Single-Score Nutrition Labels

Executive Summary

Single-score nutrition label systems promise simplicity. They deliver ambiguity. Deployed across more than forty national markets, these frameworks—Nutri-Score in Europe, Health Star Rating in Australia and New Zealand, the Guideline Daily Amount systems in Southeast Asia, and various proprietary front-of-pack formats—represent the dominant paradigm in consumer nutrition communication. They have also introduced a class of structured information failures that cascades through food trade documentation, regulatory compliance, and cross-market product positioning.

This paper traces the methodological architecture of composite scoring across five jurisdictions, examines the normative assumptions embedded in each scoring algorithm, and analyzes how these assumptions interact with cross-border export documentation requirements. The central argument is not that single-score labels are nutritionally inaccurate—they often approximate population-level dietary risk adequately—but that their methodological opacity and cross-jurisdictional incoherence make them unsuitable as the primary information vehicle for structured food trade disclosure.

The problem is not that Nutri-Score gives olive oil a “C.” The problem is that no exporter can explain why it does, and no importer can replicate the computation from the product’s own documentation.

1. The Architecture of Compression

All single-score systems perform the same fundamental operation: they collapse a multi-dimensional nutritional profile—calories, macronutrients, micronutrients, fiber, additives, processing level—into a single ordinal value. This compression is operationally useful. It is epistemologically destructive.

The process requires three architectural decisions that each scoring system resolves differently:

1.1 Nutrient Selection

Which attributes are included in the composite? Nutri-Score considers energy, saturated fat, total sugar, sodium, fiber, protein, and “fruits, vegetables, legumes and nuts” as a proxy for phytonutrient density. Health Star Rating uses energy, saturated fat, total sugar, and sodium as negative attributes, and protein, fiber, and “fruit, vegetable, nut and legume” content as positive attributes. The Gulf Cooperation Council’s voluntary scheme adds trans-fat and cholesterol. The differences are not trivial.

A product with high omega-3 content, moderate total fat, and low fiber—such as fatty fish—scores well under Health Star Rating (which credits protein directly) but middling under Nutri-Score (which does not distinguish omega-3 from saturated fat in the fat category). The same product carries different scores in France and Australia not because its nutritional composition changed, but because the algorithms resolve the fat question differently.

1.2 Weighting and Reference Values

Every composite score applies weights to component nutrients, either explicitly (assigned coefficients) or implicitly (through reference value normalization). These weights embed normative dietary assumptions—typically those of the authoring jurisdiction’s national dietary guidelines. French dietary guidance emphasizes fat restriction more heavily than Australian guidance; Australian guidance gives greater weight to protein adequacy.

Because these weights are rarely published in machine-readable format alongside the score, an importer receiving a product with a Nutri-Score “B” cannot reconstruct what nutritional composition produced that grade without re-running the proprietary algorithm against the full nutrient panel.

1.3 Category Adjustment

Most systems modify scores based on product category—cheese, fat spreads, beverages, and red meat often receive category-specific adjustments. Nutri-Score applies separate scoring tables for fats, beverages, and “general” foods. Health Star Rating applies a “dairy adjustment” that credits calcium. These adjustments are intended to prevent absurd results (skim milk scoring below sugar-sweetened beverages) but introduce a second layer of opacity: the score now depends not only on the nutrient panel but on the category assignment, which is itself a classification judgment that may not align across regulatory jurisdictions.

2. The Cross-Border Documentation Problem

2.1 Score Portability Failure

A fundamental assumption underlying single-score labels is that they are computed from objective, verifiable nutrient data. In practice, the score is an interpretation of that data under a specific algorithmic framework. When a product crosses a border from a Nutri-Score jurisdiction to a non-Nutri-Score jurisdiction, the score does not travel with it—the score is a market-specific output, not a product-intrinsic attribute.

This creates what we term score portability failure: the score cannot be reproduced by the importing country’s documentation systems, cannot be verified against the original nutrient panel without access to the scoring algorithm, and cannot be used as an interoperable attribute in trade documentation.

