Rent control in India is a body of law that answers a simple question with an increasingly indefensible answer: what may a landlord charge, and on what terms may a tenant be made to leave? In the old commercial quarters of Delhi and Mumbai, the answer was settled decades ago and has barely moved since. Rents fixed by reference to the cost of constructing a building in the middle of the twentieth century, or to the rent a premises fetched in 1940, continue to govern shops and homes whose market value has multiplied many times over. The tenant in occupation cannot be removed except on grounds a statute lists and a tribunal accepts. The tenancy itself may pass to the tenant’s family when the tenant dies. Around this frozen core, a modern rental market of eleven-month agreements, heavy security deposits and informal arrangements has grown up outside the statute book, because the statute book stopped describing the market it governs.

This article is the pillar for that body of law. It explains the architecture the states built after Independence, provision by provision: the standard rent mechanism and why it froze, the closed list of eviction grounds and the bona fide requirement at its centre, the heritability of tenancy that converts a lease into something close to a family asset, and the two worked examples that show the system in its most developed forms, the Delhi Rent Control Act, 1958 and the Maharashtra Rent Control Act, 1999. It then turns to the single most extraordinary episode in the legislative record this series covers. In 1995 Parliament passed a new rent law for Delhi. The President assented to it. The statute sits on the statute book as Act 33 of 1995, complete and enacted, and no government has ever brought it into force. When the refusal was challenged, the Union government defended it in court as a matter of executive policymaking into which judicial direction would intrude, and that defence, stated in an affidavit rather than left as unspoken practice, is the document that converts a habit this series has traced across many statutes into an articulated constitutional position. Finally, the article examines the Model Tenancy Act, 2021, the Union Cabinet’s answer to the frozen settlement: a model law with no force of its own, which states may enact or adapt, and whose deliberate exclusion of existing tenancies decides how much of the problem it can reach.

Wartime scarcity and the birth of the frozen settlement

Rent control in India was born as an emergency measure, and the emergency was war. Scarcity of housing in garrison and port cities during the First World War produced the first controls in Bombay, and the far larger dislocation of the Second World War generalised them, in several provinces first as control orders issued under the Defence of India Rules by executive authority, converted into Acts of the legislatures when the emergency passed and the scarcity did not. Delhi’s own statutory line runs through the same pressures: the Delhi High Court, recording the landlords’ challenge to the 1958 Act decades later, described a special and temporary legislation fashioned to meet emergency conditions created first by the Second World War and then by the Partition of India, which flooded the capital with displaced families and made eviction of sitting tenants a political impossibility. The pattern repeated across the country. Legislatures fixed rents at or near the level a premises already commanded, forbade eviction except on stated grounds, and described the arrangement as temporary, a bridge across an abnormal shortage.

The shortage outlasted every prediction, and the bridge became the road. Independent India’s states inherited the wartime controls and re-enacted them as permanent rent Acts through the 1940s and 1950s, each with its own definitions, its own fixation machinery and its own list of eviction grounds. Delhi’s own succession of statutes, the Delhi and Ajmer-Merwara Rent Control Act, 1947 and the Delhi and Ajmer Rent Control Act, 1952 among them, was consolidated in the Delhi Rent Control Act, 1958, Act 59 of 1958, brought into force on 9 February 1959. Bombay’s regime, the Bombay Rents, Hotel and Lodging House Rates Control Act, 1947, anchored standard rent to the rent payable on 1 September 1940. Madras legislated in 1960. The statutes differed in detail and shared a design: a controlled rent computed from history rather than from the market, security of tenure enforceable against the landlord’s ordinary property rights, and a specialised forum standing between the landlord and possession.

What none of the legislatures built was an exit. A control regime tied to historic figures contains no internal mechanism for catching up with prices, because every increase must be legislated against the organised resistance of sitting tenants, who are numerous, concentrated in identifiable neighbourhoods and able to punish a government at the next election. Landlords, by contrast, are fewer, dispersed and unsympathetic figures in public argument. The political arithmetic therefore ran one way for eight decades. Rents stayed where the statute had put them while the cost of maintaining the buildings, the municipal taxes assessed on them and the value of the land beneath them climbed without reference to the controlled figure. The courts noticed early and said so often. In Prabhakaran Nair v. State of Tamil Nadu, (1987) 4 SCC 238, the Supreme Court called for landlord and tenant law that was rational, humane, certain and capable of being quickly implemented, observed that there are weak tenants and weak landlords alike, and urged a national housing policy that would rationalise the rent structure. The observation was quoted back at governments for decades. The rationalisation arrived, in Delhi, in a form nobody expected: a complete replacement statute that was enacted, assented to and then shelved.

Who makes rent law, and where the general law of leases fits

The constitutional allocation explains the shape of everything that follows. The relation of landlord and tenant, including the collection of rents, sits in Entry 18 of the State List in the Seventh Schedule, within the entry for land. Rent control is therefore state legislation. There is no central rent control Act for India and there cannot be one for the states; each state legislature writes its own settlement, which is why the operative statutes in this field carry state names and state variations, and why advice drawn from one state’s Act misleads in the next. The Union’s roles are three, and each appears in this article. It legislates for the National Capital Territory of Delhi, which is how Parliament came to enact both the Delhi Rent Control Act, 1958 and its stillborn replacement. It formulates model legislation and commends it to the states, which produced the model rent control legislation circulated in the early 1990s and, a generation later, the Model Tenancy Act, 2021. And it defends the resulting arrangements in court, which is where the doctrine of non-commencement was articulated.

Delhi’s position within this allocation is sharper than the general rule, and it explains why the capital’s replacement statute is Parliament’s orphan rather than a state government’s embarrassment. Article 239AA of the Constitution, which created the Legislative Assembly of the National Capital Territory, withholds from that Assembly’s competence Entries 1, 2 and 18 of the State List, public order, police and land. The relation of landlord and tenant travels with the entry for land, so the Delhi Assembly cannot enact a tenancy code for the capital of its own authority: for Delhi, Parliament is not merely the historical author of the rent Acts but the only legislature the Constitution permits to replace them. Every actor in the Delhi story therefore answers to the same institution. The Parliament that passed the 1995 Act, the executive that declines to commence it and the government that defends the declinature in court are, for this subject, one constitutional complex, and the tenant or landlord of Delhi who asks which elected body might take the question up receives the answer that the body is Parliament and that Parliament acted, unanimously, in 1995.

Beneath the rent Acts sits the general law. A letting that no rent statute captures is governed by the ordinary law of leases in the Transfer of Property Act, 1882, where a lease is a transfer of a right to enjoy property for a time or in perpetuity, terminable in the ways the Act and the parties’ agreement provide, and where a landlord who wants possession on a ground the agreement supports may sue in the ordinary civil courts. The rent Acts are overlays on that base. Where a rent Act applies, it displaces the general law at the decisive points: the rent is the standard rent rather than the agreed rent where the two conflict, the grounds of eviction are the statute’s grounds and no others, and the forum is the statute’s forum rather than the civil court. Where a rent Act does not apply, because the premises fall in an exempt class or the tenancy was created outside its reach, the general law governs untouched. Much practical confusion in this field is a failure to ask the threshold question, which body of law governs this tenancy at all.

Who decides a rent dispute under the old Acts, a civil court or a Controller?

Under the Delhi Act, a specialised Controller appointed under the statute decides applications for fixation of standard rent and for eviction, with an appeal to a Rent Control Tribunal, and the civil courts are barred from the matters the Act assigns to that machinery. A landlord who sues for possession in an ordinary court is in the wrong forum.

The design is common to the first-generation Acts, though the personnel differ. The premise was that rent disputes are mass, repetitive and fact-simple, and that a dedicated authority applying a closed code would dispose of them faster and more cheaply than the civil courts, whose decrees the landlord would in any event have to execute against a protected tenant. The Maharashtra Act of 1999 routes its disputes through the courts it designates for the purpose, including the Court of Small Causes in Mumbai, rather than through an executive Controller, a reminder that even the forum is a state choice within the shared design. The Model Tenancy Act returns to the specialised model with a three-tier structure of its own, examined below. For the reader holding a notice or a summons, the practical consequence is that the letterhead matters: proceedings begin where the governing Act says they begin, and a defence or an application filed in the wrong forum consumes the one resource these disputes never replenish, which is time.

The specialised forum also explains a feature of the case law that puzzles outsiders. Because the Controller and the Tribunal work a closed statutory code, appellate courts reviewing their orders are confined to the statute’s questions: was the ground made out, was the rent correctly fixed under the formula, was the procedure followed. The large constitutional questions, whether the code itself had become unreasonable with age, could only be raised by direct challenge to the Act, which is why the history of this field is punctuated by a small number of big constitutional cases rising above a vast, quiet mass of Controller-level litigation.

Standard rent: the arithmetic of a freeze

The standard rent is the load-bearing concept of the old regime, and it is a computation, not a bargain. Its purpose was to answer the question the market could not be trusted with in a scarcity: what is a fair return on this building? The first-generation Acts answered with a formula tied to history. Under the Delhi Act, premises let before 2 June 1944 carry a basic rent determined under the Second Schedule to the Act, and the standard rent is built on that foundation; for premises let later, the Controller fixes the standard rent when asked, having regard to the rent of similar premises in the locality and to what the building cost to construct and what the land beneath it was worth when construction began. The Maharashtra legislation worked from the rent payable on 1 September 1940. In each variant the inputs are historic: a construction cost incurred in another monetary era, a land value struck before the modern city existed, a rent recorded in a ledger generations old. Apply any permitted percentage return or permitted increase to a base that old and the output is a figure the market left behind long ago.

Two features of the mechanism deserve emphasis, because together they manufacture the grievance that fills the court reports. The first is that the standard rent is a ceiling with legal teeth. The Delhi Act forbids a landlord to claim or receive rent above the standard rent and the increases the Act permits, and a stipulation in which the tenant agrees to pay more does not cure the defect. The Maharashtra Act states the prohibition as an offence: under Section 10 of the Maharashtra Rent Control Act, 1999, claiming or receiving increases above the standard rent and the permitted increases is punishable, on conviction, with imprisonment up to three months or a fine up to five thousand rupees or both. The second feature is how small the permitted movement is. The Delhi Act, by Section 6A, allows the rent to be increased by ten per cent every three years, a rate the Union government itself cited in court when defending the Act against the landlords’ constitutional challenge, as evidence that the regime was not wholly static. Maharashtra’s Section 11 permits an increase of four per cent per annum, with further room for increases tied to improvements and to taxes. Set either rate against decades of general price rise and the conclusion writes itself: the statute permits the rent to crawl while every cost around it runs.

The fixation standard conceals a circularity that practitioners recognise and the case law circles around. The Controller fixing the standard rent of premises is directed to the rent of similar premises in the locality; in a locality where control is general, the similar premises are themselves let at controlled figures struck from the same historic base, so the inquiry compares the premises against a market the statute has abolished and returns, by a route of impeccable logic, the frozen answer. Where cost of construction and historic land value supply the test instead, the inquiry is at least anchored in fact, but the facts are those of the year the building rose, recovered from records that may no longer exist and valued in a currency whose meaning has changed beyond recognition. Either route explains why fixation proceedings, though framed as technical inquiries, were so often the decisive battle of a tenancy: the number they produced would govern the premises, at permitted crawl, for the rest of the building’s life and the lives of the tenant’s successors.

The fixation machinery adds a further twist. Because the standard rent of a premises can be fixed by the Controller on the application of either side, and because a fixation, once made on the merits, is difficult to reopen absent structural alteration or a change in amenities, a single fixation order can govern a building for the remainder of its economic life. Landlords learned to avoid the machinery; tenants learned to invoke it whenever a landlord demanded a market figure informally. The informal market responded with devices the statute never contemplated, of which the pagdi premium, examined below, is the most durable: if the monthly figure cannot move, value migrates into a capital payment for the tenancy itself, outside the rent column altogether.

