The Data That Does Not Exist
Ask a simple question — what does European hair look like? — and you will discover that the answer, in any scientifically rigorous sense, is unknown.
Hair data exists, of course. L’Oreal, Procter & Gamble, Henkel, and other cosmetics corporations maintain proprietary datasets built over decades of product development. But these datasets share four structural limitations that make them inadequate for research on identity-based discrimination:
They are proprietary. Access is controlled by commercial interests and not available to independent researchers, policymakers, or the public.
They are biased toward Eurocentric hair types. Cosmetics research has historically focused on straight and wavy hair — the types that constitute the majority market in Europe. Textured hair (types 3A through 4C on the Walker classification system) is systematically underrepresented. Washington State University’s 2024 deep hair phenomics study, published in the Journal of Investigative Dermatology, documented the extent of this bias in existing hair research datasets.
They are visually assessed, not sensor-measured. The Walker 1A-4C classification system — the most widely used hair typing framework — relies on visual assessment of curl pattern. It does not account for fibre diameter, porosity, hydration, protein structure, cross-section geometry, or chemical treatment history. Two individuals classified as “3B” may have fundamentally different hair properties. Visual classification produces categories; it does not produce data.
They are not linked to psychosocial data. Cosmetics datasets characterise hair for product matching. They do not link physical hair properties to discrimination experiences, psychological outcomes, economic impact, or identity-related variables. Without this linkage, the data cannot answer the questions that CROWN’s research programme is designed to address.
The CROWN Hair Commons is designed to address all four limitations simultaneously.
What the CROWN Hair Commons Is
The CROWN Hair Commons is Europe’s first open, multi-ethnic, sensor-verified hair dataset. It aggregates anonymised CROWN Hair DNA profiles generated by CROWN’s multi-sensor diagnostic device and links them — with informed consent — to demographic data and CDI survey responses.
Each profile in the Commons includes:
Hardware-Verified Physical Data
- Fibre diameter (mean, variance, and distribution across sampled strands)
- Cross-section ellipticity index (the degree to which the fibre cross-section departs from circular)
- Cuticle layer count and condition score
- Curl pattern classification (sensor-verified, not visually assessed)
- Porosity index (measured via impedance sensing, not the “float test”)
- Hydration level (NIR spectroscopy)
- Protein integrity index (keratin structure assessment)
- Chemical treatment history (detected through NIR spectroscopic analysis of chemical residues)
- Scalp health indicators
- Breakage susceptibility index
- Environmental damage score
Linked Psychosocial Data (with separate consent)
- CDI survey responses (discrimination prevalence, severity, economic impact)
- Demographic data (age range, gender, country of residence, self-identified ethnicity)
- Self-reported discrimination experiences
- Product usage patterns
- Hair satisfaction measures (based on validated instruments including the scale used in UConn’s 2025 Hair Satisfaction Study)
This linked architecture is what distinguishes the CROWN Hair Commons from every existing hair dataset. It connects what hair is (measured objectively by sensors) to what hair means (experienced subjectively by individuals) — enabling the kind of research that neither cosmetics databases nor survey-only studies can support.
Scale and Diversity
CROWN’s target is 100,000 multi-dimensional profiles by 2030.
This target is not arbitrary. Population-scale research on hair discrimination requires sufficient sample sizes across all hair types, ethnicities, age groups, genders, and European countries to produce statistically stable estimates. Power analysis for the CDI’s intended disaggregation (by geography, sector, demographic group, and time period) indicates that a minimum of 100,000 profiles is required to support the cross-national comparisons that policymakers need.
Critically, the Commons is designed for representative diversity from inception. This is a methodological decision, not a values statement (though it is that too). An AI classification model trained on a dataset that underrepresents textured hair will produce biased classifications for textured hair. A discrimination index calibrated against hardware data that skews toward straight hair will miscalibrate its severity weights for the populations most affected by discrimination.
The Commons’ diversity requirements are embedded in its data collection protocol: stratified sampling targets ensure that all hair types (1A through 4C), all major ethnic groups represented in Europe, and all age ranges are represented in proportions sufficient for stable statistical inference.