2.2 The Audit Gap

For structured trade documentation purposes—export certificates, compliance declarations, third-party audit reports—the relevant question is not whether a product scores well, but whether the score is auditable from the underlying product information.

Auditability requires that any qualified reviewer, given access to the product’s documented attributes, can independently reproduce the claimed output. Single-score systems fail this test systematically.

Consider the audit workflow for an Egyptian food importer receiving a European packaged product:

The product carries a Nutri-Score “B” on front-of-pack
The import documentation references this score as evidence of nutritional compliance
The Egyptian auditor cannot independently verify the “B” because:
- The Nutri-Score algorithm is not part of the import documentation
- The auditor’s jurisdiction does not recognize Nutri-Score as a compliance standard
- The underlying nutrient panel (which is documented) would produce a different result under any alternative scoring system

The score therefore functions as a compliance signal without a verification pathway—precisely the kind of opaque assertion that structured disclosure frameworks are designed to eliminate.

2.3 Regulatory Multiplicity and the Conflict Problem

Fourteen distinct front-of-pack systems are currently in mandatory or voluntary use across the top twenty food export markets. A product optimizing for Nutri-Score may score poorly on Health Star Rating due to fat weighting differences. A product reformulated to improve its GCC voluntary scheme score may worsen its Nutri-Score due to category adjustment thresholds.

This creates a regulatory multiplicity burden for exporters operating across multiple markets: compliance with each system’s score thresholds requires not merely nutritional transparency but strategic nutritional positioning—reformulating products to satisfy multiple scoring algorithms simultaneously.

The burden falls disproportionately on small and medium producers who lack the technical infrastructure to model cross-system scores before product launch.

3. Governance Architecture Analysis

3.1 The Nutri-Score System (France/EU)

Algorithm transparency: Moderate. The Point System is publicly documented; category boundaries are published in regulatory text. However, the category assignment rules are complex and the reference product database is not publicly available in API-accessible format.

Cross-border interoperability: Low. Adopted in France, Germany, Belgium, Spain, Luxembourg, Switzerland, and the Netherlands as of 2025. Not recognized in the UK post-Brexit (which uses the Traffic Light system), not used in North America, Australia, or most Asian markets.

Audit reproducibility: Partial. A technically qualified auditor with access to the full nutrient panel and the scoring documentation can reproduce the score for “general” food categories. Beverages and fat spreads require access to subcategory-specific tables that are not consistently referenced in import documentation.

Documentation failure mode: Products exported from Nutri-Score jurisdictions to non-Nutri-Score jurisdictions arrive with a front-of-pack score that importing market regulators cannot verify and import documentation systems cannot systematically process.

3.2 Health Star Rating (Australia/New Zealand)

Algorithm transparency: High. The HSR Calculator is publicly available as downloadable spreadsheet. Nutrient thresholds and category adjustment rules are published in the Technical Reference Document.

Cross-border interoperability: Low by adoption breadth. Voluntary in both adopting countries; only recognized in Australia and New Zealand. Not harmonized with any other jurisdiction’s system.

Audit reproducibility: High. The public calculator makes score reproduction straightforward for any auditor with the nutrient panel. This is the most technically auditable of the major systems.

Documentation failure mode: Products reformulated for Australian market carry HSR scores that provide no compliance signal in other export markets. Exporters must maintain separate documentation for each destination market.

3.3 Warning Label Systems (Chile, Mexico, Brazil)

Algorithm transparency: High. Warning systems (EXCESS: Sugar, EXCESS: Sodium, EXCESS: Saturated Fat, EXCESS: Calories) are computationally simple—nutrient content divided by reference serving size compared against fixed threshold. No weighting, no composite calculation.

Cross-border interoperability: Moderate and growing. Chile’s warning label system has been adopted or adapted in Mexico, Peru, Uruguay, and Brazil. A MERCOSUR harmonization proposal is under negotiation. The simplicity of the threshold model makes adoption easier.

Audit reproducibility: Very high. Any auditor can reproduce the warning determination from the nutrient panel without algorithm access.

Documentation failure mode: Binary warning system conveys regulatory barrier information (this product will require warning labels in market X) but no positive nutritional quality signal. Not useful for premium product positioning or quality certification claims.