The arithmetic is best seen on a stylised building, offered as an illustration of the mechanism rather than as the record of any actual premises. Take a shop let in the mid-1940s at a monthly rent that the Act’s schedules treat as the basic rent, and carry that figure forward under the Delhi Act’s permitted movement of ten per cent every three years. Compounded across the decades since 1959, the lawful rent multiplies several times over, which sounds substantial until it is set against the transformation of the city around the shop: the same premises, let afresh in the market the statute does not govern, would command a monthly sum bearing no visible relation to the controlled figure, and the gap between the two numbers is the value the regime transfers, month by month and succession by succession, from the owner of the building to the holder of the tenancy. The illustration’s proportions, not its figures, are the point, and they are the proportions the landlords’ association carried into the Delhi High Court when it compared controlled monthly rents with a single day’s minimum wage.

It is worth pausing on what the standard rent was for, because the tenant’s case begins here and it is not a foolish case. In a scarcity market, an uncontrolled rent is not a price signal; it is a ransom. The legislature of 1947 or 1958 was not choosing between a frozen rent and a healthy market. It was choosing between a frozen rent and the expulsion of poorer households and small traders from the urban core, at a moment when no alternative housing stock existed to receive them. The formula’s tie to construction cost and land value expressed a genuine theory, that a landlord is entitled to a fair return on investment and to nothing beyond it, the scarcity premium being a windfall the community may legitimately deny him. The theory’s weakness is temporal. A return that was fair on 1940 values is not a return at all on the values of a later century, and a legislature that adopts the theory owes the landlord a mechanism that keeps the base honest. No legislature built one. That omission, repeated across states and decades, is the technical heart of the rent control problem.

The forgotten branch: fair rates for hotels and lodging houses

Both Delhi statutes carry a branch of control that public discussion of rent law has entirely forgotten, and its presence in the text is a useful reminder of how total the mid-century ambition was. The Delhi Rent Control Act, 1958 does not stop at tenancies of premises; it extends to the charges of hotels and lodging houses, defining a fair rate as the rate fixed under Section 31 of the Act and revised under Section 32, and subjecting what a hotel may charge for lodging, with or without board and other services, to fixation by the same machinery that fixes the standard rent of a shop. The Delhi Rent Act, 1995 carried the branch into the replacement code, its preamble reciting the regulation of rates of hotels and lodging houses among its purposes and its own fair rate provisions appearing at Sections 39 and 40. The theory was continuous with the theory of standard rent: lodging in a scarcity is a necessity, the price of a necessity is a fit subject for a fair-return formula, and the traveller at the mercy of a wartime city stands in the same position as the tenant at the mercy of its landlords. The branch withered in practice as the hotel industry formalised and as exemptions and market segmentation moved commercial lodging outside the fixation machinery’s effective reach, and no modern account of Indian hospitality law begins with the Controller’s fair rate. Its survival in both texts, the governing one and the uncommenced one, marks the distance between the regulatory imagination of the period that wrote these statutes and the market order that replaced it in fact while leaving it in force in law.

The working detail: exemptions, deposits and the tenant’s counter-machinery

A rent Act is best understood by the tenancies it leaves out, because the exemptions mark where the legislature accepted that the market could be trusted. Section 3 of the Delhi Act excludes premises belonging to the government from the Act’s protection altogether. More consequentially, the amendment of 1988 carved out the top of the market: premises whose monthly rent exceeds three thousand five hundred rupees stand outside the Act, and premises in buildings constructed after the amendment’s appointed point are exempt for ten years from completion, a window designed to assure investors in new construction a decade of market rents before the protective regime descends. The recalibration of 1988, which also supplied the ten per cent triennial increase in Section 6A and the special eviction rights for members of the armed forces, government servants and widows discussed below, is the nearest the Delhi Act has come to internal reform, and its logic is the logic the Model Tenancy Act would later generalise: protection withdrawn by category, where the occupant’s bargaining power or the building’s newness makes protection unnecessary or harmful. Its limits are equally instructive. A rent ceiling for exemption fixed in rupees in 1988 was eroded by the same inflation that eroded the standard rents beneath it, drawing ever more ordinary tenancies out of protection by arithmetic rather than by decision, while the tenancies beneath the line stayed frozen at their historic figures. A threshold that is never indexed performs a slow decontrol nobody voted on, in one part of the market, and none at all in the other.

Within the protected zone, the Act equips the tenant with counter-machinery that practitioners use weekly and summaries omit. Where a landlord refuses to accept rent, or a dispute exists about who is entitled to receive it, the tenant may deposit the rent with the Controller under Section 27, converting a tender the landlord spurns into a payment the law recognises and breaking the arrears ground before it forms. The landlord’s obligations run the other way: Section 44 obliges him to keep the premises in good and tenantable repair, and Section 45 protects the tenant’s essential supplies and services, empowering the Controller to intervene where a landlord cuts off or withholds water, electricity or other necessaries to squeeze the tenant out, with penal consequences under the Act’s offence provisions for the landlord who tries it. These provisions matter beyond their individual use, because they disclose the regime’s working theory of the relationship: not a bargain between equals to be enforced as written, but a continuing occupation to be policed by a standing authority, in which each side’s self-help is disabled and every pressure must be converted into an application. The landlord may not raise the rent, lock the door or cut the wire; the tenant may not withhold the rent, sublet the premises or alter their use. Both are remitted to the Controller, and the Controller’s pace becomes the pace of the housing market in the old city.

Eviction: a closed list of grounds and the bona fide requirement

If standard rent freezes the price, the eviction provisions freeze the occupation. The general law lets a landlord determine a tenancy in the ways the Transfer of Property Act and the lease permit. A rent Act substitutes a rule of the opposite character: the tenant shall not be evicted, and no decree or order for recovery of possession shall be made, unless the landlord applies to the statutory forum and proves one of the grounds the statute lists. The list is closed. A landlord whose grievance is real but unlisted has no remedy; a landlord whose ground is listed must still prove it, on evidence, in proceedings the tenant can contest and appeal.

Section 14 of the Delhi Rent Control Act, 1958 is the archetype, and its grounds repay reading as a group, because they disclose what the legislature thought a tenancy was for. The default ground protects the bargain itself: a tenant who has neither paid nor tendered the whole arrears of rent within two months of a notice of demand may be evicted, a ground softened by the tenant’s statutory opportunities to deposit the rent and save the tenancy. The conduct grounds police the tenant’s use of the premises: subletting, assigning or parting with possession without the landlord’s consent in writing; using the premises for a purpose other than the purpose of the letting; causing substantial damage. The absence ground addresses the tenancy that exists only on paper, premises let for residence in which neither the tenant nor the tenant’s family has resided for six months. The rebuilding grounds let a landlord recover premises that have become unsafe or unfit, or that are genuinely required for building or rebuilding that cannot proceed with the tenant in occupation. Special provisions carve out accelerated rights for defined classes, members of the armed forces, government servants and widows among them, whose need for their own premises the legislature thought should not wait in the ordinary queue.

At the centre of the list stands the bona fide requirement ground, Section 14(1)(e) in the Delhi Act: premises let for residential purposes are required bona fide by the landlord for occupation as a residence for himself or for a dependent member of his family, and the landlord has no other reasonably suitable residential accommodation. Every element litigates. Bona fide means the requirement must be genuine and honest, more than a pretext for reletting at a higher figure, though the courts have long accepted that the landlord is the best judge of his own residential needs and that a court will not substitute its view of what accommodation should content him. The absence of other reasonably suitable accommodation must be shown, and comparative hardship between the parties colours the exercise. For decades the ground carried a restriction that generated its own jurisprudence: it spoke of premises let for residential purposes, and commercial tenancies, however genuine the landlord’s need, stood outside it on the reasoning that a legislature may protect a family’s home more strongly than a trader’s shop. In Satyawati Sharma v. Union of India, (2008) 5 SCC 287, the Supreme Court read the provision down, holding that the classification had lost its rationale with time and that the bona fide requirement ground extends to premises let for commercial purposes as well; a provision reasonable at enactment, the Court accepted, may become unreasonable as circumstances change, a sentence that landlords would later aim at the entire Act.

The closed list does quieter work that matters more in aggregate than any single ground. It converts eviction from a property remedy into a public decision. The landlord’s title is not in question; what the statute suspends is the ordinary consequence of title, the right to recover one’s own premises at the end of a letting. Every year of suspension transfers value, because the tenant occupies at the standard rent an asset the market would price many times higher, and the transfer is not means-tested: the protection attaches to the tenancy, not to the tenant’s need. Affluent tenants in prime commercial locations enjoy it on the same terms as the poor households the regime was designed to shelter, a point the courts have made with increasing sharpness and the landlords’ associations have made the centre of their constitutional case. Against it stands the reliance interest, examined at full strength later in this article: generations of tenants, particularly in the old commercial quarters, paid substantial premiums and built businesses and lives on the faith of this security, and a legislature asked to withdraw it retrospectively is not merely correcting a price.

How eviction litigation is actually fought under the closed list

The grounds read, in the statute book, like a checklist; in the Controller’s court they behave like a siege. The landlord bears the burden throughout, and the tenancy’s antiquity works against him at every point of proof: documents from the letting are lost, the original parties are dead, the rent receipts that exist prove the controlled figure and nothing else, and the tenant’s defence is conducted by heirs whose occupation the Act protects in its own right. The arrears ground, which appears the most mechanical, illustrates the pattern. The landlord must serve a notice of demand and then show that the tenant neither paid nor tendered the whole arrears within two months of its service; the tenant answers with tenders, money orders, deposits made with the Controller where rent was refused, and the statutory opportunities to save the tenancy by payment, so that a ground framed as a simple default inquiry becomes an accounting contest across decades of small monthly sums. During the proceedings themselves, the Act arms the Controller to direct the tenant to pay or deposit the rent as it falls due, and a tenant who defaults on such a direction risks the striking off of his defence, a sanction that disciplines the tenant’s side of the record as the burden of proof disciplines the landlord’s. The bona fide requirement ground generates the deepest litigation of all, because it turns on states of mind and family circumstance: the landlord’s marriages, births, retirements and business failures are proved like facts in a family suit, the tenant answers with the landlord’s other properties and other lettings, and comparative hardship is argued household against household. Appeals carry the record to the Tribunal and writ petitions carry it beyond, so that a possession decree, when it comes, may arrive a generation after the application, against occupants who were children when it was filed. This is the lived content of the phrase security of tenure, and it cuts in two directions for the analysis. For the tenant household it is the substance of the protection Parliament promised. For the system it is a standing demonstration that a specialised forum does not by itself produce speed, because the delay lives not in the forum’s procedure but in the stakes: where possession is worth a fortune and rent is nominal, every interlocutory order is worth appealing, and the parties litigate accordingly. The Model Tenancy Act’s sixty-day design target, examined below, is written against this record, and whether a new forum can outrun the same incentives is the open empirical question of the reform.

The heritable tenancy and the pagdi premium

A lease is supposed to end. The tenant dies, the term expires, the premises return to the owner, and the market re-prices the occupation. The old rent Acts interrupt that sequence at its most sensitive point by making the tenancy heritable. The Delhi Act’s definition of a tenant extends, on the tenant’s death, to specified members of the tenant’s family in a stated order, broadly the spouse, children, parents and a daughter-in-law, who were ordinarily living with the tenant, and the Supreme Court settled the underlying question of principle in Gian Devi Anand v. Jeevan Kumar, (1985) 2 SCC 683: a statutory tenancy, the tenancy that continues under the Act’s protection after the contractual term has ended, is heritable in the same manner as a contractual tenancy, and the landlord cannot treat the tenant’s death as the tenancy’s end. The case that produced the holding shows what was at stake in it. Gian Devi Anand concerned commercial premises, a shop whose tenant had died, and the landlord’s argument was the one the trade most wanted established: that a contractual tenancy is heritable property, but the statutory tenancy, the creature of the Act’s protection that arises when the contract ends, is a personal right that dies with the tenant, at least for premises let for business. A Constitution Bench rejected the distinction, holding that the legislature which protected the tenant’s occupation during his life cannot have intended the protection to evaporate at his death for the family the occupation supported, and that heritability attaches to the tenancy whether the premises are residential or commercial. The holding was humane in the case that produced it and transformative in aggregate. A protection designed for a sitting tenant became a protection for a household across generations, renewed at each death, at a rent computed from the middle of the last century.

