The Indian education system: an independent audit

How to read this report

This report separates four things that public debate constantly mixes up:

1. Structural failure — problems of money, teachers, infrastructure and assessment that exist regardless of who governs. This is the largest part of the problem and gets the least airtime.
2. Verified factual errors and pseudoscience — claims taught or promoted that are demonstrably false.
3. Selective omission and reframing — where content has been removed or rewritten in a pattern that points in a consistent political direction. Documented, but requires judgement to interpret.
4. Contested interpretation — where historians and educationists genuinely disagree, and where calling something "brainwashing" is a political claim, not a factual one.

The brief asked specifically about ideological content introduced under the BJP-led central government. Section 4 addresses that in detail with named chapters and dates. Section 5 deliberately documents the same behaviour by non-BJP governments and by the pre-2014 Congress-led centre. This is not false balance — it is the analytical basis for the central recommendation of this report, which is that India's curriculum problem is institutional, not partisan. Any fix that only removes one party's content and installs another's will be reversed at the next election, as it was in 2004 and again after 2014.

---

PART 1 — THE SYSTEM BY THE NUMBERS

1.1 What India has genuinely achieved

Honest assessment must start here, because the failures only make sense against the scale of the success.

• Near-universal enrolment at primary level. India educates roughly 233 million school students across about 1.47 million schools.
• Teacher workforce crossed 1 crore for the first time in 2024–25 — 1,01,22,420 teachers, up 6.7% from 2022–23. About 54% are women.
• Pupil-teacher ratios are, on paper, within norms. UDISE+ 2024–25 reports 10 at foundational, 13 preparatory, 17 middle, 21 secondary — all inside the NEP's 30:1 benchmark.
• Dropout rates fell sharply in 2024–25 — preparatory 3.7% → 2.3%, middle 5.2% → 3.5%, secondary 10.9% → 8.2%.
• Higher education Gross Enrolment Ratio rose from 23.7% (2014–15) to 30.0% (2023–24), with total enrolment around 4.5 crore. SC enrolment ratio rose from 18.9 to 27.8, ST from 13.5 to 22.8, and female GER (31.2) now exceeds male.
• Single-teacher schools declined about 12% in one year to 1,04,125, and zero-enrolment schools fell about 38%.

This is real. Access has been substantially solved. The problem is what happens after the child is inside the building.

1.2 The learning crisis — the actual emergency

India has built an education system that reliably produces attendance and unreliably produces learning.

ASER (Annual Status of Education Report), the most credible independent household survey, run by Pratham across roughly 600 rural districts:

• ASER 2022: only 20.5% of Class 3 children could read a Class 2 text — below the 2012 level of 21.4%. Basic arithmetic: 25.9%.
• ASER 2023 (ages 14–18): 26% could not read a Class 2 text in their own language. 42.7% could not read simple English sentences. 56.7% could not do basic division. 34.7% could not do basic subtraction.
• ASER 2024 recorded the largest reading gains since 2022, credited to the NIPUN Bharat foundational literacy mission — but a majority of Class 3 students still cannot read a Class 2 text.

Read that last line carefully. After a decade of the strongest foundational-literacy push India has ever mounted, the median Class 3 child in rural India is functionally pre-literate by the standard of the previous year's textbook.

What this means downstream: a child who cannot decode text in Class 3 cannot learn science in Class 6, cannot pass a board exam honestly in Class 10, and enters the labour market with a certificate that does not correspond to a capability. This is the root cause of the employability numbers in Section 1.6, and no curriculum reform — of any ideological flavour — matters until it is fixed.

1.3 Teachers: qualification, deployment, and training

The teacher numbers look good in aggregate and fall apart on disaggregation.

Deployment failure. In 2023–24, roughly 40 lakh students were in about 1.11 lakh single-teacher schools — one adult teaching all grades and all subjects. Concentrations were worst in Madhya Pradesh (13,198), Andhra Pradesh (12,611), Uttar Pradesh (8,866), Jharkhand (8,353) and Maharashtra (8,196). Simultaneously, 31,981 teachers were posted in 12,964 schools with zero students. Both figures improved in 2024–25 (1,04,125 single-teacher schools; 20,817 teachers in empty schools) but the coexistence of the two tells you the core problem is not teacher supply. It is rational deployment, and it is a political problem — transfers, postings and rural hardship allowances are patronage instruments in most states.