4. The Information Architecture Argument

The central failure of single-score nutrition labeling is not design failure—these systems are often well-designed for their intended purpose (consumer guidance at point of purchase). The failure is category error: applying a consumer simplification tool as an instrument of structured trade disclosure.

Trade documentation serves an audit function. Consumer labels serve a choice-facilitation function. These are different epistemic tasks requiring different information architectures.

Consumer simplification optimizes for:

Low cognitive load at point of decision
Comparability within a product category
Rough risk signal sufficient to guide population-level dietary behavior

Structured trade disclosure optimizes for:

Exact reproducibility across computing environments
Verifiability by parties without access to proprietary scoring systems
Interoperability across regulatory jurisdictions
Auditability by third parties at any point in the supply chain

A tool optimized for the first cannot be optimized for the second. The attempt to use Nutri-Score as a trade compliance signal is functionally analogous to using a credit score as a balance sheet—it summarizes something real but cannot substitute for the underlying documentation from which it was derived.

The solution is not to abandon single-score labels. It is to stop treating them as documentation. The nutrient panel is the document. The score is a derived interpretation of that document for a specific regulatory context.

5. Implications for Structured Disclosure Systems

For platforms and documentation systems operating in cross-border food trade, this analysis suggests several architectural requirements:

5.1 Score as Attribute, Not Identity

Nutrition scores should be stored as market-specific attributes, not product-defining characteristics. A product’s documentation record should include the nutrient panel as the authoritative source and any applicable scores as derived, jurisdiction-tagged fields.

5.2 Algorithm Reference in Documentation

Where a compliance claim references a nutrition score, the documentation should include a reference to the algorithm version used to compute that score, enabling future verification and audit.

5.3 Cross-Jurisdiction Score Projection

Exporters operating across multiple jurisdictions require tools that can project a product’s likely score under multiple systems given a single authoritative nutrient panel. This is technically feasible and would substantially reduce the compliance documentation burden.

5.4 Disclosure of Category Assignment

The product category used in score computation should be part of the score citation in documentation, since category assignment is the most opaque and contested element of most scoring systems.

6. Conclusion

Single-score nutrition label systems are durable features of the regulatory landscape. They will not be replaced. What can change is how they are used—and misused—in trade documentation.

The systematic deployment of composite nutrition scores as primary compliance signals in cross-border food trade represents a structural information failure: it trades the verifiability of structured disclosure for the simplicity of a consumer-facing shorthand. The result is an audit gap that grows with every new market and every new scoring system.

The path forward is not a unified global scoring system—the political economy of such a project is prohibitive, and the scientific debates about optimal nutrient weighting are genuine. The path forward is architectural: maintaining the nutrient panel as the authoritative disclosure document while treating scores as derivative, jurisdiction-specific interpretations.

This preserves the consumer utility of single-score systems while restoring the auditability that structured trade disclosure requires.

Governance Framework Note

This paper was produced under the SGPIS (Structured Governance Papers on Information Systems) framework. The SGPIS framework applies a multi-lens governance analysis to information system design challenges in the food and agricultural product trade.

Lenses applied in this paper:

AI ethics and human interpretability (opacity of algorithmic scoring)
Cross-border regulatory coherence (jurisdictional multiplicity burden)
Consumer information architecture (compression vs. disclosure trade-off)
Structured disclosure frameworks (auditability requirements)

Footnotes

1. Nutri-Score scoring tables: Santé publique France, Mise à jour du Nutri-Score, 2023.

2. Health Star Rating Technical Reference Document v7.0, Australian Department of Health, 2022.

3. GCC Voluntary Nutritional Labeling Guidelines, Gulf Standardization Organization, 2021.

4. Chile warning label regulation: Decree 977/1996 amended by Law 20.606 (2012) and Supreme Decree 13/2015.

5. MERCOSUR Resolution GMC 44/03 (revised 2022) on nutritional labeling harmonization.

6. For a comprehensive registry of front-of-pack systems: EUFIC, Front-of-Pack Nutrition Labelling Systems in Europe and Internationally, 2024.