Heritability is the feature landlords resent most, and the resentment is intelligible. Every other element of the regime can be described as regulation of a bargain; heritability abolishes the bargain’s term. The owner who let a shop in 1960 to a trader deals, in the next century, with the trader’s grandchildren, at a rent bearing no relation to the shop’s value, and cannot recover the premises even for his own family’s business unless he proves a statutory ground against occupants who may be wealthier than he is. When the agitation against the Delhi Rent Act, 1995 is examined below, the inheritance of tenancies appears in the record as one of the provisions the trader tenants mobilised to defend, and the 1995 Act’s restriction of heritability was among the changes the government cited when it explained, in Parliament and in court, why the Act could not simply be switched on. Any reform that touches heritability touches the point where the old regime stops being a rent policy and becomes a distribution of urban property rights, which is why reformers keep failing there.

Why does a pagdi tenant behave like an owner?

Because, in substance, he is one. Under the pagdi system that grew up around the Bombay and Maharashtra legislation, the incoming tenant pays a capital premium for the tenancy itself, to the outgoing tenant or the landlord. Having bought the occupation, he treats it as an asset: held for decades, passed within the family, and sold on when he leaves.

The pagdi system is the market’s verdict on the standard rent. Where the monthly figure is frozen by law, the value of occupation does not disappear; it capitalises. A tenancy of premises in a prime Mumbai locality, carrying a nominal monthly rent and statutory protection against eviction, is worth a large capital sum, and pagdi is the mechanism by which that sum is paid and realised. The incoming tenant pays the premium to acquire the tenancy; when he leaves, he recovers a premium from his successor; the landlord, whose consent the transfer requires in practice, customarily receives a share of the consideration as the price of recognising the new occupant. None of this appears in the statute, which prohibits premiums in some contexts and ignores them in others, and all of it is common knowledge in the property market of the old city. The system’s consequences run in several directions at once. It gives the sitting tenant a capital stake that makes eviction politically and morally harder, because the tenant is no longer a mere occupier but an investor who paid for his place. It gives the landlord a grievance of a particular sharpness, since the asset being traded is his building, and the appreciation being realised is appreciation of his property, divided between the outgoing tenant and himself on terms custom sets rather than law. And it explains the otherwise baffling economics of the old quarters, in which premises officially let at negligible rents change hands, in substance, for sums that reflect their true market worth.

Pagdi also supplies the strongest concrete form of the tenant’s reliance argument, and honest treatment of this subject requires stating it without irony. The trader who paid a premium for a tenancy in an old commercial district paid it under a legal regime whose central promise was security: the premium was the price of an occupation the statute protected, of a shop his family could hold, build a clientele around and pass on. The price capitalised the protection. A reform that strips the protection without compensating the premium does not merely change the rent; it expropriates the capital the tenant paid for the protection, and transfers it to the landlord, whose asset appreciates correspondingly. One may still conclude that reform is necessary. One cannot honestly describe the tenant who resists it as a mere squatter on another’s property, and the political history of the 1995 Delhi Act, whose deemed rent and inheritance provisions threatened precisely this class of interest, is unintelligible unless the premium economy is kept in view.

The licence device: how the market routes around the statute

No account of this body of law is complete without the instrument by which the modern market escapes it. The escape route is the leave and licence. In the general law, a licence, defined in Section 52 of the Indian Easements Act, 1882, is a permission to use premises which creates no interest in the property itself: the licensee occupies by leave, acquires no estate, and can claim none of a tenant’s protection. The Supreme Court drew the line between lease and licence in Associated Hotels of India Ltd. v. R.N. Kapoor, (1960) 1 SCR 368, holding that the substance of the arrangement governs over its label, that a document calling itself a licence may in truth create a lease where exclusive possession and the parties’ intention point that way, and that the court’s task is to find the real bargain beneath the drafting. The distinction decides fortunes in this field. A tenant under a rent Act holds a heritable, price-controlled, eviction-proof interest. A licensee holds a revocable permission. Drafting therefore migrates, wherever the landlord has the bargaining power to insist on it, toward the licence form, and the eleven-month leave and licence agreement, renewed in sequence, became the standard instrument of the formal rental market in Mumbai and much of urban India: a term short enough to signal temporariness, a structure designed to keep the occupier a licensee, and a rent that moves with the market because no standard rent attaches to a permission.

Maharashtra’s legislature answered the device in part within the 1999 Act itself, and the answer shows a legislature learning from its predecessor’s evasion. Section 55 of the Maharashtra Rent Control Act, 1999 provides that any agreement for leave and licence or letting of premises entered into after the Act’s commencement shall be in writing and shall be registered under the Registration Act, 1908, that the responsibility for registration lies on the landlord, and that in the absence of a written registered agreement the contention of the tenant or licensee about the terms on which the premises were given shall prevail unless proved otherwise, the landlord’s failure to register being itself an offence punishable with imprisonment up to three months or a fine up to five thousand rupees or both. Section 56 performs the complementary legalisation: it makes it lawful for a tenant to claim or receive consideration for the relinquishment or assignment of his tenancy, and for a landlord to receive a premium or other consideration for granting or renewing a lease or for consenting to a transfer. The pagdi economy, which the 1947 Act had driven into informality, is by these provisions brought partly into the open, taxed with a registration duty and stripped of its illegality. What the 1999 Act did not do, because no state Act has done it, is close the licence route itself: the registered eleven-month licence, at a market fee, remains the governing instrument of new urban letting in Maharashtra, and the protected tenancy survives as the law of the older stock. The two instruments divide the city between them, and the date and form of a letting, rather than any assessment of the parties’ needs, decide which law an occupant lives under.

Maharashtra: the second worked example, and the judgment that forced a new Act

Maharashtra inherited three regional rent regimes, for the old Bombay State, for Vidarbha and for Marathwada, and the Bombay Rents, Hotel and Lodging House Rates Control Act, 1947 was the most litigated of them. Its standard rent was anchored to the rent payable on 1 September 1940, with permitted increases layered on top, so that by the closing decades of the century the controlled rents of Mumbai’s older buildings stood at levels that bore no relation to the cost of keeping a building standing. Landlords’ associations carried the grievance to the Supreme Court, and in Malpe Vishwanath Acharya v. State of Maharashtra, (1998) 2 SCC 1, a three-judge bench accepted the core of it. The Court held that the provisions of the Bombay Act governing the determination and fixation of standard rent were no longer reasonable, observing in substance that a fixation mechanism defensible in the conditions of 1947 had become unjust as decades passed and values transformed around it. The Court nevertheless declined to strike the provisions down, because the State was engaged in framing a new rent law and the judicial remedy of invalidation would have demolished the protection of every tenant in the state at a stroke, replacing a bad settlement with none. The judgment is the high-water mark of judicial pressure in this field: a finding that the existing law had ceased to be reasonable, coupled with a refusal to supply the replacement, addressed to a legislature told plainly that the task was its own.

Maharashtra’s answer was the Maharashtra Rent Control Act, 1999, Maharashtra Act 18 of 2000, which received the President’s assent on 8 March 2000 and came into force on 31 March 2000, repealing and replacing the three regional Acts with a single code for the entire state. The 1999 Act is instructive precisely because it is a genuine legislative attempt at rebalancing that nonetheless kept the architecture of the old regime. The concept of standard rent survived, with the court fixing it where the parties dispute it, and with the prohibition on charging above it restated as an offence under Section 10. The movement of rents was liberalised at the margin: Section 11 entitles the landlord to an increase of four per cent per annum, and the Act makes room for further increases reflecting improvements, special additions and heavier repairs, and for passing through increases in taxes. Whole classes of letting were taken out of the protective regime, the Act not applying to premises let to government, public sector undertakings, banks and comparable institutional tenants whose bargaining power needs no statutory shelter. Heritability, eviction grounds and the rest of the first-generation design were retained in modernised form. Landlord associations attacked the four per cent figure as derisory against the values Malpe had described; tenant organisations attacked the same figure as the thin end of escalation. Both reactions identify the Act’s real character. It is the old settlement with its most indefensible feature, the absolute freeze, relaxed into a slow crawl, and its enactment demonstrates the outer limit of what a state legislature, left to itself, has been willing to do: movement, yes; market rents, no.

Mumbai adds a physical chapter to the Maharashtra story that Delhi’s record lacks, because the frozen stock there is not merely uneconomic but dangerous. The island city’s older tenanted buildings, many of them a century old and maintained at rents fixed generations ago, were brought within a public repairs regime under the state’s housing and area development framework: the buildings classed as cessed buildings, whose occupants contribute a repair cess, became the charge of a dedicated repairs and reconstruction board, on the reasoning that a rent structure which made private maintenance impossible had made structural safety a public responsibility. Building collapses in the old quarters, and the casualties that attend them, are the frozen rent translated into masonry. The phenomenon completes the Malpe logic in a register the courtroom figures do not capture. A statute that fixes the price of occupation below the cost of the building’s upkeep does not abolish the cost; it defers it, compounds it, and finally collects it in the currency of dilapidation, and the bill is presented to the occupants the statute exists to protect, in the buildings least able to bear it. Redevelopment schemes for the cessed stock, with their conditions for rehousing tenants in the rebuilt premises, are the state’s attempt to unwind in one project everything the rent regime wound up in ninety years, and their slow progress measures the depth of the winding.

The comparison with Delhi is the point of placing the two examples side by side. Maharashtra, prodded by a Supreme Court finding that its law was no longer reasonable, legislated a replacement and brought it into force, however modest its content. Delhi’s replacement was legislated by Parliament with far more ambition, and never brought into force at all. The variable that differs is not drafting quality or judicial pressure. It is the willingness of the executive to spend political capital against the tenants of the old quarters, and that variable is the subject of the next three sections.

The reform consensus that produced a Bill and nothing else

By the time Parliament took up the Delhi Rent Bill, the case for reform had been assembled so often, by bodies of such different character, that the landlords’ later pleadings could present it as a catalogue. The Economic Administration Reforms Commission, chaired by L.K. Jha, reported in September 1982 on the negative effects of the rent control laws, identifying the discouragement of housing investment and the distortion of the rental market a full decade before the model legislation exercise began. The Supreme Court added its institutional voice in Prabhakaran Nair in 1987, calling for rational and humane landlord and tenant law and a national housing policy to match. The National Housing Policy supplied the framework within which the Central Government formulated the model rent control legislation and circulated it to the states, the exercise from which the Delhi Bill of 1994 directly descended, and the landlords’ pleadings in the constitutional challenge collected the later endorsements of the same diagnosis: the Jain Commission of 1998 reiterating the damage rent control does to housing supply and investment, and the Planning Commission recording the case for reform in its Tenth Five Year Plan for the years from 2002 to 2007. A consensus of commissions, courts, plans and policies therefore stood behind the Bill when the Rajya Sabha and the Lok Sabha passed it without a dissenting vote in 1995.

The judiciary’s own blueprint from within that consensus deserves separate record, because it proposed the institutional form the reform eventually took. Prabhakaran Nair, in the passage the Supreme Court quoted against the government in Common Cause a decade and a half later, urged that the idea of a national rent tribunal on an all India basis, with quicker procedure, be examined, that the tiers of appeal in rent litigation be curtailed, and that the laws be made simple, rational and clear, observing that litigation must come to an end quickly and that there are weak persons among landlords as among tenants. The Rent Authority, Rent Court and Rent Tribunal structure of the Model Tenancy Act, and of the Tamil Nadu Act that preceded it, is that suggestion enacted thirty years on, at state level and in three tiers rather than one, which suggests the blueprint was sound and only the will to build from it was rationed.

The consensus is the more damning document for what it failed to move. Every element of the reform programme that depended on a state legislature’s voluntary adoption failed with it: the model legislation circulated to the states in the early 1990s was enacted by almost none of them, and the one jurisdiction where the Centre could legislate directly, its own capital, received the model as an Act and then watched the Centre decline to commence it. The pattern deserves naming because it recurs across this series in other subjects: Indian legislative reform of a politically defended status quo characteristically completes every stage except the operative one, generating policies, models, bills, enactments and judgments in abundance, while the distribution of costs that reform would disturb remains exactly where the unreformed law placed it. Rent control is the purest instance because the operative stage here was reduced to its smallest possible form. No appropriation was needed, no machinery had to be built before the first tenancy could be governed, no state concurrence was outstanding. A single notification in the Official Gazette stood between the enacted law and its operation, and for thirty years no government signed it.

