Nine Judges Have Been Asked to Abolish Public Interest Litigation. History Says Otherwise.

On 3 September 2026, a Bench of Justices P.S. Narasimha and Alok Aradhe threw out an application in Aditi v. National Board of Examination in Medical Sciences asking for the NEET-PG examination to be held again across the country. It imposed twenty-five thousand rupees in costs, and said it would write to the Bar Council of India to act against the advocate who had filed it. The Bench told him: “It appears this person has become a full-time PIL litigant. Wherever exams are conducted, you go on to file petitions.”¹

Four months earlier, on 5 May 2026, nine judges of the same Court had spent an afternoon on a different question. They were hearing the Sabarimala reference. In front of them was the association whose petition had produced Indian Young Lawyers Association v. State of Kerala, 2018 INSC 908, the judgment permitting women of all ages into the temple. The Chief Justice of India, Surya Kant, said of that petition: “And we entertain a PIL based upon this kind of document? It should have been thrown outright in the dustbin.” Justice B.V. Nagarathna added that public interest litigation had become “private interest litigation, publicity interest litigation, paisa interest litigation, and political interest litigation.”²

Set those two days side by side and you have the argument of this essay. The jurisdiction that India invented, and that gave us most of what Article 21 now means, is being narrowed from both ends at once. At the bottom, by money: dismissal now routinely comes with a bill, and the bills have grown from tens of thousands to tens of lakhs. At the top, by doctrine: the Union of India has formally asked a nine-judge bench to abolish the thing outright, and the bench has reserved judgment.

Neither move is unreasonable on its face. Both are aimed at real abuse. What worries me is that the filter being built catches the wrong people, because it operates at the wrong moment.

What the door was for

Begin with the Court’s own description of who this jurisdiction was built for. In 1988 the full Court settled a list of the letter petitions its registry would treat as public interest litigation. The list, still on the Court’s site, is ten items long:

Bonded labour matters. Neglected children. Non-payment of minimum wages to workers and exploitation of casual workers. Petitions from jails complaining of harassment, death in jail, speedy trial as a fundamental right. Petitions against police for refusing to register a case, harassment by police and death in police custody. Petitions against atrocities on women, in particular harassment of bride, bride-burning, rape, murder, kidnapping. Petitions complaining of harassment or torture of villagers from persons belonging to Scheduled Castes and Scheduled Tribes and economically backward classes. Petitions pertaining to environmental pollution and other matters of public importance. Petitions from riot victims. Family pension.³

Read that list again with the standing question in mind. Some heads expressly contemplate the affected person writing in: petitions from jails, petitions from riot victims, petitions from persons belonging to Scheduled Castes and Scheduled Tribes. But a bonded labourer usually does not file writ petitions. A neglected child cannot. A man who died in police custody cannot. What these ten heads have in common is not that the victim never appears. It is that the jurisdiction does not require the victim to.

That is not a defect in the jurisdiction. It is the jurisdiction.

How it was opened, case by case

The opening was not a single judgment. It was about a decade of small procedural concessions, each of which looks minor and none of which the Court has ever formally withdrawn.

It starts with an attitude rather than a rule. In Mumbai Kamgar Sabha v. Abdulbhai Faizullabhai, 1976 INSC 49, decided in March 1976, Justice Krishna Iyer treated procedural objections to a union’s standing as something to be got out of the way rather than adjudicated. Procedure, on that view, serves the claim; it does not screen it.

Then the letters. In Sunil Batra v. Delhi Administration, 1978 INSC 147, a convict challenged solitary confinement and bar fetters. In the sequel, Sunil Batra v. Delhi Administration, 1979 INSC 271, decided in December 1979, the same prisoner wrote a letter to a judge about the torture of another inmate, and the Court treated the letter as a petition under Article 32. That is the moment the postcard became a writ.

In Hussainara Khatoon v. Home Secretary, State of Bihar, 1979 INSC 34, and the orders that followed through that year, the Court took up newspaper accounts of undertrials in Bihar who had been in jail longer than the maximum sentence for their alleged offence, and released them. Speedy trial became part of Article 21 in a case where nobody in custody had drafted anything.