Ratio failure at the top of the school. The national PTR average hides higher-secondary PTRs of 47 in Jharkhand, 37 in Maharashtra and Odisha, 35 in Uttar Pradesh — all above the NEP limit — while Chhattisgarh sits at 12 and Madhya Pradesh at 15. This is not a national shortage; it is a distribution collapse.

Quality of the pipeline. The structural issues, well documented and largely unaddressed:

• B.Ed and D.El.Ed degree mills. A large share of teacher-training colleges are private, non-teaching in practice, and sell credentials. The Justice J.S. Verma Commission flagged this in 2012; the NCTE's regulatory capacity has not caught up. NEP 2020's response — a four-year integrated B.Ed as the minimum qualification by 2030 — is directionally right but its rollout has been slow and the closure of substandard institutions has not happened at scale.
• DIETs (District Institutes of Education and Training) — the in-service training backbone — are chronically understaffed and in many states function as parking posts.
• In-service training is delivered as cascade model workshops (train the trainer, who trains the trainer), with well-known signal loss at each stage, and is almost never evaluated against classroom practice or student outcomes.
• Contract and para-teachers (Shiksha Mitras, Vidya Volunteers, guest lecturers) remain a significant share of the workforce in several states, on a fraction of regular pay, with weaker qualification requirements and no career path.
• Non-teaching duties. Teachers are routinely deployed for elections, census, mid-day meal administration, and now extensive app-based data entry. Instructional time loss is significant and unmeasured.

Higher education faculty. Vacancy rates in central universities were near 28% in 2021–22. A Rajya Sabha standing committee found 34,099 vacant posts across central institutions (IITs, NITs, IIMs, IISERs, central universities) in 2022–23, of which under 20% were filled that year. More recently over 5,000 teaching vacancies and about 2,000 unfilled reserved-category posts persist in central universities. Reserved-category posts going unfilled year after year, while being counted as "advertised," is a distinct and serious equity failure.

1.4 Money: the sixty-year unmet promise

The Kothari Commission recommended 6% of GNP for education in 1964–66. The National Policy on Education 1986 reiterated it. NEP 2020 reiterated it again.

• Combined centre + state education spending: around 2.7% of GDP (2023–24 Economic Survey), with other estimates of combined expenditure in the 4.1–4.6% range depending on what is counted.
• Union Budget 2026–27: ₹1.39 lakh crore to the Ministry of Education — a record in nominal terms, up 8.27% from ₹1.28 lakh crore. School Education & Literacy: ₹83,562 crore (+6.35%). Higher Education: ₹55,727 crore (+11.28%).
• That ₹1.39 lakh crore is about 2.6% of the total Union budget outlay of ₹53.5 lakh crore, and roughly 0.36% of GDP — down from around 3% of the Union budget in 2019–20.
• School education's 6.35% nominal growth is at or below inflation, meaning real-terms stagnation.

Two honest observations:

1. Education is a Concurrent List subject and states carry the bulk of the spend. Criticising only the Union figure is partly unfair.
2. But the Union sets national policy targets — GER of 50% by 2035, universal secondary education by 2030, a four-year B.Ed for every teacher — and then funds them at 0.36% of GDP while asking states to close the gap. That is an accountability mismatch, and it is compounded by the fund-withholding described in Section 6.

1.5 Assessment: what the system actually rewards

This is the least discussed and arguably most damaging failure, and it is genuinely bipartisan.

• Board examinations reward recall, not reasoning. Question papers are predictable across years; the optimal student strategy is memorising previous papers and guide-book answers. Teachers respond rationally by teaching to this.
• The coaching economy has become the real curriculum. For NEET and JEE, the school is a formality and the coaching centre is the instruction. Kota alone has drawn lakhs of aspirants annually; student suicide numbers there and elsewhere have made this a public-health issue, not merely an educational one.
• Examination integrity has degraded. The NEET-UG 2024 paper leak, which forced a Supreme Court hearing and a re-examination for a subset of candidates, and the cancellation of the UGC-NET, exposed the NTA's operational fragility at a moment when the entire high-stakes system depends on it.
• Marks inflation. Board score distributions have compressed at the top to the point where 95%+ aggregates no longer discriminate, pushing selection pressure entirely onto entrance tests and cut-offs.
• PARAKH — the national assessment body under NCERT — is the right institutional idea (comparable, rigorous, nationally benchmarked learning data). Its Rashtriya Sarvekshan and the Foundational Learning Study 2026 covering 1 lakh Class 3 students across 10,000 schools are steps forward. Its independence from the ministry that it is assessing is the open question.