The Delhi Rent Act, 1995: a law that never started

The Delhi Rent Bill, 1994 grew out of the model rent control legislation the Central Government had formulated and circulated to the states as the template for rationalising their rent laws, the same exercise the Supreme Court described in Common Cause v. Union of India when the Bill’s fate reached it. The Bill was introduced in the Rajya Sabha on 26 August 1994. The Rajya Sabha passed it unanimously on 29 May 1995; the Lok Sabha passed it unanimously on 3 June 1995; the President assented on 23 August 1995, and the Delhi Rent Act, 1995 entered the statute book as Act 33 of 1995. By every measure of the legislative process the country teaches its students, the law had been made. What the Act did not contain was a commencement date. Section 1(3) provided that it shall come into force on such date as the Central Government may, by notification in the Official Gazette, appoint. That notification has never been issued. The Delhi Rent Control Act, 1958 governs Delhi tenancies instead, as it has governed them since 1959, and the 1995 Act has spent more than three decades as the most complete specimen in Indian law of a statute that exists without operating.

The 1995 Act was no timid measure, and understanding what it proposed explains both the agitation that buried it and the stakes of the doctrine that buried it deeper. It was designed as a modern rent code for the capital: rent fixation reworked around the concept of deemed rent in place of the 1958 formula, registration of tenancies brought into the scheme, the inheritance of tenancies restricted rather than transmitted indefinitely, enhancement of rent put on a more realistic footing, eviction recalibrated, and disputes assigned to a dedicated Rent Authority and Tribunal whose chairman, members and benches the Act defines in its opening sections. Each of those features struck at a distinct interest organised under the old Act. Deemed rent threatened the frozen figures. Registration threatened the informal arrangements layered over formal tenancies. Restricted heritability threatened the family asset. The trader tenants of the old city, organised and proximate to power in a capital city whose government then depended on their support, agitated against the Act before its commencement, and the government of the National Capital Territory responded with an All Party Committee to examine the grievances, which centred, on the Union’s own later account in court, on deemed rent, registration of tenancies, inheritability of tenancies, enhancement of rent and eviction.

The Act’s own architecture, as its text discloses it, shows how complete the replacement was meant to be. Its opening chapters define the machinery before the grievances: a Tribunal with a Chairman and Members, Rent Authorities who were to be public servants under the Act and to whom pending cases under the old regime were to be transferred, a bar on the jurisdiction of the civil courts in the matters assigned to the new machinery, protection for action taken in good faith, and special provisions governing decrees passed under the Delhi Tenants (Temporary Protection) Act, 1956 and the 1958 Act itself, so that the transition from the old code to the new was drafted, clause by clause, in Sections 78 and 79 of the replacement. The Act regulated the rates of hotels and lodging houses through the concept of a fair rate, fixed under Section 39 and revised under Section 40, carrying into the new code a branch of control the capital’s older statutes had maintained. Its closing sections conferred the power to remove difficulties and the power to make rules, and Section 82 repealed the 1958 Act with savings. Nothing in this apparatus was provisional or skeletal. The statute that waits for notification is not a sketch awaiting a draftsman; it is a finished legal system, with its courts, its transition code and its repeal already enacted, lacking exactly one thing, the executive’s signature on a commencement notification that its own Section 1(3) made the final step.

The Union government’s response set the pattern for thirty years: the Act would not be notified in its enacted form; it would be amended first. The Delhi Rent (Amendment) Bill, 1997 was introduced in the Rajya Sabha on 28 July 1997 to modify the parent Act before the parent Act had lived a single day, and was referred to the Parliamentary Standing Committee, which finalised its report in December 2000. The government accepted the Committee’s recommendations on 3 April 2001 and sent notice for moving the official amendments to the Rajya Sabha in July 2001. There the trail dissolves into workload and silence. The Amendment Bill was never carried through, and in time it lapsed without enactment, leaving the 1995 Act in its original, unamended and uncommenced condition. A decade later the government changed direction without changing destination. In July 2013 the Cabinet approved the repeal of the Delhi Rent Act, 1995 and the withdrawal of the pending Amendment Bill, and the Delhi Rent (Repeal) Bill, 2013, Bill No. LIX of 2013, was introduced in the Rajya Sabha on 29 August 2013 to repeal a statute that had never been in force. The Repeal Bill was not referred to a standing committee and was never passed. The statute book therefore holds, simultaneously, a 1958 Act that governs, a 1995 Act that replaced it in form and never in fact, an amendment bill that died unpassed, and a repeal bill that died unpassed: four legislative artefacts and one operative law, the oldest of the four.

Stage	What happened	Law governing a Delhi tenancy
1958 to 1959	Parliament enacts the Delhi Rent Control Act, 1958; it is brought into force on 9 February 1959	Delhi Rent Control Act, 1958
26 August 1994	The Delhi Rent Bill, 1994 is introduced in the Rajya Sabha	Delhi Rent Control Act, 1958
29 May and 3 June 1995	The Rajya Sabha and the Lok Sabha each pass the Bill unanimously	Delhi Rent Control Act, 1958
23 August 1995	The President assents; the Delhi Rent Act, 1995 becomes Act 33 of 1995, its commencement left by Section 1(3) to a government notification	Delhi Rent Control Act, 1958
1995 to 1997	Trader tenants agitate against deemed rent, registration, restricted inheritability and eviction; an All Party Committee examines the grievances; no notification issues	Delhi Rent Control Act, 1958
28 July 1997	The Delhi Rent (Amendment) Bill, 1997 is introduced in the Rajya Sabha to amend the Act before it commences	Delhi Rent Control Act, 1958
1997 to 2001	Common Cause files Writ Petition No. 1495 of 1997 in the Delhi High Court seeking a mandamus to notify the Act; the Standing Committee reports in December 2000; the government accepts its recommendations on 3 April 2001	Delhi Rent Control Act, 1958
8 October 2003	The Supreme Court in Common Cause v. Union of India declines to direct commencement, applying A.K. Roy	Delhi Rent Control Act, 1958
July and August 2013	The Cabinet approves repeal of the 1995 Act; the Delhi Rent (Repeal) Bill, 2013 is introduced in the Rajya Sabha on 29 August 2013 and never passed	Delhi Rent Control Act, 1958
7 January 2019	The Delhi High Court decides the landlords’ challenge to the 1958 Act in Shobha Aggarwal v. Union of India, recording that a new draft Bill was favoured in 2012 and that the repeal Bill of 2013 remained pending	Delhi Rent Control Act, 1958
2 June 2021	The Union Cabinet approves the Model Tenancy Act, 2021, a model for states to enact; it does not touch Delhi’s governing statute	Delhi Rent Control Act, 1958
The pending challenge	In the landlords’ renewed proceedings in the Delhi High Court, the Union’s affidavit defends non-notification as executive policymaking; the matter awaits decision	Delhi Rent Control Act, 1958

The table’s third column is the argument of this article in miniature. Every event in the chronology, enactment, assent, agitation, amendment, litigation, repeal, replacement by model law, left the operative law exactly where it began. A reader who asks what law governed a Delhi tenancy in 1960, in 1996, in 2014 and in the years of the Model Tenancy Act receives one answer four times. Legislative history usually records change; here it records a change that was enacted and then declined, which is a different and rarer thing, and the reasons it was declined are on the public record because the government was made to state them.

The constitutional attack on the old Acts themselves

Running beneath the commencement litigation is a second line of cases in which the target is not the uncommenced replacement but the operating Act. The early challenges failed on the conventional ground: rent control in a scarcity was a reasonable classification and a reasonable restriction, and the Union’s submissions in the Delhi litigation could collect a line of Supreme Court authority upholding the classification the rent Acts draw. The landlords’ modern argument is built to escape that line rather than meet it, and its foundation is a doctrine the Supreme Court itself supplied in Motor General Traders v. State of Andhra Pradesh, (1984) 1 SCC 222: a provision that was valid when enacted may become invalid with the passage of time and the change of circumstances, because the factual basis that made the classification rational can cease to exist. Satyawati Sharma applied exactly that logic within the Delhi Act when it read down the residential limitation in the bona fide requirement ground, accepting that the distinction Parliament drew in 1958 had been emptied by the intervening decades. The landlords’ petition generalises the move from a clause to the code: if one provision can fall because its rationale has expired, a statute whose every operative figure is a figure of 1940 or 1958, whose protected class is defined by occupancy at dates two generations past, and whose temporary wartime justification the legislature itself recited, must answer for its rationale as a whole.

The Union’s defence, as recorded in the Delhi High Court proceedings, meets the argument at each step. On Article 19(1)(g), the submission runs that the Act restricts no one’s right to let premises or to carry on the business of letting; the complaint reduces to the proposition that letting under control is unprofitable, and reduced profit is not a restriction the Constitution forbids, the Act in any event permitting increase of rent by ten per cent every three years under Section 6A. On Article 14, the validity of the classification stands concluded by the Supreme Court’s repeated upholding of it, and a challenge to a classification already sustained is not maintainable merely because time has passed, the changed-circumstances doctrine being a ground for examining a provision whose rationale is shown to be gone, not a sunset clause on every old statute. On Article 21, the submission is blunt: the Act violates no one’s right to life. The petitions also carried the argument from the state’s side of the ledger, contending that the Act’s protection had been captured by tenants who needed it least, that the poor were not its beneficiaries in the old quarters, and that municipal finances and the physical fabric of the city bore the cost. The High Court’s decision of 7 January 2019 disposed of the batch without displacing the Act, and the constitutional position therefore stands where the pleadings left it: the changed-circumstances doctrine is established, its application to a provision has been demonstrated in Satyawati Sharma, and its application to an entire rent code awaits a court willing to hold that a legislature’s silence across decades can do what the legislature’s own repeal would not. The commencement affidavit described in the next section is best read as the executive’s answer to that awaiting question as well: so long as non-commencement is policymaking, the political branches retain both the frozen Act and its uncommenced replacement, and the constitutional attack must dislodge a settlement that every branch, in its own way, has chosen to leave standing.

The courts and the commencement question

The litigation began two years after the assent. Common Cause, the public interest organisation, filed Writ Petition No. 1495 of 1997 in the Delhi High Court seeking a mandamus directing the Union of India to issue the Section 1(3) notification and bring the Delhi Rent Act, 1995 into force in its enacted form. The High Court divided against itself in a manner that mapped the difficulty precisely. One judge would have directed the government to bring the Act into force within six weeks. A second held that only a limited mandamus could issue, of the kind recognised in Aeltemesh Rein v. Union of India, directing the government to consider whether the time to enforce the Act had arrived. The third judge, to whom the difference was referred, would have issued neither: since the government had already stated that it did not want the Act in its enacted form and intended to amend it first, even a direction to consider the question would be an infructuous writ, and courts do not issue writs of academic nature. A limited mandamus nonetheless issued on the majority view, the government filed an affidavit of compliance reporting that it had considered the matter, and the writ petition was disposed of, with a certificate for appeal to the Supreme Court.

The Supreme Court decided the appeals on 8 October 2003, in Common Cause v. Union of India, a bench of R.C. Lahoti and Ashok Bhan, JJ., Bhan, J. writing the judgment. The Court treated the question as concluded by the Constitution Bench in A.K. Roy v. Union of India, (1982) 1 SCC 271: an Act cannot be said to commence unless it is brought into operation by the authority the statute empowers, and where the enforcement of a statute is left to the discretion of the government without objective standards laid down, no mandamus can issue to enforce it. The appellant’s ingenious argument, that a commencement clause delegating a single appointed day without power to appoint different days for different provisions left Section 5 of the General Clauses Act, 1897 to operate, so that the Act commenced on the date of assent itself, did not survive that authority. What the record did show, and what plainly weighed with the Court, was motion of a kind: representations received, an All Party Committee appointed, an amendment bill introduced and examined by a standing committee whose recommendations the government had accepted. The government was not inactive; it was indefinitely in the antechamber of action, and the doctrine supplied no instrument for distinguishing the antechamber from the destination.

Has any other Indian statute waited decades for a commencement notification?

Yes, and the parallel is instructive because it ended. Section 30 of the Advocates Act, 1961, entitling an advocate enrolled in any state to practise throughout India, waited uncommenced for twenty-seven years before Aeltemesh Rein directed the government in 1988 to consider commencing it within six months. The notification came on 15 June 2011.

