Three definitions, one population

The UK counts disabled people in at least three distinct ways, and the resulting figures are not interchangeable — yet all three are routinely cited as though they measure the same thing.

A population, differently drawn

Start with the Equality Act 2010. Its definition of disability is a legal test: a person has a disability if they have a physical or mental impairment that has a substantial and long-term adverse effect on their ability to carry out normal day-to-day activities. "Substantial" means more than minor or trivial. "Long-term" means twelve months or more, or likely to last twelve months, or likely to recur. Some conditions — cancer, HIV infection, multiple sclerosis — are deemed to meet the definition from the point of diagnosis, without any further functional test. The Equality Act definition exists to determine who is entitled to protection from discrimination. It is a threshold, drawn by Parliament and interpreted by tribunals and courts case by case. It does not produce a count.

To count, you need a survey. The principal instrument for producing a population estimate is the Annual Population Survey, which feeds the main disability employment and pay statistics published by the Office for National Statistics and the Department for Work and Pensions. Respondents are asked whether they have a physical or mental health condition or illness that has lasted or is expected to last twelve months or more, and whether that condition reduces their ability to carry out day-to-day activities. Both parts must be satisfied. This is the survey operationalisation of the Equality Act definition — but it is operationalisation by self-report, which means the count depends entirely on how individuals interpret the question and apply it to themselves. The survey cannot verify impairment; it records identification.

The third instrument sits in the benefits system. Personal Independence Payment — the main working-age disability benefit in Great Britain — determines eligibility through a functional assessment scored against a set of daily living and mobility descriptors. An assessor, working from a report and sometimes a face-to-face or telephone consultation, scores an individual's ability to perform specified activities. Points are added; a threshold determines whether a standard or enhanced rate is awarded. The DWP publishes caseload statistics quarterly. Those figures count people receiving the benefit, which is a subset of people with disabilities defined under either of the other two instruments, filtered further by the eligibility rules, the application process and, critically, by who has applied at all.

What the instruments cannot share

Each instrument was built for a different purpose, and the population it identifies reflects that purpose. The Equality Act definition is protective: it extends the reach of discrimination law as far as Parliament intended, and its qualifying conditions and deemed inclusions are legal policy choices, not epidemiological ones. The survey definition is descriptive: it aims to estimate, within the limits of self-report, how many people in the general population experience disabling conditions, producing figures that can be tracked over time and broken down by employment status, sector, age and pay. The benefits assessment is allocative: it decides who receives a specific payment, and the criteria are financial and administrative, calibrated to the cost and intent of that particular programme.

Because the wording decides the number, changing any element of a survey question shifts the count. The transition from the previous "limiting long-standing illness" framing to the current Equality Act-aligned wording in the Annual Population Survey produced a measurable discontinuity in the series — not because the population changed, but because the instrument did. Similarly, changes to PIP assessment criteria or the descriptors used to score daily living activities change the caseload without necessarily reflecting any change in the underlying prevalence of impairment.

The distances between the three counts are not small. ONS figures from the Annual Population Survey consistently place the number of disabled people in the UK above fourteen million. DWP PIP caseload figures for Great Britain are counted in the low millions — a figure several times smaller, because PIP measures receipt of a specific benefit, not disability as a population characteristic. The Equality Act definition, were it applied mechanically to the survey population, would produce its own number, which would differ from both — partly because of how tribunals have interpreted "substantial", and partly because some conditions deemed to qualify under the Act are under-reported in self-completion surveys. The three numbers cannot be added, averaged or treated as estimates of the same quantity.

Governance: who publishes what, and who is responsible

Who publishes what is not a simple answer. The ONS publishes labour market statistics derived from the Annual Population Survey. The DWP publishes both the Family Resources Survey — which produces estimates of disability prevalence and household income — and the Equality Act-aligned statistics on the disability employment gap. The DWP also publishes PIP and other benefit caseload statistics. The Equality and Human Rights Commission monitors compliance with the Equality Act and publishes reports drawing on multiple sources. None of these bodies coordinates with the others to reconcile the figures, because reconciliation would require agreement on which instrument is authoritative — and there is no such agreement, because the instruments are not competing estimates of the same thing. They are different measurements of differently defined populations.

The Office for Statistics Regulation, an independent body operating under the Statistics and Registration Service Act 2007, can assess statistical outputs for quality, coherence and accessibility. It can designate statistics as National Statistics, indicating conformity with the Code of Practice for Statistics. It cannot mandate which definition a department uses, or require that figures produced under different definitions be accompanied by systematic comparison. The code requires producers to describe their methodology and be transparent about limitations. Whether readers of a press release or a select committee briefing then apply the appropriate caution is another matter.

What this leaves, in practice, is three separate statistical streams, each internally coherent, each labelled in a way that suggests it is counting disabled people, and each drawing on a definition that serves the function it was built for. A researcher citing employment statistics, a journalist reporting benefit caseloads, and a policy document invoking Equality Act protections may all use the phrase "disabled people" in a single paragraph while drawing, without acknowledgement, on three incompatible population definitions. The gap between those definitions is not a measurement error. It is a structural feature of the way disability is defined in UK law and administered in UK practice.

Threshold is an independent publication about disability statistics and governance in the UK. It is not an advice, advocacy or support service and does not interpret law or guidance for any individual's circumstances.