Data Governance
Access Tiers
The CROWN Hair Commons operates under a tiered access model:
Academic Access. Free or nominal cost for verified academic researchers conducting non-commercial research. Researchers submit a data access request describing their study objectives, methodology, and ethical approval. Approved researchers receive access to anonymised datasets at the level of aggregation appropriate to their research question.
Public Policy Access. Free for government agencies, legislative bodies, and international organisations conducting official policy analysis. This tier supports CROWN’s legislative advocacy mission by ensuring that policymakers have unrestricted access to the evidence base.
Commercial Access. Tiered access for commercial entities, discussed privately on a case-by-case basis. Revenue from commercial access supports CROWN’s research programme. Commercial users are bound by data use agreements that prohibit re-identification, restrict use to stated purposes, and require attribution.
Privacy and Ethics
The CROWN Hair Commons is designed around privacy-by-design principles in compliance with both the EU General Data Protection Regulation (GDPR) and the Swiss Federal Act on Data Protection (nDSG).
Anonymisation at collection. Personal identifiers are separated from hair and survey data at the point of collection. The linked dataset contains no names, contact information, or identifiers that could be used — alone or in combination — to re-identify individuals.
Layered consent. Participants provide separate informed consent for each data component: diagnostic assessment, survey completion, data linkage, and contribution to the Commons. Participants can consent to some components and decline others. Consent can be withdrawn at any time, triggering deletion of the participant’s data from the Commons.
Ethical review. All data collection protocols are subject to ethical review in accordance with the standards of the University of Geneva and applicable Swiss and EU regulations. The review process evaluates data minimisation, purpose limitation, storage security, and participant protection measures.
Data minimisation. The Commons collects only the data necessary for its stated research purposes. It does not collect biometric data sufficient for individual identification (e.g., facial images, fingerprints). Hair profile data, while detailed, does not enable re-identification of individuals.
Why “Commons”
The choice of the word “commons” is deliberate. It signals that this data infrastructure is a shared resource — not a proprietary asset to be exploited, but a public good to be governed responsibly.
The model draws on the tradition of scientific data commons — the Human Genome Project, the European Bioinformatics Institute, the World Health Organization’s Global Health Observatory — in which data generated through collaborative effort is made available under governed access conditions that balance openness with privacy and responsible use.
CROWN’s Hair Commons extends this tradition into a new domain: human appearance data linked to discrimination outcomes. The result is a dataset that serves academic research, public policy, and — through governed commercial access — the development of products and services that genuinely serve the populations represented in the data.
Relationship to Other CROWN Programmes
The CROWN Hair Commons is not a standalone data initiative. It is the data layer that connects CROWN’s four pillars:
- Research. The Commons provides the dataset for CDI research, enabling population-scale analysis of discrimination prevalence and severity.
- Innovation. The Commons provides the training data for the AI classification engine that powers the diagnostic device. The engine’s accuracy depends on the diversity and quality of the Commons.
- Healing. Anonymised clinical outcome data from CROWN’s therapeutic protocol — standardised psychological assessments administered before, during, and after treatment — contributes to the evidence base on the psychological impact of hair discrimination.
- Advocacy. Population-level Commons data, analysed through the CDI framework, provides the quantitative evidence that European legislators need.
Current Status
The CROWN Hair Commons is in its architecture and governance design phase. The data infrastructure — including storage, anonymisation pipelines, access control systems, and API design — is being developed in parallel with the diagnostic device that will generate the data.
Initial data collection will begin through the CDI pilot study, developed in consultation with the University of Geneva. As the diagnostic device moves from development to deployment — through the emerging ETH Zürich collaboration and eventual salon partnerships — the rate of data collection will accelerate.
Researchers interested in the CROWN Hair Commons — whether for potential data access, collaborative data collection, or methodological consultation — are invited to contact CROWN at contact@crown.ngo or visit our Research Partners page.
The CROWN Hair Commons is governed under CROWN’s privacy policy and in compliance with GDPR and Swiss nDSG. All data collection protocols are subject to ethical review. For questions about data governance, contact contact@crown.ngo.