On the morning of 25 July 2025, seven children died when the roof of their government upper primary school collapsed in Piplodi village, Jhalawar district, Rajasthan. They were between seven and 13 years old. They had gathered for morning prayers when the ceiling gave way. Students had reportedly noticed falling debris in the days before and alerted teachers, who dismissed the warnings. Ten days before the collapse, the Education Department had directed inspection of all unsafe buildings. Piplodi was certified safe. A day later, another school roof collapsed in Khariyawas village, Nagaur district. Fortunately, this was after school hours and a tragedy was averted.

What Piplodi indicates is a fault line in India’s current public education compact: the system is being redesigned, quietly and without fanfare, around cost reduction and efficiency.

Both these schools had been constructed around 1994 under the District Primary Education Programme (DPEP), financed substantially by World Bank credits, and engineered with an implicit working life of approximately 30 years. Those 30 years were up. Neither building had ever undergone an independent structural assessment. Under procurement rules then in force, independent third-party audits were mandatory only for construction projects above Rs. 30 crore—a threshold rural primary school construction never approached.

Of the 94 students enrolled at Piplodi, 78 were from Scheduled Tribes (STs), five from Scheduled Castes (SCs), and 11 from Other Backward Classes (OBCs). The school was theirs alone. Children of other communities attended a nearby private school.

Six weeks after the collapse, Mor Singh, a 60-year-old farm labourer, donated his two-room home as a temporary school and moved his family to a shelter on farmland. The children of Piplodi returned to class—not in a building sanctioned by the state, but in a donated house made available by someone who understood what lack of access could mean. His gesture should not be romanticised or sought to be replicated. It is not a substitute for the state.

The Rajasthan High Court has ordered 86,934 dilapidated classrooms to be locked. Expert estimates place the cost of full remediation at Rs. 20,000 crore. However, the annual state repair budget is only Rs. 375 crore. At this rate, it would take 54 years.

What Piplodi indicates is a fault line in India’s current public education compact: the system is being redesigned, quietly and without fanfare, around cost reduction and efficiency. And at the heart of that redesign sits a simple instrument—the transport voucher. This marks the boundary between two fundamentally different theories of what the state owes the child, the citizen of tomorrow.

Expansion and its Aftermath

Distance is a structural barrier to educational access. That insight—drawn from decades of work in the field into why Adivasis, Dalits, girls, and children with disabilities stayed out of school—found its legal expression in the Right to Education (RTE) Act in 2009 (Government of India 2009). Section 6 of the RTE Act mandates a primary school within one kilometre of every child’s home, and an upper primary school within three kilometres. Proximity was the mechanism of inclusion.

The neighbourhood school was understood as a shared public institution consistent with the National Curriculum Framework vision of schools as spaces where children learnt to live with difference and practice the habits of democratic citizenship (NCERT 2005). The legislation codified what the DPEP had built into infrastructure from 1994. The DPEP and its successor, the Sarva Shiksha Abhiyan (SSA), brought schools to the doorstep of the systematically excluded, on the premise that if children could not reach education, education would reach them. Between 1994 and 2014, India opened more than 160,000 new primary schools, successfully bringing a first generation of marginalised learners into the system, and in time their children (Purohit 2015).

But expansionary systems are only as durable as their life-cycle planning. The DPEP-SSA building stock was constructed with an implicit working life of approximately 30 years, yet the planning apparatus never institutionalised maintenance. The result was a nationwide cohort of buildings that aged simultaneously, deteriorating across three decades without a parallel system of inspection, repair, or renewal. Construction was systematised. Care was not.

The court acknowledged the RTE Act’s neighbourhood school mandate but ruled that administrative efficiency could override proximity.

By the mid-2010s, the expansion phase was mathematically complete—gross enrolment ratios had touched near-universal levels. At the same time, the 2011 Census recorded a five-million-child decline in the national 0–6 age group. The planning question should have shifted from how to build more to how to sustain what exists. It did not.