1.6 Higher education and the employability gap

• GER 30.0% (2023–24), against an NEP target of 50% by 2035. That requires an extremely steep trajectory that current funding does not support.
• Attrition is the hidden crisis. NUEPA estimates student attrition in higher education at around 25%. Enrolment is measured obsessively; completion is not.
• India Skills Report 2025 puts national employability at 54.81%. The State of Working India 2026 reports graduate unemployment near 40% among under-25s and around 20% for ages 25–29. India is producing unemployable graduates at scale — the clearest possible evidence that certification and capability have decoupled.
• Quality at the top. India has the highest volume of ranked institutions in the THE Asia University Rankings 2026 (128 universities) but none in the top 40. Breadth without peaks.
• Roughly 1.33 million Indian students went abroad in 2024 — a very large export of tuition fees and, frequently, of talent.
• R&D spending remains around 0.65% of GDP, far below the 2%+ of competitor economies, with the private sector contributing an unusually small share.

1.7 Infrastructure and the digital gap

Improvements are real — electrification, drinking water, girls' toilets, internet-enabled schools all rose. What persists:

• Functional-versus-existing gap. A toilet reported in UDISE+ is not necessarily a usable toilet. There is no independent physical audit.
• Science laboratories and libraries remain absent or nominal in a large share of government secondary schools.
• The digital push under NEP assumes device and bandwidth access that ASER's own data shows is unevenly distributed, especially for girls, who have lower independent smartphone access than boys in the same households.
• Self-reported data. UDISE+ and AISHE are both self-reported by institutions with only built-in validation. AISHE participation is voluntary. Every number in this report drawn from those sources should be read as an upper bound on performance.

---

PART 2 — HOW SCIENTIFIC AND RATIONAL IS WHAT IS TAUGHT?

2.1 Pedagogy

Indian classrooms remain overwhelmingly transmission-based: teacher speaks, student copies, student reproduces. The NCF-SE 2023 and the new NCERT textbooks explicitly try to change this — the Class 6–8 books emphasise inquiry, application and "what a student can do" rather than recall. This is a genuine improvement in design intent.

The gap is delivery. A teacher trained by rote, in a class of 40, assessed on board results, with no lesson-planning time, will teach by rote regardless of what the textbook asks for. Curriculum reform without teacher-capacity reform is a paper exercise. This is the single most predictable reason the NCF 2023 will underdeliver.

2.2 The 2022–23 science deletions

This is where the "is it scientific?" question becomes concrete. Under the "rationalisation" exercise announced for 2022–23, NCERT removed from Class 10 science, among other content:

• The periodic table (moved to Class 11)
• The chapter on evolution / heredity content (Darwin, natural selection — retained at Class 12)
• Sources of energy; sustainable management of natural resources
• Faraday's laws content and related electromagnetism material

NCERT's defence, given by Director Dinesh Prasad Saklani: the content was rationalised because of overlap with the same material in other subjects or higher classes, and to reduce academic load during COVID disruption. On the narrow point, this is accurate — the periodic table does appear in Class 11 chemistry and evolution in Class 12 biology.

Why the defence fails on the evidence. The relevant question is not whether the content exists somewhere in the system. It is what proportion of Indian children ever reach that class in that stream. A large majority of students never take Class 11 science. Removing evolution and the periodic table from Class 10 therefore means most Indians will complete their entire formal education without encountering the organising theory of biology or the organising table of chemistry. More than 1,800 scientists and educators, including from IISc, IITs and IISERs, signed an open letter against the evolution deletion on precisely this ground. Breakthrough Science Society and the All India People's Science Network made the same argument.

This is a clear, evidence-based failure. Whether it was ideologically motivated is arguable — evolution is a specific target for religious conservatives in many countries, and its removal alongside a broader pattern is suggestive — but the argument does not require establishing motive. The deletion is indefensible on educational grounds alone.

2.3 Indian Knowledge Systems: the legitimate and the pseudoscientific

This requires precision, because both the promoters and the critics overstate.

