do this, and they tend to have money to spend. I had money to spend, time to spend, and I was crazy enough to bet on it all. The average person does not have any of these things. So, most people won’t follow the dietary guidelines. That means most of the sugar your two-year-old consumes comes from the grocery store. Pouches, toddler snacks, yogurt, cereal, and so on. Rationing is crude, but it bets for you by what it makes available.

Which brings us to policy options. The best analogy we can draw to the ration book is creating a product standard for infant/toddler foods that restricts the amount of added sugar. WHO has recommended this in the form of a nutrient profile for infant/toddler foods. There has been discussion in the UK about restricting sugar in commercial infant foods. We currently have no such restriction in place in the United States of which I am aware. Another lever is to regulate these foods in the Women, Infants, and Children (WIC) program – the closest thing we have to a book of ration coupons in this country. WIC currently feeds a substantial portion of our infants and prescribes what foods are included in those food packages. We could update these guidelines to reflect the dietary recommendations for children under two. Lastly, we could look to restrict marketing of these products to children. There should be some rules about whether you can call your product a “fruit snack,” for example.

We have one instance of this happening in the United States. Ten states implemented a SNAP restriction on soda in 2026, and some also restricted candy. Three papers have examined grocery receipt data from these states.^5-7 All find similar patterns, though with limitations. Households that received SNAP decreased their purchases of excluded beverages by 11-13%. They also decreased total sugar/calories purchased by approximately 5-8%.⁷ These decreases were larger than what we would predict based on standard models of consumer behavior. Many people who could have purchased soda with cash did not. If states restricted some sodas but not others, households shifted their purchases to unrestricted beverages. So, a selective ban works less well. But we know that restricting access works, especially when you restrict all sources. Perceived stigma also increased when people used SNAP to buy food.⁶ By contrast, rationing in the UK was universal, so people didn’t have to worry about the perception of using a government benefit to buy soda. We could accomplish this by mandating a product standard for infant and toddler foodstuff.

Moreover, to some extent, the wartime ration evened out the diets of all classes in Britain. This was achieved by giving everyone the same ration, plus free milk, orange juice, and cod liver oil for children under five. As a result, Britain emerged from the war with taller/heavier children and decreased infant mortality.⁸ Edinburgh reported an infant mortality rate of 68/1000 in 1940 and 50/1000 in 1945.⁹ It wasn’t completely equal. Restaurants weren’t rationed, and many people with the economic means found ways to get more food.¹⁰ But the policy leaned heavily towards equality. Our current recommendations-based policy does the opposite. Those with the time and education to follow the guidelines will likely already be healthy. If the sugar dial is set in the first thousand days, then today’s class gap in what toddlers eat is tomorrow’s class gap in diabetes. A standard on the shelf is the only one of these tools that narrows it.

These results also say something about what we can intervene on. The authors of the cancer paper found that people rationed early in life continued to consume less sugar at age 50. In another paper, van den Berg et al. found that there was a short period of time in 1949 when sweets were removed from the ration while most other items, including sugar, remained on it.  Children who were exposed prenatally to this period consumed less sugar and had a slightly lower body mass index as adults.  Those with genes that make them like sugar more showed the greatest drop in consumption. Both papers highlight that we can influence how much sugar a person eats later in life by regulating how much sugar children are exposed to early in life.

Pregnancy was one of the bets I placed in 2018. But our present guidance for the under-twos says nothing to a pregnant woman. Prenatal care screens for gestational diabetes and then, for the large majority who pass, says little about sugar at all. If a third of the diabetes effect is set before birth,¹ then the cheapest intervention in this whole literature may be a sentence during prenatal visits and a change to the WIC package for pregnant women. Neither needs a randomized trial. We just need to decide that what a pregnant woman eats is a public health matter.

I want to caveat all of this. The effects we are seeing are based on a natural experiment that occurred over 70 years ago. I have put myself on a no-added-sugar diet recently, which is probably about 55 years too late. But, hey, at least I know my bet. Pascal again.

None of this requires a ration book. Just some regulations on added sugar in infant/toddler foods. We can update WIC, which is already a prescriptive policy, to include only no-sugar-added foods. Ditto for SNAP, at least on soda. Meanwhile, we can fund research to see whether these effects replicate elsewhere while we wait for this policy “experiment” to play out over the next 57 years. I may have missed the window, but it’s not too late for the next generation of Americans.

References

About the Author