Proportionality asks four questions of any decision that departs from equal treatment: is the aim legitimate, is the measure rationally connected to it, is there no less intrusive way to achieve it, and is the balance fair? All four have to hold, and the reasoning has to be recorded at the time. It is the test the Equality Act’s exceptions and the public sector equality duty both lean on, and the standard a tribunal will apply.
The four limbs at framework level
The four limbs come from Bank Mellat v HM Treasury (No. 2) [2013] UKSC 39, the Supreme Court’s authoritative statement of the proportionality test, and they are applied to the Equality Act’s exceptions and to the public sector equality duty. The test asks four questions together. First, is the objective sufficiently important to justify limiting a right — a legitimate aim, not an administrative preference. Second, is the measure rationally connected to that aim — does it actually advance it. Third, is no less intrusive means available — could the aim be met in a way that interferes less. Fourth, is the balance struck fair — does the benefit outweigh the harm to the person affected.
All four have to be answered together. A measure that passes three and fails one is not proportionate. The test is a judgement, not a checklist, but it is a structured judgement — each limb has to be addressed, not gestured at.
Why blanket rules cannot pass
A blanket rule fails the third limb almost by definition, because it does not consider whether a less restrictive option would meet the aim in the individual case. It also fails the fourth, because it applies the same harm to every person regardless of circumstance. That is why blanket exclusion is the riskiest option available, not the safe default. The manager decision framework sets this out in the single-sex context; the logic holds across any proportionality-governed decision.
The real exposure is unrecorded reasoning
The limb that catches most organisations is not the fourth — it is the absence of any written reasoning at all. A proportionality judgement that exists only in someone’s head cannot be tested, revisited or defended later. As the defensible decision-making explainer puts it, undocumented reasoning is functionally indistinguishable from no reasoning once it is challenged. The paper shield piece sets out the data behind that gap — policy without infrastructure, guidance without review, process without documentation.
What a defensible record looks like
The record needs to show, on the day the decision was taken: the question, the facts, the legal basis, the options considered, the proportionality rationale applied to the specific case, and the review date. It does not need to be a legal treatise; it needs to show the reasoning happened and was not reconstructed afterwards. That is the difference between a decision that holds up and one that does not.
The four questions are easy to state and hard to answer, because every limb turns on facts only you have: your aim, your evidence, the less intrusive options actually available. And the reasoning has to be written down when the decision is taken — that is the moment it will be judged on.