The Title Is the Asset — Why Indian Land Cannot Be Used as Capital

Pritiraj BrahmaFounding Partner, Pritiraj & Partners · 26 February 2026

India’s land records are presumptive, not conclusive. Fifteen years and ₹875 crore have produced 2.25 crore property documents and 168 districts at platinum grading — and the map that actually settles a boundary is 49 per cent complete against a textual record at 95 per cent. Every document produced remains rebuttable in court.

ProvenancePublished 26 February 2026; revised 28 August 2026 to repair a section that carried a heading with no text, to add the programme-output ledger, and to reconcile the closing claim with the counter-case. Completion figures are from the Department of Land Resources management information system as on 31 December 2023; the published disagreement on the textual share is stated rather than resolved. Bhoomi Samman grading is from the Press Information Bureau, 20 December 2023; SVAMITVA output from the Prime Minister’s statement of January 2025; Uttar Pradesh survey completion from a Lok Sabha reply of 21 July 2026. Statutory provisions are the Registration Act, 1908 and the Transfer of Property Act, 1882.

India’s largest store of household wealth is land, and India does not guarantee who owns it. The state has spent fifteen years and ₹875 crore industrialising the record — 2.25 crore property documents, 23 crore parcel numbers, 168 districts certified at platinum grading. The textual record is 95 per cent computerised and the map that actually settles a boundary is 49 per cent geo-referenced. Every document produced remains rebuttable in court.

Most Indians hold their wealth in land and property. Almost none of them hold a title the state guarantees. India operates a system of presumptive title: the record of rights, the registered sale deed and the mutation entry together create a presumption of ownership that any competing claimant may rebut in court. The state records a transaction; it does not certify the owner.

This is not a technicality. It is the reason a very large share of Indian civil litigation concerns land, the reason agricultural credit is collateralised so poorly, and one of the reasons infrastructure projects stall at acquisition rather than at construction.

Six findings anchor this analysis:

1. Registration proves a transaction, not a title. Under the Registration Act, 1908 and the Transfer of Property Act, 1882, registering a sale deed records that a transfer took place between named parties. It does not establish that the seller had good title to transfer. A chain of perfectly registered deeds can rest on a defective root, and nothing in the digitisation of that chain changes its legal character.
2. Records are fragmented across departments that do not reconcile. The record of rights sits with revenue, the map with survey and settlement, the deed with registration and stamps, and property tax with the local body. Four datasets describe the same parcel, are maintained by different authorities, and routinely disagree. Most of the practical disorder in Indian land is reconciliation failure rather than fraud.
3. Digitisation has copied the problem faithfully, and the spatial half is the half that is missing. The Department of Land Resources records computerisation of the record of rights in 6,25,137 of 6,57,397 villages — 95.09 per cent — as on 31 December 2023, against geo-referencing of cadastral maps in 3,26,776 villages, or 49.10 per cent. The textual record is nearly complete; the spatial record that locates a parcel on the ground is about half done, and it is the map rather than the text that settles a boundary. Published figures for the textual share do not agree: a Lok Sabha reply put records of rights at 98.5 per cent, a 2024 government note at nearly 96 per cent, and the Prime Minister at 98 per cent in January 2025. This Review reports the spread rather than choosing one silently.
4. The output is real and large. DILRMP was extended by the Ministry of Finance from 2021-22 to 2025-26 with an outlay of ₹875 crore, adding computerisation of revenue courts and consent-based Aadhaar linking. Under Bhoomi Samman, 168 districts across 16 states had reached platinum grading — 99 per cent and above on the core components — as on 20 December 2023, on the Press Information Bureau’s account. Under SVAMITVA, the Prime Minister stated in January 2025 that 2.25 crore property documents had been issued in five years and 23 crore Bhu-Aadhaar parcel numbers generated. In Uttar Pradesh alone, a reply to the Lok Sabha of 21 July 2026 recorded drone survey complete in all 90,573 notified villages with property cards distributed in roughly 67,000.
5. None of it converts a presumption into a guarantee. A SVAMITVA property card is a record of rights over inhabited rural land — issued faster and mapped more accurately than anything before it, and still rebuttable in court. Saturation is being certified district by district while the legal character of the record is unchanged nationally. That is the distinction this essay turns on.
6. The unresolved layers are the hardest ones, and a survey does not reach them. Inherited holdings undivided across generations, tenancy and sharecropping rights recorded informally or not at all, common and forest land under contested classification, and urban land held on lease from development authorities. A drone measures a boundary. It does not adjudicate who stands inside it.

The record that is nearly complete, and the map that is not

The gap between the two bars is the operative fact in Indian land administration. A textual record tells you that a person is recorded against a parcel number. A geo-referenced map tells you where that parcel is. Disputes are overwhelmingly about the second: where the boundary runs, whether the neighbour’s wall encroaches, whether the parcel sold is the parcel occupied. Completing the text first was the rational sequence — it is cheaper and faster — but it means the programme is reported as 95 per cent done on the half that generates fewer disputes.

What the programmes delivered, and what they did not

Three consequences follow from the unchanged legal character, and each is large.