What India needed was a second DPEP—not of new construction, but of stewardship. A programme that mapped ageing building stock, commissioned independent structural audits, funded repair, and converted demographic contraction into pedagogical opportunity. The data was visible. The demographic signal was clear. The institutional imagination was absent.

Between 2020-21 and 2024-25, India lost 18,727 government schools, their number declining from 1,032,019 to 1,013,322. In the same period, private unaided schools expanded by 8,475 (Ministry of Education 2025; Mehta 2026). The geography of closure follows predictable patterns: Rajasthan, Uttar Pradesh, Gujarat, Maharashtra—all reducing their public footprint. Only Bihar added government schools, the number increasing by 965.

This is “rationalisation” in action. Under the National Education Policy (NEP) of 2020, states have been encouraged to merge “non-viable” schools—those with low enrolment—into larger institutions (Ministry of Education 2020). The policy calls for consolidation to be carried out “very judiciously”. On the ground, it has become a mechanism for reducing cost without accounting for its consequences.

Uttar Pradesh—the state with India’s largest child population—merged 27,000 primary schools with fewer than 50 students each in 2025. The Allahabad High Court approved the merger, accepting the state’s argument that larger schools offer “better infrastructure and holistic development” (Krishna Kumari v. State of UP 2025). The court acknowledged the RTE Act’s neighbourhood school mandate but ruled that administrative efficiency could override proximity.

Statistically, the educational consequences of increased travel distances are difficult to isolate. But the broader consequences are troubling: Uttar Pradesh reported that 56,000 girls dropped out in 2025 alone, a part of the 99,218 out-of-school children in the state. Across India, 29.8 lakh girls have dropped out over the past five years, comprising nearly half of the 65.7 lakh total out-of-school children (Sehgal 2025). In Gujarat, where 5,000 government schools have shut down, there was a 341% spike in dropouts in 2025, with 1.1 lakh adolescent girls out of school—up from 54,541 the previous year (Sehgal 2025).

Dev Desai, a panchayat member from Patan district outlined what would happen next: “In my own village, after the local school shut down, children have to walk five to six kilometres to get to the nearest school. Parents are conservative and won’t send their daughters to study that far. Once a girl drops out, they wait till she is 16 and marry her off” (Sehgal 2025).

Desai can forecast what will happen in his village and surroundings. The Sustainable Action for Transforming Human Capital-Education (SATH-E) reports, by contrast, record enrolment shifts and satisfaction claims—but not how many children were never enrolled, or what combination of poverty, distance, pandemic loss, and parental hesitancy is contributing to the national figure of 29.8 lakh girls who have dropped out over the past five years. What the voucher design makes clear—and this before the question is even framed—is which theory of state responsibility is operating.

Where the Act saw a barrier, the voucher sees an expense. The distance is the same. What changed is who is asked to cross it.

The RTE Act of 2009 was built to remove distance as a barrier. The national policy that followed in 2020 made distance the mechanism of withdrawal.

Proximity is not eliminated by consolidation—it is redistributed. Families with means purchase proximity, while families with limited resources inherit the journey. For every school the state closes, the market opens alternatives, if you can pay.

This is the consolidation compact in practice: the state reduces its cost by closing the small school; the family that can afford it pays for a private alternative; the family that cannot absorbs the travel burden, the risk, or the dropout. The voucher—an attendance-linked, conditional cash transfer capped at Rs 3,000 per year—does not close this gap. It punishes the very vulnerability it claims to address: a child who misses school due to illness, household work pressures, or unsafe travel conditions receives less money the following month.

Where the Act saw a barrier, the voucher sees an expense. The distance is the same. What changed is who is asked to cross it. The shift was not announced and the child moved from being the reason the system exists—the lodestar of policy and programme—to another variable in the costing sheet. How a system built around proximity arrived at a model built around reimbursement is the question the SATH-E framework answers, without apology: the child who could not reach the school was replaced as the unit of planning by the school deemed too small and too expensive to justify.