The legitimate case. India's historical contributions to mathematics (the decimal place-value system, zero as a number, Brahmagupta, Bhaskara, the Kerala school's infinite series work predating European calculus by roughly two centuries), astronomy (Aryabhata's heliocentric elements and accurate sidereal calculations), metallurgy (the Delhi iron pillar, wootz steel), linguistics (Panini's generative grammar, genuinely foundational to modern formal linguistics), and medicine (Sushruta's surgical corpus) are historically documented and under-taught. Teaching them is not chauvinism; it is accurate history of science. UGC's IKS guidelines formally direct institutions to use authentic sources — inscriptions, historical accounts — and explicitly warn against exaggeration.

Where it has gone wrong. The implementation has, in specific documented cases, provided cover for pseudoscience:

The other IKS critique, which is separate and important. The Wire's analysis and the AIPSN statement both argue that "Indian Knowledge System" (singular) is itself a distortion: India had many knowledge traditions, frequently in conflict with each other. The UGC curriculum draws overwhelmingly from Sanskrit, upper-caste, Brahminical sources — Vedas, Upanishads, Gita, Puranas, the epics — while largely omitting Buddhist and Jain epistemology, the Charvaka/Lokayata materialist tradition (India's own indigenous rationalist school, which is an extraordinary omission in a programme claiming to teach Indian thought), Phule, Ambedkar, Adivasi ecological knowledge, and the living technical knowledge of artisan and craft communities. AIPSN specifically criticised the framing that treats science as the pursuit of a few learned men of one social group.

The consultant's assessment: IKS is a legitimate idea being implemented without epistemic quality control. The fix is not abolition. It is (a) a standing scientific review board with authority to veto content, (b) a hard firewall between history-of-science content and clinical/technical curricula, and (c) mandatory inclusion of heterodox and non-Brahminical traditions.

---

PART 3 — DOCUMENTED IDEOLOGICAL CHANGES TO THE CURRICULUM

This section deals with what the brief specifically asked about. It is organised by what changed, what NCERT said, and what the historical evidence supports — because on several of these the honest answer is more complicated than either side's public position.

3.1 Timeline

• 2022–23: "Rationalisation" of existing NCERT textbooks, publicly justified by COVID learning loss. This is when most of the deletions occurred.
• 2023: NCF-SE 2023 finalised under NEP 2020.
• 2023–2026: Entirely new textbooks written under NCF-SE 2023, introduced class by class — Class 6 (2023), Class 7 (2024), Class 8 (2025–26). These are not edits; they are new books with new architecture.
• August 2025: Following sustained criticism of the Class 8 book, NCERT constituted a high-level committee to review feedback and make "evidence-based decisions."

3.2 Category A — Deletions (2022–23)

NCERT's position: all of this was COVID load-reduction and overlap removal, with no political intent. Director Saklani stated in April 2023 that reports of Mughal chapters being dropped were false and the debate unnecessary — a claim that requires careful parsing, since the Class 12 Mughal chronicles chapter had in fact been removed while other Mughal content remained.

The problem with the defence: the Emergency and the Gujarat riots do not overlap with content in other classes. Neither do the Faiz poems, the RSS ban, or "Democracy and Diversity." A load-reduction exercise that lands disproportionately on (i) Muslim political history, (ii) episodes reflecting badly on Hindu nationalist organisations, (iii) caste as an analytical category, and (iv) protest as a democratic act is not distributed the way random load-reduction would be. NCERT also did not initially notify some of these deletions, later saying this was to avoid confusion — which is the weakest of the responses, because unannounced deletion is the definitional opposite of transparent curriculum governance.

What can be honestly concluded: the pattern is clear and the burden of explanation sits with NCERT, which has not discharged it. Whether each individual deletion was ideologically motivated cannot be established item by item. The removal of the Emergency, notably, cuts against a simple partisan reading — it is the single most damaging historical episode for the Congress party, and its removal is genuinely puzzling on any theory. This is worth acknowledging, because critics who ignore it weaken their own case.

3.3 Category B — Reframing in the new NCF-SE 2023 textbooks

The Class 8 Exploring Society: India and Beyond (2025–26) is the most significant. The Delhi Sultanate, Mughals and Marathas were moved from Class 7 to Class 8 and consolidated into a chapter called "Reshaping India's Political Map" under Theme B, "Tapestry of the Past."