In Municipal Council, Ratlam v. Vardhichand, 1980 INSC 138, Justice Krishna Iyer held that a municipality’s statutory duties to the public are enforceable at the instance of residents, and that an empty treasury is not an answer to a statutory duty. A few months later, in Fertilizer Corporation Kamgar Union v. Union of India, 1980 INSC 217, he set out why a broader class of persons should be heard when public power is misused.

The doctrinal statement came in S.P. Gupta v. Union of India, 1981 INSC 209, on 28 December 1981. Any member of the public acting bona fide, and not for personal gain or political motive, could move the court where a legal wrong was caused to a person or class unable to approach it themselves. Locus standi was not abolished. It was widened, and it was widened on a condition: bona fides, and the inaccessibility of the real victim. Everything that has gone wrong since is a failure to enforce that condition.

What followed came quickly. People’s Union for Democratic Rights v. Union of India, 1982 INSC 67, the Asiad workers case, held that non-payment of minimum wages to construction labour amounted to forced labour under Article 23, on a petition begun by a letter from an organisation with no personal stake at all. Bandhua Mukti Morcha v. Union of India, 1983 INSC 203, went further and built machinery: the Court appointed commissioners to go to the quarries of Faridabad and report back, treating fact-finding as part of its Article 32 power. Sheela Barse v. State of Maharashtra, 1983 INSC 9, began as a journalist’s letter about women in police lock-ups in Bombay.

And in M.C. Mehta v. Union of India, 1986 INSC 281, decided on 20 December 1986 after the oleum gas leak at Shriram, a five-judge bench under Chief Justice P.N. Bhagwati did three things at once: it held that Article 32 is remedial and not merely injunctive, so compensation can be awarded in a writ petition; it laid down absolute liability for hazardous enterprises; and it dealt squarely with whether letters addressed to an individual judge are entertainable. The headnote of that judgment carries “Public Interest Litigation, maintainability of” as a substantive head of law.

What came through the door

It is worth being concrete about what this procedure produced, because the debate is usually conducted as though public interest litigation were a synonym for nuisance.

Olga Tellis v. Bombay Municipal Corporation, 1985 INSC 151, brought by a journalist and pavement dwellers, put the right to livelihood inside the right to life. M.C. Mehta v. Union of India, 1987 INSC 261, closed tanneries discharging into the Ganga at Kanpur and treated Articles 48A and 51A as operative in a pollution case. Parmanand Katara v. Union of India, 1989 INSC 256, which began with a newspaper report of an accident victim shuttled between hospitals, established the duty of every doctor to give emergency treatment without waiting for police formalities.

Nilabati Behera v. State of Orissa, 1993 INSC 113, started as a letter from a mother whose son died in police custody and ended as the foundation of compensation in public law for custodial death. D.K. Basu v. State of West Bengal, 1996 INSC 1508, began as a letter to the Chief Justice and produced the arrest and detention safeguards that Parliament later wrote into the Code of Criminal Procedure as sections 41B, 41C, 41D and 50A, inserted by the Code of Criminal Procedure (Amendment) Act 2005 (Act 25 of 2005). Paschim Banga Khet Mazdoor Samity v. State of West Bengal, 1996 INSC 621, made the refusal of emergency treatment by a government hospital a violation of Article 21.

Vellore Citizens Welfare Forum v. Union of India, 1996 INSC 952, brought the precautionary principle and the polluter pays principle into Indian law. Indian Council for Enviro-Legal Action v. Union of India, 1996 INSC 237, made the chemical factories at Bichhri answer for the aquifer they destroyed, and held the petition maintainable although the polluters were private companies.

Vishaka v. State of Rajasthan, 1997 INSC 604, is the clearest case of all. A social worker was gang-raped in Rajasthan for trying to stop a child marriage. She was not the petitioner. Organisations were. The Court wrote the law on sexual harassment at work and it governed the country for sixteen years, until Parliament replaced it with the Sexual Harassment of Women at Workplace (Prevention, Prohibition and Redressal) Act 2013 (Act 14 of 2013). Under a strict rule of standing that petition is not maintainable, and there is no Vishaka. Vineet Narain v. Union of India, 1997 INSC 826, brought by a journalist over the Jain hawala diaries, produced the directions on the autonomy of the Central Bureau of Investigation and the Central Vigilance Commission that Parliament put on a statutory footing in the Central Vigilance Commission Act 2003 (Act 45 of 2003). And the right to food orders, made in People’s Union for Civil Liberties v. Union of India, Writ Petition (Civil) No. 196 of 2001, are why the midday meal became an entitlement a child could enforce rather than a scheme a government could discontinue. That petition ran for years on interim orders rather than one judgment, and the orders are themselves reported: the directions universalising the Integrated Child Development Services are 2006 INSC 1001 and 2007 INSC 730.