Efficiency as Ideology

Behind the consolidation push lies a specific intellectual architecture. In 2017, NITI Aayog launched SATH-E, with BCG and the Piramal Foundation as knowledge partners, to implement “systemic reforms” in Madhya Pradesh, Jharkhand, and Odisha (NITI Aayog 2021).

The approach announced its logic plainly. “The goal of the education system,” the report stated, “is to make each school conducive to delivery of high-quality education by optimising sub-scale schools into fewer better-resourced schools.” The aim of consolidation, it continued, is “to strike a balance between access and quality” (NITI Aayog 2021). The unit of planning had shifted—quietly, without acknowledgment—from the child who needed to reach a school to the institution that needed to be viable. Access was no longer a guarantee to be delivered. It was a variable to be weighed.

The planning arithmetic was equally candid about what it was counting. In Jharkhand, the report noted that a typical gram panchayat of 7,000–8,000 people had roughly 1,500–2,000 students. “If even 20% of these students go to private schools,” it observed, “one does not need more than five schools to accommodate the students” (NITI Aayog 2021). The children who had already left the government system — for whatever reason, under whatever pressure — were folded into the consolidation logic as a planning assumption. Their absence was not a failure to be addressed. It was a reduction in the denominator.

In Madhya Pradesh and Jharkhand, the SATH-E executed ambitious merger programmes. The results were at once impressive and numbing. The granular scale of this consolidation is detailed in Table 1 (NITI Aayog 2021, 2023).

Table 1: SATH-E School Consolidation: Scale and Outcomes by State

The SATH-E playbook became the template for rationalisation nationwide. But the NITI Aayog’s own reports acknowledge what the efficiency calculus omits. “Community backlash,” they note, could arise from “students inconvenienced if they must travel longer distances.” The mitigation strategy is not to adjust the model, but to “control the narrative so that the benefits of mergers and its impact on learning outcomes are highlighted, as opposed to negative connotations of shutting down schools” (NITI Aayog 2023). Narrative management is what is offered in place of impact assessment.

The voucher actually transfers the burden of the journey from the state to the household and the risk of non-attendance from a policy failure to a family one.

The NEP 2020 absorbed the SATH-E logic without naming it, without debating it, and without asking the question the demographic data required: if the child population is contracting across most of the country, what form should the right to education take in a landscape of fewer, more unevenly distributed children?

When SATH-E closes a village school, it leaves behind a voucher. The instrument is designed to look like provision, but as Table 2 illustrates, it actually transfers the burden of the journey from the state to the household and the risk of non-attendance from a policy failure to a family one (NITI Aayog 2021; Government of Rajasthan n.d.).

Table 2: The Cost Shift: Sub-scale School Model vs Consolidated School plus Voucher

The voucher assumes the household has already decided to send the child and provides a direct benefit transfer—transport reimbursement. Only, reimbursement is not access. It does not reach the family that pulls a daughter out when the kachcha road floods, when the consolidated school sits in an upper-caste quarter, or when the monsoon makes river crossings impossible. Because the voucher is attendance-linked, absence is penalised. The distance, which was once a barrier the state removed, is now a cost the household must factor in while making decisions.

This logic is not new. In 2001, facing a similar demographic shift, China closed more than 300,000 rural primary schools on the same efficiency rationale. The consequence: girls’ educational attainment fell sharply, ethnic minority children lost mother-tongue instruction, and rural families fled their villages, accelerating the very demographic contraction used to justify the closures. By 2012, the Chinese State Council halted consolidation, mandating the active maintenance of small-scale village schools (Hannum et al. 2021; General Office of the State Council 2012).

The SATH-E pilot, begun in 2017, had this evidence available and chose not to engage with it. What was designed as the guarantee of a right has become the instrument of its withdrawal.