What the book says:

• Babur — a brutal and ruthless conqueror who slaughtered entire city populations, enslaved women and children, and erected towers of skulls; also drawing on the Baburnama to note his curiosity and appreciation of poetry and nature.
• Akbar — a reign described as a blend of brutality and tolerance, including the massacre of some 30,000 civilians after the siege of Chittorgarh, alongside his later inclusive policies.
• Aurangzeb — the destruction of temples and gurdwaras at Banaras, Mathura, Somnath and elsewhere, presented as religious intolerance; the book also notes the argument that his actions were politically motivated.
• Jizya — reframed as a discriminatory levy rather than, as previously, a tax collected alongside land revenue.
• Malik Kafur's campaigns against southern temple centres are highlighted.
• The period is characterised as one of political turmoil, warfare, and desecration of temples and seats of learning.
• The book opens with "A Note on Some Darker Periods in History," which states the rationale for including violence, and carries a disclaimer that no one today should be held responsible for events of the past.

NCERT's defence, from Michel Danino, head of the Curricular Area Group for Social Science: Indian history cannot be sanitised into a smooth happy progression; there were bright periods and dark ones; the aim is not to demonise Akbar or Aurangzeb but to show that these rulers had limitations and committed cruel acts. Akbar himself, Danino noted, admitted brutality in his youth. The stated purpose is that understanding the origin of cruelty is how a society heals.

Assessment — and this is where care matters.

The individual factual claims are largely defensible. Babur's memoirs do record towers of skulls. The Chittorgarh massacre happened and Akbar ordered it. Aurangzeb did order temple destructions, including Kashi Vishwanath and Keshavdev at Mathura, and did execute Guru Tegh Bahadur. Jizya was a tax levied on non-Muslims and its discriminatory character is not a Hindutva invention. The older textbooks' treatment of Akbar as straightforwardly "the Great" and jizya as a routine revenue measure was itself a simplification, and the criticism that Indian textbooks had softened medieval violence to preserve communal harmony has been made by serious historians, not only by the right.

The problems are structural, not factual:

1. Asymmetry of scrutiny. The same book covers the Marathas and Sikhs as resistance narratives. Maratha campaigns in Bengal (the bargi raids of the 1740s), Vijayanagara's own temple destructions in conquered territory, and Chola raids on Southeast Asian and rival Indian temples receive nothing comparable. Applying an unflinching lens to one dynasty and a celebratory lens to others is not "unsanitised history." It is a differently sanitised history.

2. The Class 7 gap. By moving this material to Class 8, the previous Class 7 sequence now runs from the Mauryas and Shungas to the Guptas and stops. A cohort of children encounters an Indian history in which the Guptas are followed by nothing for a year.

3. The disclaimer is self-refuting in a classroom. Telling thirteen-year-olds that Babur built towers of skulls and Aurangzeb demolished their temples, then adding a paragraph that no present community bears responsibility, misunderstands how adolescents process narrative. The affective content of the chapter is not neutralised by a note. If the pedagogical intent were genuinely dispassionate historical maturity, the appropriate device is comparative treatment across all dynasties, not a disclaimer.

4. The reception tells you the function. The book was received publicly as the moment when, in the words of one widely circulated reaction, there would be "no more Akbar the Great." A textbook's effect is what it produces in the polity, not what its preface says.

5. A documented error: the Class 8 book incorrectly mapped Jaisalmer within the Maratha Empire. This is a straightforward accuracy failure, and it is the sort of thing an independent review board would have caught.

3.4 Category C — Other documented changes across recent editions

• Babri Masjid described in Class 12 Political Science as a "three-domed structure," with the account of the demolition, the subsequent communal violence, President's rule in BJP-ruled states, and the BJP's own stated regret compressed or removed. Describing a named 16th-century mosque by its architectural profile is a deliberate act of denomination, not a neutral one.
• Manipur: revisions to material on the state's merger with India.
• Harappan Civilisation: increased emphasis on Saraswati-river framing and on continuity with later Vedic culture, and reduced emphasis on the Aryan migration model. On the underlying history: the 2018–19 Rakhigarhi ancient-DNA study and the broader South Asian aDNA literature (Narasimhan et al., Science, 2019) support significant Steppe-pastoralist-derived ancestry entering South Asia in the second millennium BCE. Textbook framing that treats migration as discredited is not aligned with current genetic evidence. This is a case where the science is comparatively clear and the curriculum has drifted from it.
• Caste: reduced treatment across social science, both through the "Democracy and Diversity" deletion and through the IKS curriculum's silence on anti-caste traditions. For a country where caste remains the primary axis of material inequality, teaching less about it is a substantive analytical loss regardless of intent.
• Special modules issued by NCERT on contemporary events, including on military operations, have drawn criticism for bringing near-current political events into supplementary school material before any historical settling.