Aeltemesh Rein, (1988) 4 SCC 54, is the counterweight authority in this line, and its limits define the doctrine as precisely as A.K. Roy’s prohibition does. The Court there would not order commencement of Section 30 of the Advocates Act. It held, however, that a discretion conferred on the executive must be exercised in a just, fair and reasonable manner, and it directed the Central Government to consider within six months whether the time had come. In 2002, in Union of India v. Shree Gajanan Maharaj Sansthan, the Court returned to restraint, holding that where no objective standard governs the commencement discretion, a mandamus should not issue. The sequence yields the settled position: commencement is executive; the courts may require the executive to turn its mind to the question; the courts may not answer the question themselves. Whether a direction to consider, answered by an affidavit that records consideration and changes nothing, is a remedy or a ritual is a question the case law has never squarely faced, and the Delhi Rent Act is the case in which it bites hardest, because the consideration has been recorded, repeatedly, across three decades, and the answer has never varied.

What did the Union government tell Parliament when asked why the 1995 Act had not been notified?

It told the Lok Sabha, through the Minister of State for Urban Development, that the decision was to bring the Act into force after effecting amendments; that the steps taken included introducing the Amendment Bill, its examination by the Standing Committee and acceptance of the Committee’s recommendations; and that no time frame for introduction could be given.

That answer, given in response to a question about the grievances of property owners, is the executive’s public account of itself, and its structure deserves attention. Every step listed is a step toward amendment, not toward commencement; the Act as enacted disappears from the account entirely, replaced by an amended version that exists only as an intention; and the refusal of a time frame is stated as administrative candour rather than as a decision. Parliamentary committees returned to the subject without cracking it. The Committee on Subordinate Legislation in 1999 and the Committee on Petitions of the Rajya Sabha in 2004 both urged urgent steps to notify the Act, and the landlords’ later pleadings catalogued the assurances of successive governments alongside the committee directions, all unimplemented. A legislature’s committees may exhort; the notification remains the executive’s to give or withhold, and each exhortation that fails confirms the practical content of the power Section 1(3) confers.

The doctrine of non-commencement, stated as a position

This series has followed the notification gap from its first appearance. The same design choice, an appointed-day clause that hands commencement to the executive, is what left the 44th Amendment’s preventive detention safeguard unnotified after A.K. Roy itself, what kept the Whistle Blowers Protection Act, 2014 enacted but inert while an amendment bill travelled part of the way through Parliament and stopped, and what produced the staged and long-delayed commencements traced in the account of the labour codes’ implementation delay. In each instance the practice was documented; in none was it defended as a theory. The Delhi rent litigation is where the defence was finally written down, and that is this article’s namable claim: the doctrine of non-commencement, under which the executive’s power to appoint the day a statute begins has matured, through case law and through the government’s own pleaded position, into a power to decide whether an enacted statute shall operate at all, exercised without a duty to give effect to Parliament’s decision within any period at all.

The landlords renewed the attack on two fronts. The first was a direct constitutional challenge to the 1958 Act itself, pressed by an association of landlords led by the advocate Shobha Aggarwal, contending that a temporary wartime measure had become, through changed circumstances, an arbitrary and discriminatory regime offending Articles 14, 19(1)(g) and 21, that it subsidised tenants selected by historical accident rather than by need, and that its frozen rents had reached levels the petitioners illustrated by comparison with the daily minimum wage of an unskilled worker in Delhi. The Delhi High Court heard the challenge over years, at one stage remarking in the course of proceedings that under the Act tenants appeared to be smiling and landlords crying, and decided the batch on 7 January 2019, recording the government’s narrative of committees, drafts and the pending repeal bill without displacing the 1958 Act. The second front returns to commencement itself. In the proceedings that followed, the challenge to non-implementation was pressed again, and this time the Union of India set out its position in an affidavit in terms the brief for this article anticipated and the public record confirms: the decision whether and when to notify the 1995 Act involves balancing the interests of multiple stakeholders and falls within the executive’s policymaking domain; the court should exercise judicial restraint, since a direction to enforce the Act would offend the separation of powers; and non-notification of the Act violates no fundamental right of the petitioners. The matter awaits the High Court’s decision, and is described here as pending.

Stated at full strength, the Union’s position is a serious one, and its strongest plank is the one the columnist and advocate Lakshita Mohan, writing on the notification veto, concedes before answering: Parliament wrote Section 1(3). No one compelled the legislature to surrender the commencement date; it chose, with the text of A.K. Roy available to it, to hand the appointed day to the Central Government without a deadline, a default date or a duty to explain delay. An executive that does nothing under such a clause is exercising the discretion the clause confers, on the very question, timing, that the clause reserves to it, and a court that orders commencement is not enforcing Parliament’s will but overriding the allocation Parliament enacted. Add that commencement of a rent code is genuinely polycentric, implicating the readiness of the Rent Authority machinery, the treatment of lakhs of existing tenancies and the political settlement of a capital city, and the case for judicial restraint is complete on its own terms.

The objection, at full strength, begins where A.K. Roy itself began, because the Constitution Bench saw the destination and named it: Parliament, the Court observed, could not possibly have intended to confer on the government a sort of veto over its own legislation by the simple device of declining to bring a provision into force. A power to decide when is a power over scheduling; exercised without limit of time it becomes a power to decide whether, and a power to decide whether an enacted law shall operate is a legislative power, held by the branch the Constitution did not give it to, exercisable without debate, vote or assent, and unaccompanied by the repeal that would at least force the question into the open. The objection has acquired a doctrinal instrument since Shayara Bano v. Union of India, (2017) 9 SCC 1 recognised manifest arbitrariness as a ground for striking down legislation: the argument, pressed in commentary on the pending petition and acknowledged by its proponents to be an extension rather than an application of that case, is that a commencement mechanism whose practical operation across thirty years renders a legislative scheme wholly dormant is arbitrary in the same sense, whatever the clause’s formal validity. Between the two positions the law stands where A.K. Roy and Common Cause left it. The executive holds the power; the courts hold a remedy that requires the executive to consider; and the statute at the centre of the dispute continues to govern nothing at all.

The drafting remedies are not obscure, and commentary on the pending petition has assembled them. A commencement clause can carry a long-stop: the statute commences on the appointed day or, if none is appointed, on the expiry of a fixed period from assent, so that silence produces commencement rather than defeating it. A direction of the Aeltemesh Rein type can be given content by requiring the executive, after an inordinate interval, to place its reasons for continued non-commencement before the legislature whose Act waits, converting an affidavit in litigation into a report to the body that enacted the law. And repeal, honestly used, remains available: a government persuaded that an assented Act should never operate can ask Parliament to repeal it, as the Cabinet of 2013 in fact resolved to ask in this very matter, and take the political question to the forum the Constitution assigns it to. Each remedy preserves what is defensible in commencement discretion, the need for machinery, rules and preparation time, while denying it the one attribute the Delhi record exposes, the attribute of being answerable to no clock and no chamber. That no such amendment of practice has been adopted, in the forty years since A.K. Roy named the veto it feared, is itself part of the doctrine’s history: the beneficiaries of the power are the governments that would have to surrender it.

The landlord’s case and the tenant’s case, stated plainly

The landlord’s case rests on arithmetic and requires no rhetoric. A rent fixed by reference to the construction cost of a building erected before Independence, or to a rent recorded on 1 September 1940, and thereafter permitted to move by ten per cent every three years in Delhi or four per cent a year in Maharashtra, is not a low rent; it is a negligible one. The landlords who pressed the constitutional challenge in Delhi placed before the High Court the comparison their association had compiled: a large share of the properties under rent control in the capital fetched a monthly rent lower than the daily minimum wage of an unskilled worker. Against that receipt the landlord carries the obligations of ownership undiminished: municipal taxes assessed on the property’s real value, the cost of structural maintenance the statute lays on him, and the decay of a building whose income cannot fund its own repair. The predictable consequences have all been documented in committee reports and judgments across decades: buildings in the old quarters deteriorate because no one’s interest is served by maintaining them; owners who might let property keep it vacant or sell it rather than surrender it to a protected tenancy, withdrawing stock from a market short of homes; new rental construction aims exclusively at the exempt and informal segments; and the protection itself is distributed by the accident of who occupied what in the middle of the last century, sheltering prosperous trading houses in prime locations on the same terms as the vulnerable households the regime exists to protect. The Economic Administration Reforms Commission under L.K. Jha reported the negative effects of rent control on housing as early as 1982, and the landlords’ pleadings collected the endorsements of that view from committee after committee, including the Planning Commission’s own plan documents. On the facts of the frozen rent, the landlord’s case is as strong as the brief for this article directs it to be stated: it is very strong.

The tenant’s case rests on reliance, and it is equally real. The old commercial quarters of Delhi and Mumbai were not populated by squatters awaiting a market correction. They were populated by families and firms that paid for their occupation, in pagdi premiums and in key money and in the patient building of businesses at addresses whose value lay in continuity, under a statutory promise that the occupation was secure and heritable. A shop’s goodwill is its location embodied; the clientele of a trader in Sadar Bazar or Crawford Market attaches to the premises, and eviction or market rent destroys the business, not merely its cost base. The tenants who mobilised against the 1995 Act were defending capital they had paid for and livelihoods built in reliance on the law as it stood, and a legislature that had sold them that security, through the premiums the regime made rational, bears a responsibility for the reliance it created. The class point cuts the other way as well: beneath the prosperous trader of the landlords’ examples stands the genuinely poor tenant, the widow in a two-room tenancy, the pensioner in an old building, for whom the standard rent is the difference between a home and destitution, and who has no premium and no capital, only the statute. Reform that treats the two occupants as one class, because both hold protected tenancies, will ruin the second while correcting the windfall of the first.

The honest verdict, and it is the verdict this pillar records, is that the failure was one of design rather than of intention on either side. A transition was always available in principle: rents decontrolled in steps over a generation, protection withdrawn from premises and persons above defined thresholds while retained for the vulnerable, premiums compensated or amortised, heritability phased out for future successions while honoured for lives in being. Successive expert bodies sketched versions of such a path. No legislature enacted one, because a transition pleases no organised constituency in the present while its costs are immediate and its beneficiaries diffuse, and because the executive discovered, in the appointed-day clause, an instrument that let it accept the credit of enacting reform in 1995 while paying none of the price of commencing it. The freeze is not a policy anyone chose in the open. It is the residue of a choice no one would defend in the open, maintained by a doctrine defended in court as principle. The general account in this article is information about a body of law, not advice on any tenancy; which Act governs, which exemptions apply and which forum hears a dispute depend on the premises and the state, and the notice and limitation rules in eviction matters are unforgiving of delay.

Vacancy in a shortage: what the frozen stock does to a city

The aggregate consequences of the settlement are visible from any old quarter of Delhi or Mumbai, and they were catalogued in the record long before they were litigated. Buildings under frozen rents decay, because the income they produce cannot fund their maintenance and because neither party holds the incentive the other lacks: the landlord will not invest in an asset whose return the statute caps at a nominal figure, and the tenant will not invest in premises he does not own and cannot sell except by surrendering the tenancy that is his asset. Municipal corporations lose the house tax the landlords’ pleadings emphasised, since properties assessed on real values and occupied at controlled rents generate arrears their owners plead inability to pay, and tenants resist paying for fear that engagement with the tax record will disturb the tenancy. Owners respond rationally to the incentives the statute sets: a vacant premises is an appreciating asset fully in the owner’s control, while a let premises is the same asset transferred, in substance, to the tenant’s family at a frozen price, so owners keep homes empty in cities whose defining scarcity is homes, and successive censuses have recorded vacant houses in the millions nationally alongside official estimates of urban shortage. New supply routes around the protected sector entirely, into the licence market and the exempt categories, so that the formal protections govern a shrinking and ageing stock while the growing market organises itself outside the statute’s sight. Redevelopment, the one process that could renew the old quarters, stalls on the same rock: a building of pagdi and protected tenants cannot be rebuilt without each occupant’s price being met, and the Maharashtra Act’s conditions for rebuilding evictions, designed to return tenants to premises in the redeveloped building rather than extinguish their interest for cash, testify to how completely the tenancy had become a property right that redevelopment must buy out rather than a contract it could wait out. The frozen rent thus produces, in sequence, disrepair, vacancy, shortage and the political renewal of control, each stage supplying the argument for the next, which is why the settlement has survived every committee that condemned it.

