Every year, millions of data points flow into the Unified District Information System for Education Plus, known to most of us simply as UDISE+. They describe India’s schools in remarkable detail — enrolment by management type, teacher availability, electricity connections, internet access, libraries, smart classrooms, and dozens of other indicators that together sketch the physical and administrative shape of the country’s classrooms.

Most readers open these reports looking for a single answer: are government schools improving? Far fewer ask the second, harder question — are parents responding to that improvement? When the UDISE+ reports for 2022–23 through 2025–26 are placed side by side, they reveal one of the more unsettling patterns in contemporary Indian education. Government schools are, by most physical and administrative measures, getting better nearly every year. Parents, meanwhile, continue to walk away from them.

Should this trajectory hold, India may be approaching its most significant structural shift in schooling since Independence — not because government schools are failing outright, but because parental aspiration is moving faster than public-sector reform can follow.

01The Great Shift

The enrolment numbers, taken on their own, might look unremarkable. Government schools accounted for roughly 51.9 per cent of enrolment in 2023–24; that share fell to 50.83 per cent the following year and to 49.66 per cent in 2025–26. Private unaided schools moved in the opposite direction over the same period, rising from about 36.5 per cent to 37.97 per cent.

These are not small movements. India has close to 250 million school-going children, so every single percentage point of shift represents an estimated 2.5 million children changing the kind of school their family chooses for them. This is not a handful of schools losing favour with their catchment area. It is millions of households, across very different states and income levels, arriving independently at the same decision.

02Yet Government Schools Keep Improving

Here is where the story becomes genuinely surprising. If parents were leaving because government schools were deteriorating, the infrastructure data ought to confirm it. Instead, it shows the reverse. Computer availability in government schools rose from 47.7 per cent in 2022–23 to 69.9 per cent in 2025–26. Internet access rose from 49.7 per cent to 67.4 per cent over the same window. Electricity supply, already high, edged towards near-universal coverage at 95 per cent. Libraries expanded modestly but steadily, and the pupil–teacher ratio at the primary level improved from 23 students per teacher to 19.

Nearly seven in ten government schools now have a computer. Internet availability has climbed sharply within a few years. Electricity coverage is approaching completeness. Class sizes at the primary level are shrinking, and library access continues to widen. This is, by most conventional measures, exactly the outcome that decades of education policy set out to achieve — and it is happening in parallel with continued enrolment decline, not instead of it.

03Naming the Paradox

“A hospital that receives better doctors, better equipment, and cleaner buildings every year — but fewer patients — does not need another building. It needs to understand why people no longer trust it.”

That analogy is worth sitting with, because it reframes the entire question. If a hospital kept receiving new equipment and better-qualified staff while patient numbers steadily fell, no sensible administrator would respond by ordering more equipment. They would ask why trust had eroded even as capability improved. Indian government schooling appears to be in exactly that position: two lines on the same chart, infrastructure rising and enrolment falling, moving in opposite directions across the same four years.

04Where the Trend Points, If Unchanged

Projections of this kind are often mistaken for predictions, which they are not. They function more like an estimate of where a car will be after an hour if it holds its current speed — useful for understanding direction, not a claim about the destination. A driver may stop for tea, take a different road, or accelerate; a school system may respond to policy, migration, or a change in parental sentiment in ways no straight line can anticipate. With that caveat firmly attached, a simple linear extension of the last four years’ movement is still informative.

If the present direction were to continue unchanged, the gap between government and private-unaided enrolment could narrow substantially within five years, and private unaided schools could conceivably educate a larger share of India’s children than government schools within a decade. Whether that comes to pass has little to do with arithmetic and everything to do with the decisions made in the interim.

05What UDISE+ Cannot Tell Us

The UDISE+ dataset is a record of enrolment, not of motivation. It can tell us that a child moved from a government to a private school; it cannot tell us why. For that, one has to turn to household surveys such as ASER and the broader literature on school choice, where a handful of recurring themes emerge.

• Parents increasingly associate English-medium instruction with better long-term opportunity for their children.
• Teacher accountability and consistent, visible classroom engagement carry real weight in parental trust.
• School discipline and regular communication with families shape perceptions of quality as much as test scores do.
• Education is increasingly framed by families as a route to upward mobility, not simply a public entitlement.
• Even modestly priced private schools frequently market aspiration more effectively than government schools communicate their own genuine improvement.

The pattern beneath these themes is consistent: parents are not only comparing two sets of classrooms against each other. They are comparing two imagined futures for their child, and choosing the one that feels more legible and more within their control.

06From Hardware to Software

For much of the last two decades, Indian education policy concentrated, understandably, on the hardware of schooling — classrooms, toilets, electricity, computers, internet connectivity, and libraries. Those investments have produced real, measurable gains, several of which are visible in the tables above. The next phase of reform will have to reckon with something less tangible and considerably harder to install: the software of schooling — classroom teaching quality, language confidence, school leadership, sustained parent engagement, student wellbeing, community trust, and outcomes that are visible to families in everyday terms, not only in annual reports.

A newly painted building can draw a parent’s attention for a single visit. A classroom where children read confidently, speak up, and are known individually by their teacher is what earns a family’s continued confidence over years.

Suppose a parent walked into your school tomorrow, unannounced. Would they notice the infrastructure first — or would they remember the warmth of the teachers, the confidence of the children, and the visible joy of learning in the room?

That answer may, in the end, matter more to a school’s future than any single figure in next year’s UDISE+ report.

Data: Unified District Information System for Education Plus (UDISE+), Reports 2022–23 to 2025–26, Department of School Education & Literacy, Ministry of Education, Government of India. Trend projections are illustrative linear extrapolations only and should not be read as forecasts.