Not one of those petitioners had a personal grievance in the ordinary sense. That is the point that gets lost. If the ordinary principles of locus standi had applied, the list above is not a list of cases decided differently. It is a list of cases never heard.

The Court’s doubts are thirty-five years old

The honest part of the story is that the Court began worrying about misuse almost immediately, and has never stopped.

In Chhetriya Pardushan Mukti Sangharsh Samiti v. State of U.P., 1990 INSC 230, the Court found the environmental petition before it to be a vehicle for a private feud. In Subhash Kumar v. State of Bihar, 1991 INSC 3, it held in the same breath that Article 21 includes the right to pollution-free water and air and that the petition before it was a personal grudge dressed as public interest. In Janata Dal v. H.S. Chowdhary, 1991 INSC 210, arising out of Bofors, it dealt at length with the busybody and the meddlesome interloper.

The warnings then harden into a settled line of authority. Raunaq International v. I.V.R. Construction, 1998 INSC 468, dealt with petitions aimed at public contracts and the costs a losing petitioner should bear. BALCO Employees Union v. Union of India, 2001 INSC 601, held that economic policy is not a matter for public interest litigation. Guruvayur Devaswom Managing Committee v. C.K. Rajan, 2003 INSC 394, surveyed the whole jurisdiction and its limits.

Then a run of cases in the 2000s that reads exactly like the remarks made in May 2026, twenty years earlier. Ashok Kumar Pandey v. State of West Bengal, 2003 INSC 645, held that only a person acting bona fide and with sufficient interest has standing, and that a petition must not be publicity oriented or founded on personal vendetta. Dr. B. Singh v. Union of India, 2004 INSC 160, said the same of petitions built on newspaper clippings and nothing else. In Dattaraj Nathuji Thaware v. State of Maharashtra, 2004 INSC 712, the petitioner was an advocate who had been caught taking money from the parties he was suing; the High Court found blackmail, and the Supreme Court upheld the exemplary costs. Holicow Pictures v. Prem Chandra Mishra, 2007 INSC 1239, and, on the same day, Divisional Manager, Aravali Golf Club v. Chander Hass, 2007 INSC 1240, added the theme of judicial restraint.

The synthesis is State of Uttaranchal v. Balwant Singh Chaufal, 2010 INSC 47, decided on 18 January 2010. It ends with a set of directions to the High Courts: frame rules to encourage genuine petitions and discourage those filed for extraneous reasons; verify the credentials of the petitioner before entertaining the petition; satisfy yourself of the correctness of the contents; ensure that substantial public interest is involved; and impose exemplary costs where it is not.

The best illustration of what that filter is supposed to catch is Kalyaneshwari v. Union of India, 2011 INSC 60. A petition sought a ban on asbestos. The Court refused it in January 2011, and then, in Kalyaneshwari v. Union of India, 2011 INSC 386, issued contempt notice, having found that the petitions had been filed against one business rival at the behest of another. And in Santosh Singh v. Union of India, 2016 INSC 549, a petition to make moral science a compulsory school subject was dismissed on the ground that ideology and policy design do not become justiciable merely because somebody has filed a petition about them.

So the diagnosis announced from the bench in 2026 is not new. It has been the settled position for a generation. What is new is the remedy.

The price

Costs in public interest litigation used to be measured in the low tens of thousands. They are not any more.

