The impairment is in the body.
The disability is in the arrangements.
The social model of disability, made clickable. Six Indian outcomes drawn as barrier stacks — what stands between a person and school, paid work, a bus, a public building, being counted at all, and the certificate that unlocks the rest.
Open any barrier for the survey, statute or judgment behind it, and for what it leaves unsettled.
The distinction is UPIAS’s, from the Fundamental Principles of Disability (1976). Michael Oliver named it the social model (1983) and developed it in The Politics of Disablement (1990). Anita Ghai has made the Indian case, including where the model travels badly. ImpactMojo added the Indian evidence. Full credits ↓
43.6% of persons with disability entered a public building in a year. 63.9% of those who did found it difficult.
Both figures are from the same survey. Read one way they describe a population; read the other they describe a set of buildings. The model’s claim is that the second reading is available for almost every disability statistic India publishes, and is almost never the one taken.
What the model does not claim
It does not claim that impairment is unreal, or that pain, fatigue and illness are socially produced. That misreading is common enough that Oliver spent much of his later work rejecting it. The claim is narrower: that the distance between having an impairment and living an excluded life is filled by arrangements, and that arrangements were chosen.
The most serious objection comes from within disability studies rather than outside it. A model built to shift attention away from the body can end up unable to talk about the body at all — about chronic pain, or about conditions where treatment is exactly what a person wants. Tom Shakespeare has pressed this; the honest position is that the social model is a political tool of great precision and an incomplete account of a life.
And it translates imperfectly to India. Anita Ghai’s argument is that a model developed against a functioning welfare state assumes a state that has already promised access and failed to deliver it. Where services are thin for everyone, the line between a barrier faced because of disability and one faced because of poverty is not clean — and the family, which the British model often reads as a site of dependence, is frequently the only support that exists.
What actually stands in the way
Every stack below is built from these four. They are not ranked, and they are not independent: the attitudinal barrier is usually why the other three were built the way they were.
Six outcomes, and what is between
Pick an outcome. Each band’s width is how much of the way that barrier blocks; the clear channel on the right is what gets through. Click or tab to a band for the detail behind it.
Whose model this is
The distinction
The Union of the Physically Impaired Against Segregation, Fundamental Principles of Disability (1976). The separation of impairment from disability, and the claim that disability is imposed on top of impairment, are stated there first — by disabled people, about themselves.
UPIAS, with the Disability Alliance
The model
Michael Oliver, who named it in Social Work with Disabled People (1983) and developed it in The Politics of Disablement (1990). Vic Finkelstein and Colin Barnes carried the argument alongside him.
And spent later years rejecting the misreading that it denies the body
The Indian argument
Anita Ghai, whose work sets out what the model gains and loses in a context where the state has not already promised access, and where family support and poverty complicate the line the model draws.
Credited because this page uses the model in exactly that context
What ImpactMojo added
The six barrier stacks, the weighting of each barrier, and the Indian evidence: the survey, statute or judgment behind each one, with a named source and year, plus a note on what each stack does not settle.
ImpactMojo, 2026 · content CC BY-NC-ND 4.0 · code MIT
Sources drawn on
MoSPI, NSS 76th Round — Persons with Disabilities in India (2018) · Rights of Persons with Disabilities Act (2016), ss. 2(r), 16, 32, 34, 44–46, 57 and 58 and the Schedule · Rajive Raturi v Union of India (2017) · Vikash Kumar v UPSC (2021) · Accessible India Campaign (2015) · Unique Disability ID.
Using this in a workshop
Take a service the group runs and build its stack. Then ask the question the model exists to force: for each barrier, who decided it, when, and what would it cost to undo? A barrier with a named decision behind it is a different problem from one that has simply always been there, and most of them turn out to have a decision behind them.
The most useful disagreement is usually about the attitudinal band, because it is the one nobody can be required to remove and the one that explains why the other three were built as they were.
Read alongside Who Counts, which holds the measurement gap this page’s fifth stack describes, and the Capability Approach, whose conversion factors are the same argument in another vocabulary.
If you reproduce the model, credit UPIAS and Oliver. If you reproduce this evidence or the weightings, credit ImpactMojo — and the weightings are ours, so argue with them.