Credit. A lender against land is lending against a presumption. Banks respond rationally: they lend conservatively, demand additional security, and prefer borrowers whose title has already been tested. The farmer or micro-enterprise with an unclear record is pushed to informal credit at far higher rates.

Litigation. Land disputes are among the largest categories of civil case, and they are slow by nature: they turn on documents decades old, boundaries never precisely surveyed, and family arrangements never written down. Judicial delay and title uncertainty are not two problems; they are one problem observed at two points.

Delivery. Every road, transmission line, industrial park and irrigation channel requires land. Where title is unclear, acquisition becomes a dispute-resolution exercise before it becomes an engineering one — and the delay is attributed to the project rather than to the record system that caused it.

What conclusive titling would require, and why it stalls

Conclusive titling means the state guarantees ownership and indemnifies anyone who suffers loss from an error in its own register. That is the model in force in several jurisdictions abroad, and a version of it has been drafted for India more than once.

It stalls on three real obstacles, which advocates tend to understate. First, cost and risk: a guarantee implies an indemnity fund, and no state wishes to underwrite the accumulated errors of a century of record-keeping. Second, sequencing: guaranteeing a bad record simply converts a private dispute into a public liability, so a resurvey must come first — and the resurvey is the 49 per cent bar. Third, and most difficult, the state is a party: a large volume of contested land involves government, forest or common land, and a titling exercise forces adjudication the state may prefer to postpone.

The counter-case, honestly stated

Three arguments cut against this analysis, and each has force.

First, the credit argument may be aimed at a channel the market is already leaving. This Review’s own analysis of small-enterprise finance identifies the most consequential recent development as the shift from collateral-based to cash-flow-based underwriting, using the Account Aggregator framework, GST returns and digital payment histories — a move made precisely because lending against pledged assets was the constraint. If lenders are learning to underwrite a firm’s cash flow without a mortgage, then guaranteeing title improves a route that is becoming less central. Conclusive title would still help the household with no digital trail. But the claim that it would unlock more household capital than any credit scheme yet designed is an assertion, and the catalogue in which it usually appears contains the counter-example.

Second, and most seriously, conclusive title can dispossess. A state guarantee works by extinguishing rebuttable claims — and the claims least likely to be documented are those of tenants and sharecroppers, of women whose inheritance was never recorded, and of communities occupying land classified as common or forest. Converting a presumptive record into a guaranteed one at a fixed date rewards whoever appears on the paper on that date and permanently forecloses whoever does not. An objection window is exactly the instrument that fails people without documents, legal advice or literacy, and it converts a slow injustice that remains contestable into a fast one that does not. This is the strongest argument against the reform, and it is the reason the recommendations below stop short of a national conversion.

Third, the foreign precedent is weaker than it appears. Guaranteed-title systems were largely built on land the state had already surveyed completely, with professionalised conveyancing, modest parcel counts, and — in several of the jurisdictions most often cited — on holdings the state had itself recently appropriated and could therefore register without adjudicating centuries of competing claim. India would be retrofitting a guarantee onto a century of contested records across a parcel count without international precedent, in a system where the state is itself a major disputant. That does not make conclusive title wrong. It means the comparison is doing rhetorical work its underlying similarity cannot support.

What we would do

Two of these survive every objection above, and they are the ones to do first regardless of what is decided about guarantees: resurvey before guaranteeing anything, and force the four records to reconcile. Neither extinguishes a right.

1. Resurvey before you guarantee, and publish the spatial completion figure alongside the textual one. Modern survey — drone or differential GPS — of one district at a time, with the draft map published and objections heard locally, is the precondition for everything else. It is also the number that should lead every progress statement: 49.10 per cent, not 95.09 per cent. Several states have run credible pilots; almost none have scaled them.
2. Make the four records reconcile by law. A single statutory parcel identifier carried across revenue, survey, registration and municipal tax, with a duty on each department to update on the others’ entries. Bhu-Aadhaar has generated 23 crore parcel numbers; what it lacks is a statutory obligation on four departments to treat one number as authoritative.
3. Record tenancy honestly, before any titling exercise, not after. Tenancy that cannot be recorded without risking the tenant’s or the owner’s position stays invisible, and invisible cultivators cannot access credit, insurance or procurement. They are also the people a guarantee would extinguish. Several states have legislated recorded leasing without loss of ownership rights; it works where the record is trusted.
4. Where conclusive title is attempted, do it district by district and after the resurvey. Guarantee title only where the survey is complete and objections have been heard locally, with legal assistance funded for undocumented claimants during the window. A single statewide switch is what makes the reform both unaffordable and unjust, and therefore permanently deferred.
5. Publish a district land-dispute index. Pending land cases per thousand parcels, and median age. It would tell a state exactly where to resurvey first, and it is computable today from data the courts already publish.

India’s poorest households hold their savings in an asset the state declines to certify. Fifteen years of programmes have made that asset faster to look up, more accurately mapped and better graded — and no more certain. Whether a guarantee should follow is a genuinely open question, because the people a guarantee would extinguish are the people least able to appear at an objection hearing. What is not open is the sequence: the map before the guarantee, the tenant on the record before the register closes, and the 49 per cent figure published where the 95 per cent figure is quoted.