Protected characteristics are specific attributes defined under the UK's Equality Act 2010 that are safeguarded against discrimination. Groups with these characteristics refer to people who share one or more of them, meaning public authorities must proactively address inequalities affecting such groups in their policies and decisions. For the aim of advancing equality of opportunity (which includes removing disadvantages, meeting different needs, and encouraging participation), the relevant protected characteristics are:Age: Covers people of a particular age or age range (e.g., children, young adults, elderly individuals).
Disability: Includes those with physical or mental impairments that have a substantial, long-term adverse effect on daily activities (e.g., people with mobility issues, chronic illnesses, or neurodivergence).
Gender reassignment: Applies to transgender individuals who are proposing to undergo, are undergoing, or have undergone a process to reassign their sex (no medical intervention required).
Pregnancy and maternity: Protects women during pregnancy and for 26 weeks after giving birth (or longer in some cases, like maternity leave).
Race: Encompasses colour, nationality, ethnic or national origins (e.g., Black, Asian, White, or mixed heritage groups; also includes groups like Gypsies and Travellers).
Religion or belief: Covers religious groups (e.g., Christians, Muslims, Jews, Hindus) and non-religious beliefs (e.g., atheism, humanism, or philosophical beliefs like pacifism).
Sex: Refers to men or women (biological sex, not gender identity).
Sexual orientation: Includes heterosexual, gay, lesbian, or bisexual orientations.
Note that "marriage and civil partnership" is a protected characteristic under the Act but does not apply to the specific aim of advancing equality of opportunity—it's mainly relevant for eliminating discrimination. Authorities assess impacts on these groups through tools like Equality Impact Assessments to ensure equitable outcomes.
Recent EHRC guidance highlights its intersection with data protection laws . Can you explain what guidance is given?
The recent guidance from the Equality and Human Rights Commission (EHRC), published in September 2024 and referenced in subsequent updates, focuses on the interplay between the Public Sector Equality Duty (PSED) under the Equality Act 2010 and data protection laws, primarily the UK GDPR and Data Protection Act 2018.
equalityhumanrights.com
It aims to help public authorities collect and use equality information effectively while ensuring compliance with privacy requirements. Below, I'll outline the key elements of the guidance, structured by principles, requirements, examples, pitfalls, and tips.Main PrinciplesThe guidance emphasizes that PSED compliance often involves processing personal data related to protected characteristics (e.g., age, disability, race, religion, sex), which must align with data protection principles like lawfulness, fairness, transparency, data minimisation, accuracy, storage limitation, integrity, confidentiality, and accountability.
equalityhumanrights.com
While there's no explicit PSED requirement to collect equality data, it's often necessary to evidence "due regard" to the duty's three aims: eliminating discrimination, advancing equality of opportunity, and fostering good relations. Special category data (e.g., health-related aspects of disability or racial origin) requires extra safeguards, and proxy data (e.g., postcodes implying ethnicity) should be treated cautiously as it can indirectly reveal protected characteristics.Legal RequirementsGeneral PSED Alignment: Authorities must integrate data handling into their functions without breaching privacy laws. For special category or proxy data, a lawful basis under Article 6 of the UK GDPR (e.g., public task) and an Article 9 condition (e.g., substantial public interest) are needed.
equalityhumanrights.com
Country-Specific Duties: In England (and Great Britain-wide bodies), publish annual equality information on workforce and services (for organizations with 150+ staff), measurable objectives every four years, and gender pay gap reports for 250+ employees.
In Scotland, produce biennial mainstreaming reports, gender pay gaps every two years, equal pay statements every four years (for 20+ employees), and equality outcomes with impact assessments.
In Wales, collect and publish relevant information periodically, including pay differences and annual reports on progress, with equality objectives covering all protected characteristics and action plans for gender pay gaps.
equalityhumanrights.com
Data Protection Overlaps: All processing must be documented, with risks mitigated, especially in AI or automated systems where biases could arise from proxies.
Examples of ComplianceEquality Impact Assessments: Gather anonymized data on protected groups to evaluate policy effects, then publish aggregated findings if impacts are significant (e.g., adjusting a service based on disability data without identifying individuals).
equalityhumanrights.com
Workforce Monitoring: Collect optional data on employees' protected characteristics for annual reports, ensuring it's anonymized and used only for equality purposes.
AI and Proxy Use: In fraud detection, remove proxy indicators like "carers' days off" (which might correlate with sex) to prevent discriminatory outcomes, as seen in guidance referencing real-world cases like biased algorithms.
equalityhumanrights.com
Publishing Reports: Share gender pay gap data with action plans, focusing on aggregated metrics to demonstrate progress without privacy breaches.
Potential PitfallsExcessive Data Collection: Gathering more information than needed violates data minimisation, potentially leading to fines up to £17.5 million or 4% of turnover.
equalityhumanrights.com
Proxy-Induced Bias: Using indirect data (e.g., surnames for race) in profiling without checks can cause indirect discrimination, as in the Dutch benefits scandal where AI unfairly targeted based on nationality.
Lack of Transparency: Failing to inform people about data use or rights, or not securing data properly, risks accountability breaches and legal challenges (e.g., third-party contracts without oversight, per the Bridges case on facial recognition).
Inadequate Monitoring: Not reviewing data retention or impacts could exacerbate inequalities, inviting judicial review.
Practical TipsPre-Collection Planning: Define necessities, justify processing, and make data provision optional; explain uses clearly on forms.
equalityhumanrights.com
Security Measures: Limit access to trained staff, implement policies, and inform individuals of protections.
Documentation: Maintain records of processing activities, including lawful bases, per Information Commissioner's Office (ICO) guidance.
AI-Specific Advice: Analyze proxies for correlations with protected characteristics, monitor for biases, and adjust models accordingly.
Engagement and Review: Consult affected groups for evidence, review data periodically, and refer to ICO resources on anonymisation, AI, and data principles for further support.
equalityhumanrights.com
This guidance promotes a balanced approach, ensuring equality efforts enhance trust without compromising privacy. For the full document, visit the EHRC website.
