Open letter to FSANZ about mandatory Health Star Rating proposal

Background

Food Standards Australia New Zealand (FSANZ) is currently considering a proposal to make the Health Star Rating mandatory across much of the packaged food supply in Australia and New Zealand. The Health Star Rating has operated as a voluntary front-of-pack labelling system since 2014, giving products a rating from 0.5 to 5 stars based on an algorithm that considers factors including energy, saturated fat, sodium, sugar, protein, fibre and fruit, vegetable, nut and legume content.

Under the proposal, many food manufacturers that currently choose not to display a Health Star Rating would be required to add one to their packaging. FSANZ is also considering a simplified “stars-only” version of the symbol, which would remove the accompanying energy and nutrient information currently shown on some labels. The proposal would therefore represent a major change to how packaged foods are labelled and how consumers are encouraged to judge their nutritional quality.

As a health food company, of course Chief supports the goal of helping people make healthier and more informed food choices. However, we're concerned that FSANZ is proposing to mandate the current Health Star Rating algorithm without first conducting a full review of how it works. FSANZ’s own supporting documents acknowledge that a full algorithm review was not undertaken as part of this proposal, despite known questions about the treatment of saturated fat, protein and different food categories.

The current algorithm has been assessed against the 2013 Australian Dietary Guidelines, while updated guidelines are expected to be published in late 2026. Introducing mandatory packaging changes before those updated guidelines are finalised creates a real risk that businesses will be required to redesign and reprint packaging twice - first to comply with the proposed mandate, and again if the algorithm changes following the dietary guideline review. These costs will ultimately affect product prices, packaging waste and the ability of smaller businesses to compete.

In the interests of transparency, we're including the key parts of our response to FSANZ.

Do you support mandating the HSR rather than keeping it voluntary? Why/why not?

No. Chief does not support mandating the Health Star Rating (HSR) system in its current form and at this time. We support the policy goal of helping consumers make healthier choices, but the proposal as drafted would rely on an algorithm that FSANZ has chosen not to review and on dietary guidelines that are simultaneously being rewritten.

Our reasons are set out below.

The algorithm is being mandated without being reviewed

FSANZ proposes to lock the current algorithm into the Food Standards Code while expressly deferring known issues to a future review. SD3 states plainly that "a full review of the algorithm was not undertaken as part of this proposal" (SD3, s1.5). The detail is in our answer to Question 4.

It's anchored to 13 year old guidelines that are under revision

The whole justification for the proposal is that the algorithm aligns with the dietary guidelines, and that alignment was measured against the 2013 Australian Dietary Guidelines (ADGs) (SD3, s2.1), which are due to be replaced in late 2026 (CFS, s2.3.1). Mandating now risks forcing a second, avoidable round of packaging changes. We make this argument in full in our answer to Question 19.

The cost-benefit case is unproven and falls hardest on small business and consumers

Benefits are entirely unquantified; most costs are unquantified; FSANZ itself notes the burden is proportionately higher for small manufacturers (SD6, s2.2.7); and these costs will be passed on to consumers in higher grocery prices. 

The consumer evidence does not support a mandate

FSANZ's own literature review found that while laboratory experiments suggest the HSR can shift choices, "field experiments showed limited impact on purchasing decisions, except for those who were motivated to use the HSR" (CFS, s3.1.2).

Two-thirds of consumers do not understand that the HSR is only meant to compare similar products (CFS, s3.1.2). The nutritionists and health practitioners we talk to find it just as confusing. Most can't tell you how the number is worked out or which foods are in the same category. If the experts can't follow it, they can't guide their clients with it.

This is not a minor misunderstanding: the HSR is calculated using six different category formulas (Categories 1, 1D, 2, 2D, 3 and 3D), each with different components, thresholds and a different score-to-star conversion table (CFS, Attachment C). A consumer comparing a flavoured milk, a muesli bar and a soft drink at face value is comparing numbers produced by entirely different calculations. Mandating a single star symbol on almost everything, with no indication of category on pack and (under the stars-only proposal) no nutrient context, actively invites the cross-category comparison FSANZ says is wrong. 

Trust is moderate (59% in Australia, 50% in New Zealand) and is reduced precisely by consumers not understanding how the rating is calculated and by the belief that ratings can be manipulated (CFS, s3.1.2). FSANZ also concedes it "does not have a mandate or resources to deliver broad, ongoing public education" (CFS, s5.1).