In December 2022 the Supreme Court declined to disturb a cost of one lakh rupees imposed by the Bombay High Court on an organisation that had challenged a Worli redevelopment while, on the High Court’s reading of its own memorandum of association, having no ecological object at all. The then Chief Justice, D.Y. Chandrachud, said that petitions of that kind “could become an instrument of blackmail when it is an issue of an infrastructure project”.⁴

In July 2025 a division bench of the Gujarat High Court dismissed a petition against a builder’s development permission with costs of twenty lakh rupees on each of seven petitioners, one crore forty lakh in all, payable to the State Legal Services Authority. The court’s reason was that the petitioners had not disclosed who they were: “Who are these people, nobody knows. What business they are doing, what their occupation is, nothing.”⁵

In January 2026 the Supreme Court refused to reopen a five lakh rupee cost imposed in 2023 on a Lucknow advocate who had petitioned about the wording of a High Court chief justice’s oath. He said the figure announced in open court had been twenty-five thousand. The Bench, led by Chief Justice Surya Kant, answered: “We will go by the order, and it contains Rs five lakh.” And then: “These all are publicity petitions.”⁶

In August 2026 the Delhi High Court dismissed a petition about the sale of National Stock Exchange shares by IFCI with exemplary costs of five lakh rupees, because the petitioner had a substantially similar case pending in Bombay and had sworn on paper that he did not.⁷

Alongside the orders there is the warning, which costs the Court nothing and the petitioner a great deal. In May 2025 a Bench led by Justice Surya Kant asked the petitioner, advocate Vishal Tiwari, who had filed about tourist safety after the Pahalgam attack: “Why have you filed this kind of PIL? What’s your real motive?... I think you are inviting some exemplary cost for filing this PIL.” Days earlier the same judge had told a petitioner in a related matter, “Be responsible before filing such PILs. You have some duty towards your country also.”⁸ In March 2026 the Court dismissed five petitions by one advocate, including one asking for a scientific study of whether onion and garlic carry tamasic energy, and said it would have imposed exemplary costs had he not been an advocate.⁹ In August 2025 it turned away a petition by an advocate, Narendera Kumar Goswami, seeking a safety audit of Air India, with the question, “Why target Air India only which recently witnessed an unfortunate tragedy?”¹⁰

Some of those petitions deserved everything they got. The onion petition is indefensible. The IFCI petitioner swore a false statement. The Gujarat petitioners hid who they were.

But look at what a cost order is, structurally. It is a liability of unknown size, fixed after the event, by the judge who has just decided s/he does not believe you. It does not distinguish between the petitioner who lied and the petitioner who was merely wrong. Balwant Singh Chaufal contemplated exemplary costs as the tail end of a screening process that begins with credentials at the admission stage. What happened instead is that the screening was never built and the penalty grew. Anyone deciding today whether to file about a bonded labour camp, a lock-up death or a landfill has to price in a possible twenty lakh rupees. A busybody with a commercial backer can absorb that. An organisation working on bonded labour cannot.

That is what losing teeth actually looks like. Not a doctrine repealed. A doctrine priced.

The nine-judge bench

The doctrinal move is more direct.

When the Supreme Court disposed of the review petitions against its 2018 Sabarimala judgment, it did not decide them. In Kantaru Rajeevaru v. Indian Young Lawyers Association, 2019 INSC 1248, decided on 14 November 2019, a five-judge bench referred seven questions to a larger bench. Six are about Articles 25 and 26: the scope of judicial review of religious practice, the meaning of morality, the relationship between individual and denominational rights. The seventh is not about religion at all. It asks whether a person not belonging to a religious denomination can question that denomination’s practice by filing a public interest litigation.¹¹

That question sat unheard for six years. On 7 April 2026 a nine-judge bench headed by Chief Justice Surya Kant, with Justices Nagarathna, Sundresh, Amanullah, Aravind Kumar, Masih, Varale, Mahadevan and Bagchi, began hearing it.

The Union of India used the opening. In written submissions the Solicitor General, Tushar Mehta, asked the Court to remove public interest litigation as a jurisdictional exception and to restore the ordinary principles of locus standi. The submissions describe India’s present position, in which any member of the public may challenge any law without demonstrating personal injury, as a “global constitutional anomaly”. They say the culture has “metastasised into a general licence for ideological, political, and denominationally hostile litigation”. They describe the Sabarimala petition as “a PIL filed by persons with no personal stake, representing no identifiable class of inaccessible victims, against a practice they did not follow, in respect of a deity they did not worship”. And they attribute a catastrophic expansion of the docket to the jurisdiction, putting filings at roughly twenty-five thousand in 1985 and over seventy thousand in 2019.¹²

On 5 May the bench said what it said. On 14 May, after sixteen days, it reserved judgment. As I write, nearly four months later, the judgment has not been delivered.¹³

There is an oddity in the position the Court is being invited to take. The Sabarimala petition of 2006 is now said to have been unworthy because the petitioners had no stake in the temple. But the reference itself, and the six constitutional questions the nine judges have spent sixteen days on, exist only because that petition was heard. If a stranger to a denomination cannot question its practice, the bench’s jurisdiction over everything else in front of it is an inheritance from a petition it says should have gone in the bin.

