Does One More Health Star Really Make a Food Healthier?

Geoffrey Chen

When Numbers Start Making Decisions · Season Two, “When the Measure Becomes the Target” · Article 9

1. Two cereal boxes on the same shelf

A shopper picks up two breakfast cereals. One displays four Health Stars on the front and the other three. She does not have time to compare sugar, sodium, fibre and ingredients per 100 grams, so she chooses the four-star product. This is the decision Australia and New Zealand’s Health Star Rating system is designed to support: a quick comparison of the nutritional profiles of similar packaged foods.

Suppose a manufacturer raises a cereal from three stars to four by reducing sugar and sodium and adding fibre. The formulation and the rating have both improved. Now suppose it adds a score-positive protein or fibre ingredient while leaving a high degree of processing and other unmeasured features largely unchanged. The star can still rise. Whether the food has become healthier in the context of a whole diet needs a more qualified answer.

A Health Star is not an essential property of food. It is a nutritional proxy designed for comparison at the point of purchase. Official guidance says to compare like products—for example one yoghurt with another, not yoghurt with potato chips. It also warns that the algorithm cannot account for every ingredient, processing method, vitamin and health effect and does not replace dietary guidance or professional advice. Health Star Rating: “How ratings are calculated”

The answer is therefore not a simple yes or no. An additional star generally indicates a more favourable profile on the characteristics measured by the system. It does not prove that portion size, dietary role, processing and individual need are all healthier.

2. How are the stars calculated?

Health Star Ratings run from half a star to five stars. For most products, the calculation uses values per 100 grams or 100 millilitres. It begins with baseline points for energy, saturated fat, sodium and total sugar, then applies modifying points associated with protein, fibre and the proportion of fruit, vegetables, nuts and legumes. Rules differ among food categories before the final score is converted into a star rating.

The method addresses a real information problem. The Nutrition Information Panel contains multiple values, and shoppers must interpret units, serving sizes and interactions among them. A front-of-pack star compresses that work into a visible relative judgement. The algorithm is public and supported by calculators and implementation guidance, so manufacturers, researchers and members of the public can inspect the rules instead of accepting a purely commercial black box.

Compression necessarily contains institutional choices. Why use 100 grams rather than an ordinary serving? When should favourable components offset unfavourable ones? Should sugar be total or added sugar? Do dairy, beverages and oils need different categories? These disputes do not arise after a neutral calculation. They are the conditions that make calculation possible.

Each weight and category forces foods, consumption patterns and different bodies into a limited format in exchange for comparability at scale. A responsible system does not pretend this compression is costless. It makes the scope visible and continually checks for anomalous results.

3. When the star enters the formulation meeting

The rating affects more than shoppers. It can shape product design. A higher star displayed on the front has marketing value, and even a manufacturer that does not display it can use the algorithm internally. A development team can simulate whether reducing sodium, sugar or saturated fat, or adding fibre, will move a product across the next boundary.

This can be the system’s ideal feedback loop. A company improves its score by reformulating in the direction of public-health objectives. The Five Year Review found that the system had increased awareness and use, encouraged positive reformulation and was broadly aligned with dietary guidance. Health Star Rating System: “Five Year Review Report”

The same loop can narrow attention to variables that yield points. The operative question becomes not “how does this product fit a healthy diet?” but “what is the lowest-cost way to reach four stars?” If adding a favourable component is easier than removing an unfavourable one, a producer may choose the former. A label may also display a small serving while the algorithm standardises by 100 grams, leaving actual eating patterns to a separate judgement.

Not all algorithm-directed reformulation is gaming. Reducing sodium is exactly the behaviour the system seeks. The appropriate test is whether the change improves the public-health condition represented by the measure or exploits compensation among proxies without a comparable reduction in nutritional risk.

4. Voluntary display creates a selection problem

As at August 2026, displaying Health Star Ratings remains voluntary in Australia and New Zealand. A manufacturer calculates the result and chooses whether to show it on an eligible product. Voluntary adoption reduced initial regulatory cost and allowed the system to develop incrementally.

It also creates selective visibility. High-rated products have a stronger incentive to display. Low-rated products can omit the label. A shopper seeing “five stars” and “no stars” cannot assume the unlabelled product has a low score; it may simply come from a business that does not participate. A system built to compare like with like works best only when like products supply comparable information.

Food Standards Australia New Zealand began Proposal P1067 in 2026 because the industry had not met a target of displaying the rating on 70 per cent of intended products by November 2025. FSANZ says consumer awareness and use are high, but inconsistent adoption limits effectiveness and affects trust and understanding. It is evaluating whether the system should become mandatory. FSANZ: “P1067—Health Star Rating System”

The power of a measure depends not only on the number of stars but on who enters measurement. Selective participation makes the strongest performers more visible and turns market comparison into self-selection.

5. Which dietary questions does one star omit?

The system compares a product’s nutritional composition. It does not know how much a particular person eats, how often, or what accompanies it. Four-star cereal is not unlimited food. A low-star celebration food is not harmful in every small and occasional serving.

The algorithm does not directly represent every method of processing and every additive, and cannot substitute for individual needs involving allergy, diabetes or kidney disease. Per-100-gram comparison creates a useful standard but can differ from the practical experience of foods consumed in radically different quantities.

Health Stars mainly apply to packaged products. Fresh fruit and vegetables strongly recommended by dietary guidance do not necessarily need a prominent rating. If consumers infer that a package carrying five stars is inherently healthier than unpackaged food with no label, an explanatory device can redirect attention towards heavily marketed products.

The official message “the more stars, the healthier” must therefore stay connected to “compare similar products”. Without that second clause, a relative aid becomes an absolute certificate of health.

6. How can the stars guide real improvement?

First, coverage should include eligible comparable products as consistently as possible. If display becomes mandatory, implementation needs a transition for small businesses, uniform verification and correction of error. A company should not be able to display only the energy icon instead of the complete rating.

Second, the algorithm needs version control. Nutrition evidence, products and commercial adaptation change. When a category repeatedly receives results that conflict with dietary guidance, the system should publish the anomaly, model the effects of reform and revise the method. Breaks in historical comparability should be marked.

Third, evaluation should study the direction of reformulation, not only the average number of stars. Reductions in sugar, sodium and saturated fat should be distinguished from the addition of a point-earning component that changes the total with less reduction of harm.

Fourth, the label should continue telling people to compare like products and make the underlying nutrient panel and ingredients easy to find. The total narrows the choice; the components explain why.

Fifth, the final evaluation should connect display with shopping and dietary outcomes. Do people use it correctly? Does the basket improve? Does reformulation benefit groups across income levels? Counting packages that carry stars is only an activity measure.

Conclusion: one more star is useful evidence, not a complete verdict

The Health Star Rating solves a real problem. A shopper cannot fully calculate every nutrient during each purchase. A transparent, consistent front-of-pack measure can improve comparison within a category and create an incentive to reformulate. Imperfection is not a reason to abandon it.

My judgement is that a star rating can compare the relative nutritional profile of similar packaged foods. It should not declare a food healthy in every dietary context. The system should move towards consistent coverage, independent verification and continuing revision, and should treat the quality of reformulation—not the count of higher-star packages—as its ultimate outcome.

Once a star influences formulation, it participates in making the supermarket reality it measures. A good indicator makes the easiest route to a higher score the route that reduces real nutritional risk. A bad one teaches food to look healthier. The difference lies not in how brightly the star appears but in whether its algorithm, coverage and correction keep pace with those adapting to it.

Principal sources

Discover more from Geoffrey Chen

Subscribe to get the latest posts sent to your email.