What Chief Asks

1. Defer mandating the HSR until the revised ADGs are published and the algorithm has been reviewed against them, so that any consequential changes are made once, not twice.
2. Undertake a full review of the HSR algorithm before any mandate takes effect, addressing the saturated fat and protein-eligibility issues FSANZ has identified but deferred.
3. If FSANZ nevertheless proceeds, adopt the longer transition requested by industry (3 to 5 years plus a 2-year stock-in-trade period) and align it with routine label-change cycles to minimise cost and packaging waste.
4. Collect and publish robust small-business cost data (including the data in this submission) before concluding that the proposal delivers a net benefit, and account for the flow-on cost to consumers.

Proposed approach to calculating the HSR / the algorithm

Chief does not support mandating the current algorithm, for three reasons drawn from FSANZ's own analysis.

1. It has not been reviewed. SD3 states that "a full review of the algorithm was not undertaken as part of this proposal" (SD3, s1.5). Mandating an unreviewed algorithm entrenches known problems in the Code.
2. FSANZ concedes the saturated fat treatment may not reflect current evidence, yet declines to act. On cheese, SD3 records industry requests to review saturated fat "to better align with recent evidence suggesting saturated fat content in dairy foods is not harmful to health", and to review protein because "most cheeses do not currently receive protein points ... even though they are a good source of protein". FSANZ's response is to await the ADG review (SD3, s3.2.1). 391 of 710 core cheeses (55%) score below 3 stars. The same saturated fat penalty drives Chief's packaged meat products (biltong, beef bars and beef sticks) to ratings of 0.5 to 1 star.
3. The algorithm penalises our prepared meat products for what real meat is made of, and denies them credit for their protein. 

Treatment of FVNL content, and the protein-points rule

The algorithm marks our products down for what real meat is made of, then denies them credit for their protein. This is the clearest demonstration that the system measures the wrong thing.

To be clear on scope: we are not referring to a plain cut of beef, which is not required to carry a rating. We are referring to the prepared, packaged meat products that are required to carry one, namely our biltong, beef bars and beef sticks. These are marked down for two things that are intrinsic to real, minimally processed meat:

• the saturated fat naturally present in the beef; and
• sodium. In dried and cured meats this is not discretionary seasoning. Salt is one of the food-safety hurdles that keeps the product shelf-stable and controls pathogens such as Salmonella and E. coli, and it is a documented critical control in our process. It cannot be removed without compromising safety.

The HSR treats both as simple negatives, which alone drives our products to 0.5 to 1 star.

It then compounds this through the protein rule. A food scores protein points only if it has fewer than 13 baseline points, OR has 13 or more baseline points AND also scores 5 or more fruit, vegetable, nut and legume (FVNL) "V" points (CFS, Attachment C). A meat-based product that is high in protein but necessarily higher in saturated fat and sodium exceeds 13 baseline points, and because it is built on meat rather than plants it has little or no FVNL content. So one of the most protein-dense foods on the shelf earns no protein points at all, while a more processed product with added fibre, added FVNL and intense sweeteners replacing sugar can clear the gate, bank its protein points, and out-rate it (SD3, s2.4.2.2; see Question 9).

This is not a marginal anomaly. It is the algorithm working as designed, and it is the opposite of what dietary guidelines intend when they recommend lean meats, poultry, fish and eggs as a core food group. A front-of-pack system that rates a real-meat product below a processed snack or a diet soft drink does not help consumers, and should not be mandated.

Algorithm overrides

The proposed overrides illustrate the system's inconsistency. Water and fresh or minimally processed fruit and vegetables receive automatic ratings of 4.5 to 5 stars, while a minimally processed meat product is capped near the bottom of the scale by the standard algorithm.

At the same time, FSANZ's own data shows highly processed products scoring well: 398 of 844 carbonated drinks (47%) and a substantial share of sugar-based confectionery reach 3 stars or higher, largely through "the use of added intense sweeteners to replace sugar", and 211 of 347 misaligned snack products (60.8%) such as crisps score 3 stars or more (SD3, s2.4.2.2).

A system in which intense-sweetened drinks and crisps can out-rate a real-meat product undermines, rather than supports, informed choice.

A stars-only symbol

Chief does not support restricting the symbol to the stars element only. Removing the energy and nutrient icons strips away the context a consumer needs to interpret a low rating.

For a minimally processed meat product that rates poorly chiefly because of its saturated fat and sodium, the nutrient icons are the only on-symbol means of conveying, for example, low sugar or high protein. Mandating a single, context-free star count makes a misleading headline rating harder, not easier, for consumers and health practitioners to put in perspective. It also worsens the cross-category problem described in Questions 1 and 19, because the star number is presented with nothing on the symbol to signal that different food types are scored differently.

Have all major impacts been identified in Table 2 of SD6?