3.5 What the assessment adds up to

Stating this plainly, since the brief asked for a clear finding:

There is a documented, consistent pattern of curriculum revision under the current central government that (a) reduces the historical visibility of India's Muslim past, (b) removes episodes damaging to Hindu-nationalist organisations, (c) reduces caste as an analytical category, (d) reduces the teaching of protest and dissent as democratic practice, and (e) in the science curriculum, removed evolution and the periodic table from the terminal year most students complete.

At the same time: several individual changes are historically defensible; NCERT's stated rationale is not always false; the Emergency deletion complicates a simple partisan reading; and the word "brainwashing" is a rhetorical characterisation rather than an analytical one. What is happening is better described as curriculum capture — the use of a state monopoly on textbooks to shift the frame of national memory. That is a precise, defensible charge, and it is more damaging to defend against than the imprecise one.

---

PART 4 — WHY THIS IS AN INSTITUTIONAL PROBLEM, NOT A PARTISAN ONE

Any report that stops at Part 3 is politically useful and analytically incomplete. The following is documented history.

1977–79 (Janata government): Ministry attempts to withdraw NCERT textbooks by Romila Thapar, Bipan Chandra and others on the grounds of anti-national and pro-Marxist content.

1998–2004 (NDA-I, HRD Minister Murali Manohar Joshi): The first systematic "saffronisation" episode. Textbooks rewritten; the UGC issued a notification in February 2001 inviting universities to propose courses in Jyotir Vigyan (Vedic astrology) — the origin of the astrology-in-universities problem that continues today. Astrology as a university subject was challenged in the Supreme Court, which declined to strike it down.

2004–2014 (UPA): Comprehensive reversal. The NCF 2005 and a new textbook set replaced the NDA-era books wholesale. The new books were genuinely better by most pedagogical measures — but the mechanism was identical: a change of government produced a change of national historical narrative, executed through a body reporting to the ministry.

Kerala (LDF/CPI(M)): In 2023 the state government produced supplementary textbooks restoring the NCERT deletions — Gandhi's assassination, the Gujarat riots, India under Nehru — and made them examinable. On the merits of restoring deleted content this is defensible. But Kerala has its own record: the 2008 Class 7 social science textbook controversy over the "Mathamillatha Jeevan" chapter, which triggered mass protests and was withdrawn, and a broader pattern in which the state's curriculum reflects Left historiography with the same confidence that NCERT now reflects a different one.

Tamil Nadu: Dravidian-movement historiography is embedded in state textbooks. The Aryan-invasion framing, Sanskrit's role, and the characterisation of North Indian political influence are handled in ways that would be contested by historians outside that tradition.

Karnataka: Textbook revision committees were reconstituted and content reversed with the change of government in 2023, and again contested.

Rajasthan, Maharashtra, Gujarat: Repeated revisions to content on Savarkar, Nehru, Gandhi, Tipu Sultan and regional figures tracking changes in state government.

The conclusion this forces: India has no institutional protection for curriculum from the executive of the day. NCERT's director is a government appointment. Curriculum area groups are constituted by the ministry. There is no statutory public consultation requirement, no mandated disclosure of deletions (as the unannounced 2022 deletions demonstrated), no independent historical or scientific review with veto power, and no fixed revision cycle insulated from electoral cycles.

Every party has used this vulnerability when in power and objected to it when out of power. A reform proposal that treats the problem as "the BJP's textbooks" will produce another reversal in 2029 or 2034. A reform proposal that treats it as "the ministry's unchecked control over national memory" can, in principle, attract cross-party support because every party has been on the losing side of it.

---

PART 5 — FEDERALISM AND GOVERNANCE

Education is on the Concurrent List. The current mechanism of central-state coordination has broken down in a way that directly harms children.

The PM SHRI dispute. The Union government has linked release of Samagra Shiksha Abhiyan funds — the core scheme financing government school education — to states signing the PM SHRI (PM Schools for Rising India) memorandum, which embeds NEP 2020 implementation including the three-language formula.

