If only 22% of India's estimated children with special needs are enrolled in school, what does that 22% actually measure - the children who exist or only the children the system has learned to see?

The Enrollment Gap That Is Really a Counting Gap In 1995,

Parliament passed India's first disability rights law and recognised exactly seven categories of disability. It took until 2016 - twenty-one years, one Supreme Court-adjacent civil society campaign and India's 2007 ratification of the UN Convention on the Rights of Persons with Disabilities - for that number to jump to 21, folding in conditions like autism, specific learning disabilities, dwarfism and acid-attack injury that the 1995 law simply had no category for.

This is not an unusual pattern in Indian rights legislation: the law expands first, on paper and the machinery that is supposed to count, fund and serve the people the law now covers catches up years, sometimes decades, later.

The UDISE 2025-26 data is the latest instance of that lag showing up as a number that looks like failure but may actually be a filing problem. Only 21.66 lakh Children With Special Needs are enrolled in schools - a figure that has barely moved in eight years - against an estimated 92-99 lakh CWSN in the country.

That 22% enrollment rate is being read, understandably, as evidence that India has failed to get disabled children into classrooms. But The Hindu's own reporting surfaces something sharper: lawyer Shashank Pandey's observation that many schools filing UDISE data are still sorting children into the 2011 Census's older, narrower disability buckets rather than the 21 categories the RPWD Act legally requires.

A child with a specific learning disability or an autism spectrum condition - categories that did not exist in the state's paperwork before 2016 - may simply not be getting recorded as CWSN at all. This is not a defence of the numbers.

Even a generous reading leaves a real, serious gap: the drop-off between primary and higher secondary is sharper for CWSN than for students overall - enrollment nearly halves between upper primary and secondary - and only 43 of every 100 enrolled CWSN are girls.

Those are retention and access failures a data-classification fix cannot explain away. But the 1995-to-2016 pattern tells us where to look for the fix. When Indian law expands a category of rights faster than its administrative categories can absorb, the resulting "gap" is really two gaps stacked on top of each other: a real one, in access and retention and a measurement one, in how the state counts who it has failed.

Solving only the first without fixing the second means India could spend a decade improving actual enrollment while UDISE data shows almost no change - because the children being newly enrolled under expanded categories were never subtracted from the "missing" pool to begin with.

The two-decade wait between the 1995 Act and the 2016 Act's 21 categories is the real precedent here - and it suggests the next fix, updating how schools actually classify disability, will take years unless someone forces the administrative system to catch up faster than the law did.