Tamil Nadu: the state that actually replaced its Act

Against Delhi’s three decades of suspension, Tamil Nadu supplies the control experiment. The state legislated on the Centre’s draft model tenancy framework before the Model Tenancy Act, 2021 existed, and carried the replacement through. The Tamil Nadu Buildings (Lease and Rent Control) Act, 1960, Tamil Nadu Act 18 of 1960, was the state’s first-generation settlement, of the same family as Delhi’s 1958 Act. The Tamil Nadu Regulation of Rights and Responsibilities of Landlords and Tenants Act, 2017, Tamil Nadu Act 42 of 2017, received the Governor’s assent on 4 August 2017 and was brought into force on 22 February 2019, repealing the 1960 Act. The replacement is a working prototype of the model design. Tenancies of urban premises are to be created by written agreement, and the agreement is to be intimated to a Rent Authority, an officer not below the rank of Deputy Collector appointed by the Collector, which registers the tenancy and issues a tenancy registration number. A Rent Court and a Rent Tribunal stand above the Authority. The civil courts are barred from entertaining suits insofar as they relate to the Act’s provisions. Rent revision travels by written notice given three months ahead. The obligations of repair and maintenance are divided between landlord and tenant in the text, structural work falling to the landlord and the daily upkeep of fixtures to the tenant, and the withholding of essential supplies is made a matter for the Authority’s intervention rather than a weapon in the parties’ hands.

How did Tamil Nadu replace its rent control Act when Delhi could not commence its own?

By legislating for itself on the Centre’s model draft and accepting the political cost within its assembly. The 2017 Tamil Nadu Act built the Rent Authority, Rent Court and Rent Tribunal structure, required written agreements and their registration with the Authority, and came into force on 22 February 2019, repealing the 1960 Act outright instead of leaving a replacement uncommenced.

The Tamil Nadu scheme also writes its urgency into numbers, a discipline the older Acts gesture at and the model adopts. Under the Act’s design, the Rent Court is to decide the generality of eviction applications within ninety days of filing, and specified urgent categories within thirty days, timelines that function as a legislative confession about the parent regime, in which the same applications were measured in years and decades. The Rent Authority registers the agreement and issues the tenancy registration number within thirty days of the application reaching it, and maintains the particulars on the state’s tenancy portal, so that the existence of the tenancy is a matter of public record rather than of the parties’ competing recollections. A regime that began, in 1960, by fixing rents against the market had been replaced, in the same state and within the span of one Act’s life, by a regime that registers bargains and promises to adjudicate them in a season.

One feature of the Tamil Nadu statute deserves particular attention, because it marks the boldest divergence from the model that followed it. The 2017 Act did not leave pre-existing tenancies wholly in the old world. Section 4(2) required that where no written agreement existed for a tenancy created before the Act’s commencement, the landlord and tenant were to enter into a written agreement within ninety days of commencement, a window the legislature then extended by a further one hundred and twenty days by the amendment of 2019 after representations from both sides showed the machinery straining. Formalisation of the existing stock, in other words, was attempted, with deadlines that had to be extended because the stock was vast and the Authority new. The Model Tenancy Act, 2021, examined next, chose the opposite course and left existing tenancies entirely untouched. The comparison isolates the design question on which the practical fate of tenancy reform turns: a reform that governs only new bargains is easy to commence and slow to matter; a reform that reaches the existing stock matters at once and is correspondingly hard to carry. Tamil Nadu reached further and accepted the friction. The Union, for Delhi, would not commence even the reform that Parliament had enacted in prospective form twenty-six years before the model was approved.

The Model Tenancy Act, 2021: a statute that is not a statute

On 2 June 2021 the Union Cabinet approved the Model Tenancy Act, 2021, and on 7 June 2021 the Ministry of Housing and Urban Affairs circulated it to all states and union territories for adoption, by enacting fresh legislation or by amending existing rental laws to conform. The document must be described with precision, because the most common error in public discussion of it is to treat it as law. It is not an Act of Parliament. It was not introduced in either House, debated, passed or assented to, and it binds no landlord and no tenant anywhere in India. It is a model: a complete draft code, prepared by the Union government under its Entry 18 disability, offered to the legislatures that do hold the power, in the hope that enough of them enact it to produce something like national coherence without a constitutional amendment. Its provisions therefore describe the law of no place, and the law of any state that adopts it, in the form that state gives them.

The design repudiates the first-generation settlement at almost every point where that settlement was distinctive. Rent is not computed by formula; it is whatever the written agreement records, and revision of rent follows the terms the agreement itself specifies, the landlord being required to give notice three months before a revision takes effect, as the Ministry’s own summaries of the model describe the scheme. A written tenancy agreement is not optional: no person shall let or take on rent any premises except by an agreement in writing, and the parties must inform the Rent Authority of the agreement within two months of its execution, whereupon the Authority issues a unique identification number and places the tenancy on a public digital record. The security deposit, the unregulated toll of the informal market, is capped at two months’ rent for residential premises and six months’ rent for non-residential premises, refundable when vacant possession is taken over, after due deductions. Eviction during the currency of the agreement is barred except on the grounds the model lists, which track fault and necessity: refusal or failure to pay the agreed rent, arrears left unpaid after demand, parting with possession without the landlord’s written consent, misuse of the premises persisted in after written notice, and structural alteration without consent. The landlord’s self-help is disabled from the other side as well: no landlord or property manager may withhold an essential supply or service to the premises in an attempt to force the issue. And the overstaying tenant, the figure at the centre of the old system’s anxieties, meets the model’s sharpest provision: under its enhanced rent rule, a tenant who fails to vacate at the end of the tenancy or on its termination by order pays twice the monthly rent for the first two months and four times the monthly rent thereafter for as long as the occupation continues, a liquidated pressure designed to make holdover expensive rather than comfortable.

The adjudicatory design completes the repudiation. Disputes travel to a three-tier structure built for the purpose and standing outside the ordinary civil courts, whose jurisdiction over the model’s matters is barred. The Rent Authority, an officer not below the rank of Deputy Collector, is the first point of contact, registrar of agreements and authority for the run of applications. The Rent Court, presided over by an officer of the rank of Additional Collector or Additional District Magistrate, hears the contested matters and appeals from the Authority. The Rent Tribunal, presided over by a District Judge or an Additional District Judge, is the final appellate tier within the scheme. The stated design target, in the Union Housing and Urban Affairs Minister’s presentation of the model, was disposal of disputes within sixty days, a figure that functions as an indictment of the Controller litigation under the old Acts, in which an eviction application could consume a generation. The model also carves out categories of premises from its own scheme, among them premises owned by government and local authorities and premises belonging to religious and charitable institutions, leaving those occupations to the arrangements that already govern them.

The model’s allocation of the parties’ duties is written into the agreement the statute makes compulsory, and its terms invert the old Acts’ assumptions in a way worth stating. Structural repair is the landlord’s charge: the model places on the landlord the activities of keeping the premises sound, whitewashing of walls and painting of doors and windows, changing and plumbing pipes when necessary and the internal and external electrical wiring, while the tenant carries the daily register of drain cleaning, switches and socket repairs, kitchen fixtures, replacement of glass panels and the maintenance of gardens and open spaces, damage the tenant causes being the tenant’s own account. Sub-letting is not left to the general law: the model prohibits it unless a supplementary agreement permits it, and the sub-tenancy must then be jointly informed to the Rent Authority within two months of its execution, so that the derivative occupation enters the same public record as the original. The landlord’s right of entry is likewise converted from an incident of ownership into a term of the contract: the agreement must specify the reasonable causes for the landlord’s entry into the premises, and entry outside those causes is entry without right. Where a property manager interposes between owner and occupant, the model regulates the manager as it regulates the landlord, and the prohibition on withholding essential supplies binds the manager expressly. Read together, these provisions disclose the model’s theory of the transaction, which is the first-generation theory turned inside out. The old Acts distrusted the written bargain and replaced it with a status policed by a Controller. The model trusts the written bargain and polices the writing: get the terms recorded, register them with a public authority, price the breaches in advance, and adjudicate departures quickly in a forum built for nothing else.

A tenancy under the model generates a paper trail by design: the written agreement, its intimation to the Authority, the identification number, the receipts and notices by which revision and termination travel. For the landlord and the tenant alike, that trail is the tenancy’s legal substance, and keeping it complete, ordered and private across years is the discipline on which rights under the new regime will stand or fall; it is the same discipline by which journalists assemble and protect their source files on VaultBook, and a reader entering a tenancy under a state law built on the model will find that a single, private, well-kept file of the agreement, the Authority’s acknowledgment and every notice exchanged is worth more in a dispute than any recollection either party brings to the Rent Court.

Set beside the Delhi Rent Act, 1995, the model discloses a lineage its drafters did not advertise. The 1995 Act, drafted from the model legislation of the early 1990s, anticipated the 2021 design in its essentials: adjudication by a dedicated Rent Authority and Tribunal rather than the civil courts, registration of tenancies as a condition of the scheme’s order, heritability cut down from a perpetual family interest, and rent released from the 1940s arithmetic toward figures the market could recognise. The 2021 model adds what the intervening decades taught: the deposit cap against the informal market’s entry toll, the unique identification number and the digital register as the tenancy’s public face, the priced overstay in place of the interminable holdover, and the express disablement of self-help on both sides. The continuity matters for reading the model’s prospects. The ideas in it are not untested novelties awaiting a first trial; their Delhi prototype was enacted by Parliament a generation ago and defeated at the commencement stage, by the constituency the ideas threatened. A state enacting the model in the belief that the argument was won in 1995 will discover, as the Union discovered, that enactment is the stage at which the argument begins.

Adoption: the model meets the states

A model law’s biography is written in other legislatures, and this one’s is short. The Ministry of Housing and Urban Affairs has told Parliament, in answers given by the Minister of State for Housing and Urban Affairs, that four states have enacted or revised tenancy legislation on the model’s lines: Andhra Pradesh, Assam, Tamil Nadu and Uttar Pradesh. The enumeration requires the care the brief for this article demands, because the four names conceal four different relationships to the model. Tamil Nadu’s Act of 2017 precedes the 2021 model and follows its draft ancestor, and is a full replacement code in force since 2019. Uttar Pradesh enacted the Uttar Pradesh Regulation of Urban Premises Tenancy Act, 2021, replacing its own first-generation statute of 1972 with a code on the model’s pattern of written agreements, a Rent Authority and capped deposits. Assam legislated on the model’s lines in the same period. Andhra Pradesh revised its rental law framework to conform. Every other state and union territory, on the Ministry’s own account, has not followed, which means the map of tenancy reform in India consists of a small number of adopting states, of whom the earliest acted on a draft, surrounded by the unreformed first-generation Acts, of which Delhi’s remains the most litigated and the most frozen.

The reasons for the slow uptake are the reasons this article has traced throughout. A state legislature asked to enact the model is asked to spend the same political capital the Union declined to spend in Delhi for thirty years, against tenant constituencies that are organised and landlord constituencies that are not, in exchange for benefits, a functioning formal rental market, unlocked vacant stock, housing for migrants and young workers, that are real but diffuse and deferred. The Union’s own stated objectives for the model were framed in exactly those market terms: to create an adequate rental housing stock for all income groups, to institutionalise rental housing as a business, to unlock vacant houses whose owners fear the old Acts too much to let them. Each objective concedes the diagnosis of the first-generation regime. None of them compels a state to act, and the constitutional allocation ensures that nothing can. The same urban land scarcity that the urban land ceiling experiment tried to legislate away by capping ownership presses on the rental market from the other side: in cities where land is the scarcest of goods, a law that makes letting it hazardous suppresses the supply of homes precisely where demand concentrates, and the shortage then renews the political case for the controls that caused it. Rent control’s trap is circular, and the model’s fate in the states is the circle turning.