• Kerala: ₹953 crore withheld across 2023–24 and 2024–25 as of December 2024, later cited at higher figures, affecting by the state's account around 40 lakh students from marginalised communities.
• Tamil Nadu: ₹2,151 crore withheld for 2024–25; the state moved the Supreme Court in May 2025, which declined urgent hearing in June 2025. By May 2026 the figure owed across 2024–25 and 2025–26 exceeded ₹3,500 crore, and the state government confirmed it would maintain its two-language policy and not sign.
• West Bengal: also withheld, in part over branding requirements — schools must carry the PM SHRI prefix even though states bear 40% of the cost and assume the schools after five years.
• As of February 2025, zero SSA funds had been disbursed to Kerala, Tamil Nadu or West Bengal. Punjab initially refused and subsequently signed under fiscal pressure.

Assessment. Set aside the merits of the three-language formula, on which reasonable people differ. The mechanism is the problem: a general-purpose school funding stream, to which states have a claim, is being used as leverage to compel adoption of a separate discretionary scheme. The children who lose teaching-learning materials, bridge courses and inclusive-education support in Chennai and Kozhikode are not parties to the dispute. This is the clearest single governance failure in Indian school education today and it is causally attributable, not a matter of interpretation.

It is also self-defeating for the centre. It converts the NEP — which contains genuinely good ideas on foundational literacy, flexible streams and vocational integration — into a coercion symbol, guaranteeing resistance in exactly the states with the strongest public education systems.

---

PART 6 — RECOMMENDATIONS

Ordered by expected impact per rupee, not by how much attention each gets.

Tier 1 — Highest impact

1. Treat foundational literacy and numeracy as a national emergency with a hard deadline and independent measurement. Extend and fully fund the FLN mission (the Rajya Sabha standing committee's 363rd report already recommends extension to 2032). Every child reading at grade level by Class 3, measured by a body that does not report to the ministry being assessed. Nothing else in this report matters if this fails.

2. Fix teacher deployment before hiring more teachers. - Publish a school-level teacher-requirement map against enrolment and rationalise within 24 months. Zero teachers in zero-enrolment schools is achievable now. - Eliminate single-teacher schools by consolidation with guaranteed transport, or by mandatory second posting — not by closing schools and leaving children with no option. - Make rural and remote postings career-advantageous rather than punitive: hardship pay, promotion weighting, fixed tenure, housing. - Remove teachers from non-education duties by statute, with a defined exception list.

3. Rebuild the teacher pipeline. - Close substandard B.Ed and D.El.Ed institutions. This requires NCTE to be given inspection capacity and political cover; it will be resisted because these institutions are locally owned and politically connected. - Staff and fund DIETs properly; make them accountable for district learning outcomes. - Replace cascade training with school-based mentoring and subject-specific coaching, evaluated against classroom observation. - Regularise para-teachers with a qualification pathway rather than perpetuating a two-tier workforce.

4. Move to 6% of GDP with a legislated glide path. A record nominal allocation that is 0.36% of GDP and below inflation in real terms is not a serious answer to the numbers in Part 1. A statutory multi-year commitment with annual reporting to Parliament, disaggregated by centre and state, is the minimum credible mechanism.

Tier 2 — Curriculum integrity

5. Establish a statutory, independent National Curriculum Authority. This is the central structural recommendation. Design principles: - Fixed-term appointments for members, staggered so no single government appoints a majority. - Composition including working historians, practising scientists, teachers, and state SCERT nominees — with formal representation for opposition-ruled states. - Mandatory public consultation with a defined comment period before any curricular change. - Statutory duty to publish every deletion, addition and reframing in a public change log with reasons. The 2022 unannounced deletions would have been impossible under such a rule. - A scientific review board with veto power over any content contradicting established scientific consensus, and a historical review panel with the same power over content contradicting the documentary and archaeological record. - A fixed revision cycle — for example, ten years, with an emergency amendment procedure requiring supermajority.

This should be pitched as protection against the next government, whoever that is. That is its honest selling point and its only realistic path to cross-party support.

6. Restore evolution and the periodic table to Class 10. On educational grounds. No further argument is needed.

7. Apply a single evidentiary standard across all of history. If Aurangzeb's temple destructions are taught, so are Maratha raids in Bengal, Vijayanagara's conquests, Chola expansionism, and the caste violence of every period. If disclaimers about present-day responsibility are used, they are used uniformly. Comparative rigour is the only defensible answer to charges of bias from any direction — and it is genuinely better history.

8. Split IKS into two streams with a firewall. - History and philosophy of Indian science — taught with full rigour, including Charvaka materialism, Buddhist and Jain logic, anti-caste thought, Adivasi ecological knowledge, and artisan traditions. This is valuable and underdone. - Practices without evidentiary support — astrology, Vastu — may be studied anthropologically as belief systems. They may not be credentialed as vocational or scientific disciplines, and they may not enter clinical curricula. Withdraw astrology from the Ayurveda curriculum and from medical training pathways immediately; the risk there is to patients.