The Southern Alternative

In June 2026, Karnataka expanded its free student bus pass programme from girls to all school-going children. At a time when many states are debating school consolidation and transport vouchers, the decision is notable not only because it is unique, but also because it reflects a longer tradition. Across the region, shrinking child populations did not automatically produce a retreat from educational access. Instead, distance was treated as a collective responsibility rather than an individual burden.

Kerala has operated student bus concessions since 1963, providing passes at roughly 20% of the regular fare, with free travel available to students through higher secondary education (Narayana 2011). Andhra Pradesh was providing free travel concessions to schoolchildren by the mid-1980s. After bifurcation, both successor states continued the scheme (Narayana 2011). Tamil Nadu provides fare-free bus passes to school and college students on government buses as a matter of entitlement, allocating Rs 1,782 crore in 2025-26 to benefit approximately 28 lakh students (Metropolitan Transport Corporation n.d.). Despite differences in design and administration, these schemes rest on a common principle: when location impedes children’s participation, the state assumes responsibility for helping children make the journey.

Under the voucher model, the household must first organise the journey and then seek reimbursement. Responsibility for access shifts from the state to the family.

This is not the SATH-E voucher. It is public infrastructure, state-arranged and publicly accountable. The critical distinction is not the money—it is who is responsible for the journey: the state or the household. Under the southern model, the state retains the responsibility for route planning, vehicle provision, safety, and frequency. If the bus does not run, there is a government agency answerable for that failure. Under the voucher model, the household must first organise the journey and then seek reimbursement. Responsibility for access shifts from the state to the family.

The Real Choice

The NEP 2020 does not distinguish between the planning requirements of Tamil Nadu—where the population of the 0–6 age group has fallen for two consecutive census decades and the total fertility rate (TFR) stands at 1.4—and Bihar, where fertility remains above the replacement level and the child population is still growing. The next Census will make the landscape visible with greater precision. It will show where children are, and where they are not. What it will not do is decide what comes next. That decision is institutional. It requires choosing whether contraction will be treated as an opportunity to deepen the promise of education, or managed as a problem of reducing cost.

The bus and the voucher are not two versions of the same policy. They are two different understandings of the relationship of the child with the state. One treats education as a service the state subsidises. The other treats it as a right the state guarantees. The NEP 2020 chose the voucher. The southern states chose the bus. The southern states demonstrate that demographic contraction does not inevitably lead to educational withdrawal. Their experience shows that distance can still be treated as a public responsibility, for proximity is a choice.

Piplodi has already shown the cost of not choosing. The next phase of India’s schooling system will not be defined by how many schools it builds, or by how efficiently it consolidates them, but by how it reorganises what it already has in a landscape where children are fewer, more unevenly distributed, and differently placed within society. The census is counting. The Republic must decide whether the child walks to school, rides a bus the state provides, or stays home because the voucher could not buy safe passage across a flooded road. A policy instrument that pays less when the child most needs to travel is not solving the problem of access. It is pricing it.

The Kothari Commission had proposed a pedagogic standard, 20 children to a teacher. On fiscal grounds, that was diluted to 30 in the National Education Policy in 1986 and has remained unchallenged, despite the child population shrinking. Finland, whose population of 0–4-year-olds declined by 22% between 2013 and 2023 (OECD 2025), educates 13 students per teacher in primary, and nine in lower secondary (OECD 2024). The Republic, without debate, has chosen longer journeys over smaller class sizes.

Small versus large schools, efficiency versus sentiment, quality versus access. The proximity debate is about none of these. This is a larger debate on who bears the responsibility for the journey. For three generations, that responsibility belonged to the state. The voucher shifts it to the household, without debate. Before more schools are closed, and that inheritance is squandered, the shift deserves to be debated.

Vijayalakshmi Balakrishnan (baroquepodcast@gmail.com) is the author of Growing Up and Away: Narratives of Indian Childhoods (Oxford University Press, 2011).