No. Table 2 of SD6 omits:

• the disproportionate burden on small manufacturers (SD6, s2.2.7 acknowledges this but does not quantify it);
• the flow-on cost to consumers, as label-change, reformulation and ongoing education costs are passed into the cost of goods during a cost-of-living crisis (FSANZ's own focus groups noted that ensuring implementation costs are not passed to consumers was considered important, CFS, s4.1);
• the cost of a potential second relabelling following the ADG review (see Questions 18 and 19);
• the loss of marketing and compliance flexibility resulting from a stars-only symbol (see Question 11); and
• the packaging waste created by relabelling against large print minimum order quantities (see Questions 17 and 19).

Any other comments not related to specific questions

Chief wishes to make four points that cut across the specific questions.

The HSR should not be mandated ahead of the Australian Dietary Guidelines review

This is, in our view, the single strongest reason to defer the proposal.

The entire justification for the proposal is that the HSR algorithm aligns with the dietary guidelines. But that alignment was measured against the 2013 Australian Dietary Guidelines (SD3, s2.1), which the Call for Submissions confirms are "currently under review, with updated guidelines due for publication in late 2026" (CFS, s2.3.1).

FSANZ's own approach is internally inconsistent. It is willing to wait for the revised ADGs before resolving the cheese and saturated fat issue (SD3, s3.2.1), before considering the use of colour in the symbol (CFS, s4.4.4), and it excludes ultra-processed foods from the algorithm because they are "not covered in the current dietary guidelines" (CFS, s4.3.7). In other words, FSANZ accepts that the revised guidelines are likely to change what the algorithm should measure, yet proposes to make the current algorithm mandatory before those guidelines exist. It cannot be right to defer those fixes to the revised guidelines while simultaneously locking the unrevised algorithm into law.

The practical consequence is a real risk of two rounds of mandatory packaging change. If the revised ADGs prompt algorithm changes, which FSANZ's own deferrals suggest is likely, every business that has just relabelled to comply would have to relabel again. SD6 treats label change as a one-off cost and does not account for this. For Chief, a second round could double the costs set out in Question 17. Because packaging is committed to large minimum order quantities (400,000 to 700,000 units per pack), each relabelling also means writing off existing pre-printed stock. This falls on a food supply that is roughly 70% affected and dominated by small manufacturers.

This is entirely avoidable. The proposal still has a second round of consultation, a draft variation, ministerial consideration and a transition period ahead of it. Sequencing the mandate to follow the ADG review is therefore achievable without abandoning the policy goal. We ask that the HSR algorithm be reviewed against the revised ADGs and finalised before any mandatory obligation commences.

These costs will be passed on to consumers during a cost-of-living crisis

The costs in Question 17 are not absorbed by business alone. For a small manufacturer they must be recovered through the cost of goods, which means higher prices on the shelf. Mandating relabelling, and potentially a second round after the ADG review, during a cost-of-living crisis, would raise grocery prices at the worst possible time. FSANZ's own focus group participants identified that ensuring implementation costs are not passed on to consumers was important (CFS, s4.1). The current cost analysis does not address this flow-on effect at all.

Consumers and health practitioners compare across categories without realising the scores are calculated differently

The HSR is not a single scale. It is six different category formulas (Categories 1, 1D, 2, 2D, 3 and 3D), each using different components, different thresholds, different scaling, and a different score-to-star conversion table (CFS, Attachment C). A 4-star drink, a 4-star yoghurt and a 4-star muesli bar are not produced by the same calculation.

FSANZ states the HSR is only meant to compare similar products, and concedes two-thirds of consumers do not understand this (CFS, s3.1.2). Mandating a single star symbol on almost the entire packaged food supply, with no indication of category on the pack, and (under the stars-only proposal) no nutrient context, does not fix this problem. It makes it worse, by putting more cross-category star comparisons in front of consumers who do not know the scores are not comparable. A system that invites the very comparison its designers say is invalid is not fit to be mandated in its current form.

References

In-text references in this submission use FSANZ's own P1067 documents, identified as follows:

• CFS - FSANZ 2026, Proposal P1067 Health Star Rating System, 1st Call for Submissions (12 June 2026).
• SD1 - FSANZ 2026, Proposal P1067, Supporting Document 1: Health Star Rating Consumer Literature Review.
• SD2 - FSANZ 2026, Proposal P1067, Supporting Document 2: Nutrition Labelling Focus Groups.
• SD3 - FSANZ 2026, Proposal P1067, Supporting Document 3: Performance of the Health Star Rating algorithm.
• SD6 - FSANZ 2026, Proposal P1067, Supporting Document 6: Preliminary consideration of costs and benefits.

Prepared by Sarah Thornborough
QA Manager, Chief Nutrition