Whether the docket claim is true

The strongest empirical claim on the abolition side is docket expansion. It should be tested, because it is testable.

The only recent systematic count I have found goes the other way. A study of 3,190 such petitions filed in the Bombay High Court between 2021 and 2025 found that they never exceeded half a per cent of that court’s annual docket in any year, and that their share fell over the period. Of the 1,408 that had concluded, the petitioner succeeded in 29 per cent, more than half were dismissed, and about a fifth were withdrawn. The authors are careful to say the finding cannot be extrapolated to the Supreme Court.¹⁴

Take the two claims together. If half a per cent of a High Court’s file is the order of magnitude, no docket is being drowned. And a 29 per cent success rate is not the profile of a jurisdiction full of frivolous petitions. The number doing the work in the Union’s submission is the seventy thousand, and it deserves an answer from somebody with access to the Supreme Court’s own filing data, which is not published in a form that lets an outsider check it.

The access premise

The other limb of the argument is that the justification of the 1980s has expired. Legal aid exists now, there is a legal services authority in every district, cases can be filed online, and a poor litigant is no longer structurally locked out.

Some of that is true. Enough of it is not. The 2025 India Justice Report records that the number of paralegal volunteers, who are the last mile of legal aid, fell by 38 per cent between 2019 and 2025. Per capita spending on free legal aid is about six rupees. Electronic filing assumes bandwidth that most rural households do not have, and only a small fraction have a connection stable enough to upload a paginated petition.¹⁵

And there is a category problem that no amount of legal aid fixes. Legal aid helps a person who can instruct an advocate. Look again at the 1988 list. Bonded labour. Neglected children. Death in custody. Atrocities against Scheduled Castes and Scheduled Tribes. Those are precisely the situations in which the affected person often cannot instruct anybody, because s/he is confined, or dead, or a child, or frightened of the police station s/he would have to walk into to complain about the police. A legal services authority is not a substitute for a jurisdiction that does not require the victim to appear. It is a different thing entirely.

What should actually change

The answer is not to defend every petition filed under this label. The Court is right that a great deal of what reaches it is private grievance, competitor warfare or advertising. The answer is that the filter is in the wrong place and made of the wrong material.

Almost every abusive petition in the cases above was detectable at the threshold, on facts the petitioner was already obliged to disclose. In Dattaraj the petitioner was taking money from the respondents. In Kalyaneshwari the petition was filed for a business rival. In the Gujarat case the petitioners refused to say who they were. In the Delhi case an identical petition was pending in another High Court and the petition said otherwise on oath. None of that requires a hearing on the merits. It requires the registry to ask four questions before admission: who are you, what is your interest, who is funding this, and what else have you filed on the same facts. That is what Balwant Singh Chaufal told the High Courts to build sixteen years ago, and it is largely not built.

Verification at admission is the opposite of a cost order at dismissal. It is cheap, it is knowable in advance, and it screens on conduct rather than on outcome. A petitioner who answers the four questions honestly and loses has done nothing wrong. A petitioner who cannot answer them never gets a number.

The distinction that matters is not between the deserving cause and the undeserving one, which is a judgment on the merits dressed up as a rule of standing. It is between the petitioner who is standing in for someone who cannot come, and the petitioner who is standing in for a private interest. S.P. Gupta drew that line in 1981. It has never needed replacing. It has only ever needed enforcing.

The nine judges may yet say something narrow: that a stranger to a denomination cannot litigate its internal practice. That would answer the seventh question without touching bonded labour or custodial death. The wider answer, restoring ordinary locus standi across the board, is not available without accepting what it costs, and the honest way to state the cost is by name. No Hussainara Khatoon. No Bandhua Mukti Morcha. No Vishaka. No D.K. Basu. No Nilabati Behera.

The letter Sunil Batra wrote from Tihar was not a loophole in the Court’s procedure. For a while it was the Court’s best answer to the problem that the people with the strongest claim on Article 21 are the people least able to file anything. That problem has not been solved. It has only stopped being fashionable.