The non-adopting majority includes the jurisdictions where the frozen settlement is deepest, Maharashtra among them, and the reasons visible in its record generalise. Maharashtra legislated its own settlement in 1999, within the memory of the Malpe finding, and a legislature that has already paid the political price of one rebalancing, however modest, does not volunteer to pay it again for a central model whose marginal gain over its own Act it can plausibly dispute: the written agreement, the registered licence and the legalised premium are already its law, and what the model adds, the deposit cap, the priced overstay and the three-tier forum, must be weighed against reopening every settlement its tenants organised to defend in 1999. Multiply that calculation by the number of state capitals in which the old quarters vote, and the adoption map explains itself without recourse to inertia or ignorance. The model’s four adopters are not the states that understood it best; they are the states where the local constellation made its price payable, and the map will change, if it changes, one constellation at a time.

Delhi’s own position in the adoption story is the paradox that completes it. The model route is designed for state legislatures, and Delhi has no legislature competent to enact a tenancy code for the capital on this subject in place of Parliament; the adopting authority for Delhi is Parliament itself, the very body that enacted the replacement of 1995. For the capital, therefore, adoption of the model and commencement of the 1995 Act are the same political act wearing two different drafting histories, and neither has been performed. The government that commends the model to twenty-eight states and eight union territories governs its own capital city under a rent Act of 1958 that its own Parliament replaced before some of the model’s drafters were born, and defends that arrangement in court as policymaking. Whatever else the adoption map records, it records that the reform’s difficulty was never located in the states.

What the model changes, and what it cannot reach

The model’s most consequential sentence is its quietest: it applies prospectively, and it does not affect existing tenancies. The Union government said so expressly when the Cabinet approved the model, and the design is deliberate and defensible. To reach into existing tenancies would be to reopen every reliance interest this article has described, to invite a decade of constitutional challenge on the Satyawati Sharma logic in reverse, and to guarantee the mobilisation that buried the Delhi Rent Act, 1995 in every adopting state at once. A model that no state dares enact reforms nothing; a model confined to new tenancies at least begins. The price of the choice is written into the same sentence. The frozen stock, the lakhs of tenancies at standard rents computed from 1940 and 1944, the heritable occupations, the pagdi premiums, the decaying buildings of the old quarters, lies outside the model’s reach entirely, governed by the first-generation Acts until those tenancies end by the slow attrition of the statutory grounds. India therefore acquires, in the adopting states, two rental laws running in parallel for a generation: a modern code for the new letting, written, registered, market-priced and quickly adjudicated, and a museum code for the old, frozen at figures from the age of the telegram. The boundary between them is the date of the tenancy, and every inequity the boundary creates, neighbours in one building paying rents a hundredfold apart for identical premises, is the price of the transition mechanism the reformers of 1995 could not design and the reformers of 2021 declined to attempt.

For the tenant of the new market, the migrant worker, the young professional, the family that moves to a city for a posting or a school, the model’s payload is not the deposit cap alone, considerable as that is against the entry tolls of the informal market. It is the registered tenancy itself. A written agreement, informed to a public authority and carrying an identification number, is evidence of residence that the informal letting never supplies: the tenant of the licence market holds a document drafted to deny that he is a tenant at all, revocable in form and precarious in substance, while the tenant of the model regime holds a registered place in a public record, a rent fixed by an agreement a tribunal will enforce, and protection against eviction and against the cutting of supplies for the term he has bought. The Union’s stated case for the model dwelt on unlocking stock for owners; the symmetrical benefit for occupants is the conversion of urban renting from a tolerated informality into a documented civil status, and it is the benefit most likely to be felt first where adoption occurs, because a generation of urban Indians rents without ever having held a document that calls them a tenant.

The landlord’s payload under the model deserves the same concrete statement, because the reform’s political premise is that owners must be given reasons to let. Under the code a state enacts on the model’s lines, the landlord recovers the incidents the first-generation Acts suspended: a rent set by his own bargain and revised on the agreement’s terms, an occupation that ends when the term ends, with the overstay priced at twice and then four times the rent rather than rewarded with a protected succession, an eviction remedy on stated grounds through a forum designed to decide in sixty days rather than a Controller’s list measured in years, and protection against the tenancy hardening into an inheritance, the model’s tenancy being the parties’ contract and not a family estate. The model is not a landlord’s charter in disguise, and its balances are real: the deposit is capped, the supplies may not be cut, eviction mid-term requires a ground and an order, the deposit must be refunded on vacant possession with deductions accounted, and a landlord who delays a refund the model requires owes interest on the sum withheld at the rate the state prescribes. The symmetry is the design’s claim to durability. The old Acts purchased the tenant’s security with the landlord’s property and paid for it, over decades, in vanished supply; the model attempts to purchase both parties’ security with procedure, writing, registration and speed, and to charge neither side’s capital for it. Whether procedure can bear that weight, in districts where the Rent Authority is one officer among his other duties and the Rent Court’s sixty days meet the incentives that made the Controller’s list eternal, is the question adoption will answer state by state.

For a landlord or a tenant, the practical payload is consequently a question before it is an answer: which regime governs this tenancy? In a non-adopting state, the first-generation Act governs old and new tenancies alike within its exemptions, and the model is a policy document of no legal consequence to the parties. In an adopting state, a tenancy created under the new code is governed by its written agreement, its deposit cap, its Rent Authority and its enhanced rent rule, while the tenancy across the landing, created before the code, continues under the old Act’s standard rent and closed list of grounds. In Delhi, uniquely, the question has a third layer: the governing Act is the Act of 1958, the enacted replacement of 1995 waits uncommenced, and the model of 2021 has not been enacted for the capital whose own replacement Parliament passed a generation ago. The reader who wants to know what the law requires of a particular letting must therefore begin with the state, the date and the premises, in that order, and any account of Indian rent law that skips those three questions, including most of what circulates about the Model Tenancy Act as though it were a national statute in force, is wrong in a way that costs people money.

The transition that was never designed

It remains to state, with the specificity the subject usually avoids, what a designed transition out of the frozen settlement would have had to do, because the absence of that design is the article’s verdict and it should not be allowed to rest as a slogan. Four problems required solving together. The rent problem required a path from the standard rent to a defensible figure that did not arrive as a single shock: staged decontrol of the kind the 1988 Delhi amendment gestured at with its rupee threshold and its decennial exemption for new buildings, except indexed, so that the path could not be silently repealed by inflation. The premium problem required the law to decide what a pagdi payment buys under the new order: if the tenant’s capital, paid for perpetual security, is to be extinguished by decontrol, the transition must either honour it in possession for a term, amortise it against rising rents, or compensate it, and each choice has a payer. The succession problem required heritability to be phased rather than abolished or perpetuated: honoured for the generation in occupation, closed to the next, so that the grandfather’s tenancy ends as a family interest with the grandchildren the statute names as its last holders. The vulnerability problem required the protection withdrawn from the prosperous to be retained, expressly and permanently, for the occupants the regime was built for, by means-tested or premises-tested lines of the kind the exemption clauses already draw in reverse. None of these devices exceeds ordinary legislative craft; versions of all four appear in committee writing across the decades this article has surveyed. What no version survived was the encounter with the interests: staged decontrol offers tenants a dated loss to mobilise against, compensation for premiums presents the exchequer with a bill for the capitalised value of the entire frozen stock, phased heritability satisfies neither the landlord who wants his building nor the tenant who reads any sunset as dispossession, and means-tested retention demands an administrative inquiry into lakhs of households that no government wished to conduct. The executive’s alternative, discovered in the appointed-day clause, solved every one of these problems by declining them, and the Model Tenancy Act’s clean prospectivity is the same solution in modern dress: the transition designed, at last, by designing the existing stock out of the reform. Naming that choice accurately is not a counsel of despair. It is the precondition of any future reform that intends to govern the tenancies that exist rather than the tenancies that would be convenient.

Adoption, where it happens, is also only the first of two labours, and the second is the one the Delhi record teaches observers to watch. A state that enacts the model must still appoint its Rent Authorities district by district, constitute its Rent Courts and Tribunals in consultation with its High Court, frame its rules and forms, and build the digital platform on which agreements are to be recorded and identification numbers issued; Tamil Nadu’s experience, with its tenancy portal and its registration machinery built alongside the Act’s commencement, and with the window for reducing old tenancies to writing extended when the machinery strained, shows the labour’s scale even where the political question was answered. A commencement notification for the new state Act, rules notified under it and authorities actually appointed are the operative events, and a state can perform the enactment while postponing every one of them, reproducing within its own borders, in miniature, the structure of the Delhi suspension. The reader who wishes to know whether the model governs a tenancy in an adopting state must therefore ask not only whether the state enacted it but whether the state commenced it and built it, three questions where public discussion habitually asks one.

A reader’s map: identifying the law that governs a tenancy

The doctrine and the history resolve, for the person holding a notice or drafting an agreement, into an order of questions, and the order matters more than any single answer. The first question is the state, because the subject is a state subject and no instinct formed under one Act travels safely into another: the Delhi tenant’s standard rent, the Maharashtra tenant’s pagdi and the Tamil Nadu tenant’s registration number are creatures of three different statutes, and the Model Tenancy Act is the law of none of them until a legislature makes it so. The second question is the premises and the letting’s date and form, because within a single state the exemptions and the instrument divide the field: in Delhi, whether the rent exceeds the three thousand five hundred rupee line of the 1988 exemption and whether the building is within its ten protected years; in Maharashtra, whether the occupation rests on a registered tenancy, a registered leave and licence, or an unregistered arrangement in which the occupant’s version of the terms prevails; in an adopting state, whether the tenancy was created before or after the new code commenced, the boundary on which the frozen and the market regimes divide. The third question is the forum, which the answers to the first two determine: the Controller in Delhi, the designated courts in Maharashtra, the Rent Authority and Rent Court in Tamil Nadu, each with its own appellate ladder and each closed to the others’ business. The documents that decide disputes are the documents the regimes themselves privilege, the fixation order and the rent receipt under the old Acts, the registered agreement and the Authority’s acknowledgment under the new, and the party who kept them holds the tenancy’s history in admissible form. A reader who works through the three questions in that order will locate any Indian tenancy in its governing regime; a reader who begins, as most public discussion begins, with the Model Tenancy Act’s provisions will mislocate most of them. The map has one further use, which is diagnostic: a tenancy that cannot be located on it, because the letting was oral, the instrument ambiguous and the premises’ status contested, is a tenancy whose parties will spend their first years of dispute establishing which law they live under, at prices neither budgeted for.

Verdict: the freeze, the veto and the model

The pillar’s assessment can be stated in three sentences. The first-generation rent Acts solved a real emergency with a formula that had no self-correcting mechanism, and legislatures that would not pay the political price of correction allowed the formula to harden into a transfer of urban property value from owners to occupiers, distributed by historical accident, corrosive of maintenance and of rental supply, and condemned as unreasonable in its effects by the Supreme Court itself in the Maharashtra litigation. The remedy the Constitution appears to offer, legislation, was supplied in Delhi in its most complete form and nullified by the appointed-day clause, and the executive’s defence of that nullification, that commencement is policymaking, that judicial direction would breach the separation of powers, and that no fundamental right is engaged, stands as the articulated doctrine of non-commencement at the head of a line of authority running from A.K. Roy through Common Cause. The Model Tenancy Act is a competent draft of the replacement settlement, market rent by agreement, writing and registration, capped deposits, specialised and speedy forums, priced overstay, and its confinement to new tenancies and to the states that choose it means that the question this body of law poses has not changed since 1995: not what the replacement should say, which has been known for thirty years, but who will pay the political price of switching it on. The specialist articles in this cluster take up the threads this pillar has deliberately left at summary depth, the passage and afterlife of the individual state settlements, the constitutional litigation over commencement across the statute book, and the general law of leases beneath the rent Acts. The pillar’s closing observation belongs to the record rather than to argument. A tenancy in Delhi is governed by an Act of 1958 because an Act of 1995, passed unanimously and assented to by the President, has been waiting for a notification for over thirty years, and the government of India has explained to a court, in writing, why it is entitled to keep it waiting. Until a legislature designs the transition the record calls for, or an executive accepts the price of the commencement its own Parliament enacted, the frozen settlement will continue to govern the old quarters of India’s cities, and the replacement law will continue to be written, approved, circulated and praised everywhere except in the one place that matters, which is force.

Frequently Asked Questions

Q: Who makes rent control law in India?