Tier 3 — Assessment and higher education

9. Reform board examinations toward application. Publish question-design frameworks; mandate a rising share of unseen-passage, data-interpretation and multi-step reasoning items; audit papers annually against the framework. This is the highest-leverage way to change classroom practice, because teachers optimise for the exam regardless of what the curriculum says.

10. Rebuild examination integrity. Independent audit of the NTA, published security protocols, statutory timelines and compensation for candidates affected by cancellations.

11. Fill higher education vacancies on a statutory clock. Publish vacancy data quarterly by institution and category. Reserved posts that remain unfilled after two advertisement cycles should trigger mandatory reporting to Parliament with reasons.

12. Measure completion, not just enrolment. Publish degree-completion, time-to-degree and graduate-outcome data by institution. The 25% attrition estimate should be a headline national statistic, not a footnote in a research paper.

13. Independent audit of self-reported data. UDISE+ and AISHE should be subject to third-party physical verification on a random sample of institutions annually, with results published. Policy built on unverified self-reporting is policy built on sand.

Tier 4 — Federal relations

14. Delink Samagra Shiksha from PM SHRI, by rule. Core school funding to which states have an entitlement should not be conditional on adopting discretionary schemes. If the centre wishes to incentivise PM SHRI, it should fund PM SHRI. Resolve the Kerala, Tamil Nadu and West Bengal arrears. This is not a concession to those state governments; it is the removal of children from a political dispute they did not create.

15. Make the three-language formula genuinely flexible in practice, not only in text. NEP 2020 states no language is imposed. If that is meant, states choosing two languages should face no funding consequence. If it is not meant, the policy should say so and be defended openly.

---

PART 7 — WHAT WOULD MAKE THIS REPORT WRONG

An honest audit states its own vulnerabilities.

• Self-reported data. UDISE+ and AISHE numbers are institution-reported. Improvements in dropout rates and PTRs may partly reflect improved reporting rather than improved reality — and may equally understate genuine gains. Neither can be resolved without independent verification.
• Attribution of intent. The pattern in Section 3.5 is documented. The motive behind each individual change is not directly evidenced and this report does not claim otherwise. Readers who conclude the deletions were bureaucratic rather than political have to explain the pattern; readers who conclude they were political have to explain the Emergency.
• The Class 8 review committee. NCERT constituted a high-level panel in August 2025 to examine feedback on the Class 8 book with a mandate for evidence-based decisions. If it produces substantive revisions, that is a meaningful counter-datapoint on institutional self-correction and this report's assessment should be updated. If it does not report publicly, that is also information.
• ASER 2025/2026 data will test whether the 2024 foundational-learning gains were a genuine trend or a post-COVID rebound. That single data series should update the urgency ranking in Part 6 more than any argument in this report.
• This report relies on secondary reporting for textbook content in several places rather than direct examination of every edition. Anyone acting on Section 3 should obtain the physical books.

---

Principal sources

• ASER Centre / Pratham — Annual Status of Education Report 2022, 2023 (Beyond Basics), 2024
• Ministry of Education — UDISE+ 2023–24 and 2024–25 reports; AISHE 2022–23 and 2023–24
• Union Budget 2026–27, Ministry of Education demand for grants; Economic Survey 2023–24 and 2025–26
• NCERT — NCF-SE 2023; Exploring Society: India and Beyond (Class 8, 2025–26); rationalisation notifications 2022–23; statements of Director D.P. Saklani and Michel Danino
• UGC — Guidelines for Incorporating Indian Knowledge in Higher Education Curricula
• Rajya Sabha Department-related Parliamentary Standing Committee on Education (reports on vacancies and on FLN mission extension)
• All India People's Science Network; Breakthrough Science Society; Indian History Congress statements
• Narasimhan et al., Science (2019), on South Asian population genetics; Mukunda et al., IISc (1974), on the Vaimanika Shastra
• ORF, Ideas for India, IMPRI, The Hindu, Indian Express, The Wire, Swarajya, Careers360, Science.org reporting, 2022–2026

Sources spanning the political spectrum have been used deliberately, including outlets supportive of the current government's curriculum changes, so that the defenders' case is stated in their own terms rather than through their critics.