Once a bill gets the President's assent, most people assume that's the end of the story. Fewer stop to ask the second question: when does it actually start applying? For at least 3 Indian statutes, no one has given a real answer to that in years. In one case, it's been more than 3 decades.
Take the Delhi Rent Act. The President assented to it on August 23, 1995, but it has remained uncommenced ever since. This is because the decision on this lies with a Central government notification under Section 1(3), which has not been made in over three decades now. Advocate Shobha Aggarwal took the matter to court in 2025.
Parliament passed the Act. It was signed by the President; then it lay there, as nothing in the statute obliged the government to justify itself in any time frame. The Delhi Rent Act, 1995 is, therefore, now an Act that is on the statute book but not implemented to this day, waiting for a notification “on such date as the Central government may appoint.”
And this is not an isolated incident. Another example of such a law is the National Environment Tribunal Act, 1995, which met the same fate and was finally done away with by Parliament through a new National Green Tribunal Act in 2010. There is also the Whistleblowers Protection Act of 2014, which became law on May 9, 2014 and has never been operational. According to a December 2025 parliamentary reply, the Act must be changed so that disclosures do not threaten India’s sovereignty and security. The Amendment Bill, introduced in 2015, passed the Lok Sabha that year but could not reach the Rajya Sabha before the 16th Lok Sabha was dissolved in 2019.
This pattern repeats across the 3 statutes: Parliament passes the Act the President signs the Act and the executive simply forgets to enforce the Act. What I refer to as a “notification veto” is not a legal theory, but merely a description of the constitutional impact of the failure to commence legislation.
For more than 4 decades, the Supreme Court has grappled with this problem without answering its hardest version: when does commencement discretion stop being discretion and turn into indefinite refusal?
Look back on AK Roy v. Union of India, delivered in December 1981. The Court declined to direct the government to notify a provision of the 44th Amendment, holding the discretion to be executive, not judicial. Yet, the same judgment had foreseen the danger, observing that “Parliament could not possibly have intended to grant Government a sort of veto simply by declining to bring a provision into force.” As early as 1981, the Court saw this coming. It never provided a solution.
Then, in 1988, Aeltemesh Rein v. Union of India took up Section 30 of the Advocates Act, 1961, still uncommenced 27 years after it received assent. Once again, the Court would not order commencement directly. It merely ordered the government to seriously review, within 6 months, whether sufficient time had elapsed, laying down a principle that has remained in force since: such discretion must be exercised in “a just, reasonable and fair manner.”
Things swung back toward restraint with Union of India v. Shree Gajanan Maharaj Sansthan in 2002, where the Court held that mandamus should not issue if there is no objective standard for commencement discretion. Common Cause v. Union of India, decided in 2003, dealt with this same Delhi Rent Act and took the same reserved approach, but only after being convinced that something was going on. Representations were being taken, an all-party committee formed an amendment bill drafted and referred to a standing committee. Not inaction; just slow action.
This takes us back to Shobha Aggarwal's petition, which, 23 years after Common Cause, raises essentially the same question: can “slow” continue endlessly? The plea argues that abeyance of a law for over 30 years defeats legislative purpose, citing a 2004 Rajya Sabha Committee observation that had already called such delay undesirable. The government's May 2026 affidavit maintains that notification is a policy decision and that judicial interference would violate the separation of powers.
This is a distinction case law has never quite pulled apart cleanly. There is a legitimate case for commencement discretion: the government may genuinely need time to develop regulations before switching on and Parliament can provide for that.
But there is another variant that looks the same on the surface and acts nothing like the first: a law left uncommenced for decades, each government explanation sounding reasonable alone, until the reasons quietly add up to a functioning refusal to commence the law at all. The possibility to decide when something is to begin, if it has no outer restriction, tends to morph into the possibility to decide whether something will begin at all. This is where 40 years of case law have been pointing, without ever stating it explicitly.
The government's best defence of keeping everything just as it is is not a cynical one and it deserves fairness. Parliament enacted Section 1(3) in its current form; nobody compelled it to surrender commencement to the Central government; it did so willingly. The executive is, therefore, not flouting Parliament's wishes when it does nothing about the notification.
Delegating discretion over timing is not necessarily the same as delegating an unfettered discretion to freeze the core policy in place indefinitely, across changes of government. Whether timing can be allowed to covertly become “never” is an entirely different question that AK Roy itself posed in no uncertain terms.
Shayara Bano v. Union of India, decided in 2017, is where manifest arbitrariness first appeared as a ground for invalidating primary legislation. But that case concerned the content of an existing provision, not a method that prevents a legislation's operation altogether.
There is a tougher issue worth considering. Can such logic apply to a commencement mechanism whose practical consequence, after an unreasonably long period, is to render a legislative scheme totally dormant? It would amount to expanding the precedent of Shayara Bano rather than following it and should be acknowledged as such. If manifest arbitrariness invalidates the content of legislation, why can't the validity of its operation be questioned as well, under Article 14?
No reasonable person wants judges deciding when commencement should take place. That's all AK Roy and Common Cause were right about. But a court could still ask a narrower question: once a statute has remained uncommenced for an inordinate time, does the government have a genuine, live justification or is it reaching for an old excuse because asking costs nothing? 3 years is suggested here as a workable trigger, not a deadline, and no guarantee of automatic commencement. Inordinate delay would create an obligation to explain, not to commence. The burden shifts to the government and the remedy, in line with Aeltemesh Rein, would be a direction to reconsider within a fixed period and put its reasoning in writing.
A simpler fix lies with Parliament: instead of leaving future commencement provisions open-ended, build in an automatic commencement date after a fixed period, unless the government notifies earlier. That is a drafting solution, not a constitutional one.
Successive governments have all been left with the same unanswered question: whether and when to give force to these Acts. Individuals who might have relied on whistleblower protection since 2014 have been denied its benefits. Tenants and landlords have lived for 30 years under a 1958 law that Parliament itself determined, in 1995, was obsolete.
A statute passed by Parliament and assented to by the President cannot be left indefinitely dormant, dependent on the executive's whim to decide whether to implement it. Every case of non-commencement starts out as a straightforward issue of when. Non-commencement may continue as long as the government keeps offering reasons that barely pass judicial muster until “when” turns into “never.”
Lakshita Mohan is an advocate practising at the Delhi High Court.
Views are personal.