State legislatures do. The relation of landlord and tenant, including the collection of rents, falls within Entry 18 of the State List in the Seventh Schedule to the Constitution, the entry for land, so each state enacts its own rent control statute with its own definitions, rent fixation machinery and eviction grounds. There is no central rent control Act governing the states. The Union legislates for the National Capital Territory of Delhi, which is why Parliament enacted the Delhi Rent Control Act, 1958 and its uncommenced replacement of 1995, and the Union also prepares model legislation, such as the Model Tenancy Act, 2021, which states are free to adopt, adapt or ignore.

Q: What is standard rent and how is it fixed?

Standard rent is the maximum rent lawfully chargeable for premises under a first-generation rent control Act, computed by statutory formula rather than by the parties’ bargain. For older premises it is built on a historic base, in Delhi the basic rent determined under the Act’s Second Schedule for premises let before 2 June 1944, in the old Bombay legislation the rent payable on 1 September 1940. For other premises the Controller fixes it having regard to the rents of similar premises in the locality and to the cost of construction and the value of the land when the building went up. Permitted increases are then added at the modest rates the Act allows.

Q: On what grounds can a landlord evict a tenant under rent control law?

Only on the grounds the governing Act lists, proved before the statutory forum. Under Section 14 of the Delhi Rent Control Act, 1958 the principal grounds are non-payment of arrears left unpaid within two months of a notice of demand, subletting or parting with possession without the landlord’s written consent, misuse of the premises for a purpose other than the letting, non-residence in residential premises for six months, substantial damage, premises required for repairs or rebuilding that cannot proceed while occupied, and the landlord’s bona fide requirement of the premises for his own or his family’s occupation. A ground not on the list, however genuine the landlord’s grievance, supports no eviction.

Q: Why was the Delhi Rent Act 1995 never brought into force?

Because the Union government decided not to notify it in its enacted form. The Act, passed unanimously by both Houses and assented to on 23 August 1995, left its commencement under Section 1(3) to a government notification. Trader tenants agitated against its provisions on deemed rent, registration of tenancies, restricted inheritability, rent enhancement and eviction, and an All Party Committee examined their grievances. The government resolved to amend the Act before commencing it; the Amendment Bill of 1997 stalled and lapsed, a repeal bill of 2013 was never passed, the Supreme Court in Common Cause declined to order commencement, and the notification has never issued.

Q: What does the Model Tenancy Act 2021 propose?

It proposes a complete replacement tenancy code for states to enact. Every tenancy is to be created by a written agreement, informed to a Rent Authority within two months, with rent set by the agreement rather than by formula. Security deposits are capped at two months’ rent for residential premises and six months’ rent for non-residential premises. Eviction during the agreement is confined to listed grounds such as non-payment, unauthorised parting with possession and misuse, a tenant who overstays pays enhanced rent, and disputes go to a dedicated Rent Authority, Rent Court and Rent Tribunal outside the ordinary civil courts, designed for disposal within sixty days.

Q: Is the Model Tenancy Act binding on states?

No. Rent control falls within the legislative competence of the states under Entry 18 of the State List, so the Union could not enact a tenancy code for them. The Model Tenancy Act, 2021 is a draft approved by the Union Cabinet on 2 June 2021 and circulated to states and union territories on 7 June 2021 for adoption by fresh legislation or by amendment of existing rental laws. It has no legal force anywhere by itself. A state’s tenants and landlords are governed by it only if, and in the form in which, their own legislature enacts it, which is why its practical reach depends entirely on state adoption.

Q: Does the Model Tenancy Act apply to existing tenancies?

No, and the exclusion is the model’s most consequential design choice. The Union government stated on its approval that the model applies prospectively and does not affect existing tenancies. Tenancies already running under a state’s first-generation rent Act, with their standard rents, closed eviction grounds and heritable occupation, remain under that Act in any state that enacts the model; the new code governs tenancies created after the state law commences. The consequence is two parallel regimes in adopting states for a generation, a market-priced written regime for new lettings and the frozen settlement for the old stock, separated only by the date the tenancy began.

Q: Is a tenancy heritable under Indian rent control law?

Yes, under the first-generation Acts. The Delhi Rent Control Act, 1958 extends its definition of a tenant, on the tenant’s death, to specified members of the tenant’s family in a stated order, broadly the spouse, children, parents and a daughter-in-law, who were ordinarily living with the tenant. In Gian Devi Anand v. Jeevan Kumar, (1985) 2 SCC 683, the Supreme Court held that a statutory tenancy protected by the Act is heritable in the same manner as a contractual tenancy, so the landlord cannot treat the tenant’s death as ending the tenancy. Heritability converts a time-limited letting into a family interest and is the feature landlords most strongly resent.

Q: What does the Model Tenancy Act say about security deposits?

It caps them for the first time in a national template. Under the model, the security deposit payable in advance by the tenant may not exceed two months’ rent for residential premises and six months’ rent for non-residential premises. The deposit is to be refunded by the landlord when the tenant hands over vacant possession, after making due deductions for the tenant’s liabilities. The cap addresses one of the informal market’s heaviest entry costs, the large advance deposits landlords demand where no statute restrains them, and it operates only where a state has enacted the model, since the cap is a provision of a draft until a state legislature gives it force.

Q: What happens if a tenant refuses to vacate after the tenancy ends under the Model Tenancy Act?

The model prices the overstay instead of merely prohibiting it. A tenant who fails to vacate at the end of the tenancy, or on its termination by an order, is liable to pay the landlord enhanced rent at twice the monthly rent for the first two months of the overstay and four times the monthly rent for every month thereafter until possession is given up, subject to a short allowance where a force majeure event such as flood, fire or earthquake has affected the locality. The landlord may also pursue eviction through the Rent Authority and the Rent Court on the model’s grounds. The design replaces the old system’s indefinite protected holdover with an escalating financial pressure to leave.

Q: What is the pagdi system, and why does it survive in Mumbai?

Pagdi is the premium economy that grew around frozen rents in Mumbai under the Bombay legislation of 1947 and its successor, the Maharashtra Rent Control Act, 1999. Because the monthly rent of a protected tenancy is fixed at a historic figure far below market value, the occupation itself carries a large capital worth, and an incoming tenant pays a substantial premium, to the outgoing tenant or the landlord, to acquire it, the landlord customarily taking a share when a transfer is recognised. The system survives because the legal rent cannot move: value that cannot be collected monthly is collected once, in capital, and the tenant who has paid it holds the tenancy as an asset to occupy, bequeath in practice and sell on departure.

Q: What does a Rent Authority do under the Model Tenancy Act?

The Rent Authority is the model’s first-tier institution, an officer not below the rank of Deputy Collector for the area. Its core function is registration: every tenancy agreement must be informed to the Authority within two months of execution, and the Authority issues a unique identification number to the parties and maintains the record on a digital platform, giving the tenancy a public, verifiable existence. The Authority also receives and decides the run of applications the model assigns to it, including matters arising from the withholding of essential supplies and the deposit of rent, and its orders are appealable to the Rent Court, with a further appeal to the Rent Tribunal presided over by a District Judge.

Q: How can rent be increased under the Model Tenancy Act?

By the agreement, and only as the agreement contemplates. The model abandons formula fixation altogether: the rent is the figure the written tenancy agreement records, and revision of rent is governed by the terms and periods for revision that the agreement itself must specify. A landlord proposing a revision is required to give the tenant notice three months before the revised rent takes effect, as the Ministry of Housing and Urban Affairs describes the scheme in its summaries of the model. There is no statutory percentage cap of the kind the old Acts impose, and equally no unilateral mid-term increase: if the agreement does not provide for a revision, the rent recorded is the rent payable for the term.

Q: Did the Supreme Court direct the government to bring the Delhi Rent Act 1995 into force?

No. In Common Cause v. Union of India, decided on 8 October 2003, the Supreme Court applied the Constitution Bench ruling in A.K. Roy v. Union of India, (1982) 1 SCC 271, and declined to issue a mandamus for commencement. Where a statute leaves its enforcement to the discretion of the government without laying down objective standards, the Court held, no writ can direct the government to enforce it. The Delhi High Court before it had divided three ways, and the most that emerged from the litigation was a limited direction that the government consider whether the time to commence had arrived, a direction the government answered with an affidavit recording that it had considered the matter and preferred amendment first.

Q: What happened to the Delhi Rent (Amendment) Bill 1997 and the Delhi Rent (Repeal) Bill 2013?

Both died unpassed, leaving the uncommenced Act of 1995 untouched on the statute book. The Amendment Bill, introduced in the Rajya Sabha on 28 July 1997 to modify the Act before its commencement, was examined by the Parliamentary Standing Committee, which reported in December 2000; the government accepted the Committee’s recommendations on 3 April 2001 and gave notice for moving the official amendments, but the Bill was never taken through and in time lapsed. The Repeal Bill, Bill No. LIX of 2013, introduced in the Rajya Sabha on 29 August 2013 after the Cabinet approved repeal of the 1995 Act, was not referred to a committee and was never passed either.

Q: Which states have adopted tenancy laws on the lines of the Model Tenancy Act?

On the account the Ministry of Housing and Urban Affairs has given Parliament, four: Andhra Pradesh, Assam, Tamil Nadu and Uttar Pradesh. The four cases differ. Tamil Nadu’s Regulation of Rights and Responsibilities of Landlords and Tenants Act, 2017 follows the model’s draft ancestor and has been in force since 22 February 2019. Uttar Pradesh enacted a tenancy Act of 2021 on the model’s pattern, replacing its 1972 statute. Assam legislated on the model’s lines, and Andhra Pradesh revised its rental framework to conform. Every other state and union territory, on the Ministry’s account, retains its earlier rental law unreformed, so the model’s map remains a small set of adopting states within an unreformed majority.

Q: How much can rent rise each year for premises under the Maharashtra Rent Control Act 1999?

Four per cent a year, and the landlord needs no permission to take it. Section 11 of the Maharashtra Rent Control Act, 1999 entitles the landlord to increase the rent of premises by four per cent per annum from the Act’s commencement, computed year by year. Further increases are available on defined accounts, including a reasonable addition where the landlord has carried out improvements or special additions, and recovery of increases in taxes levied on the premises. The entitlement sits alongside the Act’s standard rent ceiling: Section 10 makes it an offence, punishable with imprisonment up to three months or a fine up to five thousand rupees or both, to claim or receive rent above the standard rent and the permitted increases.

Q: Why did the Supreme Court not strike down the frozen standard rent provisions in the Malpe Vishwanath Acharya case?

Because invalidation would have abolished tenant protection across Maharashtra at a stroke, and the State was already framing a replacement law. In Malpe Vishwanath Acharya v. State of Maharashtra, (1998) 2 SCC 1, a three-judge bench held that the Bombay Act’s provisions for determining and fixing standard rent were no longer reasonable, the fixation having become unjust as decades passed and property values transformed around a 1940 base. Having found the provisions indefensible in their operation, the Court nevertheless refrained from striking them down, allowing the legislature the opportunity to enact the modern law it was preparing, which arrived as the Maharashtra Rent Control Act, 1999.

Q: Does the Model Tenancy Act require every tenancy agreement to be in writing?

Yes. The model’s opening design choice is that no person shall let or take on rent any premises except by an agreement in writing. The agreement must record the rent payable, the period of the tenancy, the terms and period for revision of rent, the security deposit, the grounds on which the landlord may enter the premises and the parties’ maintenance responsibilities, and the landlord and tenant must jointly inform the Rent Authority of the agreement within two months of its execution. A purely oral letting has no place in the scheme. The requirement binds only where a state has enacted the model; in states that have not, oral and informal tenancies continue under their existing law.

Q: Can a landlord cut off essential services to force a tenant out?

No. Self-help of that kind is unlawful under both the old settlement and the model. The Delhi Rent Control Act, 1958 protects the tenant’s enjoyment of essential supplies and services, and a landlord who cuts off or withholds them to compel eviction exposes himself to orders for restoration and to penalty under the Act. The Model Tenancy Act, 2021 carries the prohibition into the new design: no landlord or property manager may withhold any essential supply or service to the premises occupied by the tenant, and the Rent Authority is the forum to which the tenant carries the complaint. A landlord’s remedy for default is an application on a statutory ground, never the water tap or the electricity meter.