On September 23, a two-judge Bench of the Supreme Court of India, comprising Justices Dipankar Datta and Satish Chandra Sharma, delivered a split verdict on a set of petitions which assailed the validity of the Chief Election Commissioner and other Election Commissioners (Appointment, Conditions of Service and Term of Office) Act, 2023. Remarkably, the judges were divided not on the merits of the challenge, but on whether a Bench of two was competent to decide the matter at all.

An appointment framework

The statute, which was enacted in December 2023, provides that the Chief Election Commissioner and the Election Commissioners shall be appointed on the recommendation of a selection committee comprising the Prime Minister, the Leader of the Opposition in the Lok Sabha, and a Union Cabinet Minister nominated by the Prime Minister. This law replaced an interim arrangement that a Constitution Bench of five judges had put in place in Anoop Baranwal vs Union of India (2023). Under that arrangement, the Chief Justice of India (CJI) sat as the third member in place of a Minister.

Midway through the hearings in the matter, which began in May, the Union government changed tack and argued that the case raised a substantial question of law as to the interpretation of the Constitution. It therefore claimed that the challenge, in terms of Article 145(3), ought to be heard by a Bench of at least five judges. Curiously, this plea was made even though it found no place in the Union’s counter-affidavit.

While Justice Datta rejected the request, Justice Sharma accepted it. Ordinarily, a difference of opinion on a Bench of two is resolved by a Bench of three. But reasoning that a ruling on which of their opinions was correct would only delay matters further, both judges jointly placed the papers before the CJI asking him to consider constituting a Constitution Bench. The upshot is this: the petitioners, who came to court in early 2024, must now wait again.

As Justice Datta rightly held, Article 145(3) merely prescribes that a case involving a substantial question of constitutional interpretation ought to be decided by a Bench comprising a minimum of five judges. It does not ask whether a case is important or whether a statute under challenge is new. As far back as in 1959, in State of Jammu and Kashmir vs Thakur Ganga Singh, the Court explained, albeit in a different context, what a question of constitutional interpretation entails. “The question of interpretation can arise only if two or more possible constructions are sought to be placed on a provision — one party suggesting one construction and the other a different one,” the Court had held. “But where the parties agree on the true interpretation of a provision or do not raise any question in respect thereof, it is not possible to hold that the case involves any question of law as to the interpretation of the Constitution.” A Bench of three judges later applied the same reasoning to Article 145(3) itself, in People’s Union for Civil Liberties vs Union of India (2003).

A deeper institutional fault

What, then, did Anoop Baranwal settle? Justices Datta and Sharma agreed that the committee the judgment had created was a stopgap, meant to last only until Parliament enacted a law. To that extent, the Union may well be right in claiming that the CJI’s place on the committee is by no means constitutionally entrenched. But the Constitution Bench also read Article 324 to require that appointments to the Election Commission be insulated from exclusive executive control. This was a clear declaration of law, binding on every court under Article 141. As Justice Datta held, whether a committee in which the executive holds two votes out of three meets that standard is merely a question of applying this settled rule to a new statute.

Justice Sharma’s opinion adopted a different test. He held that since “there is no authoritative decision of this Court till date pronouncing upon the constitutional validity of the [2023] Act, the matter must be heard by a Constitution Bench of not less than five Judges.” But almost any challenge to a fresh statute would, in this sense, be the first of its kind. On such logic, Benches of two judges could not have decided Shreya Singhal vs Union of India (2015), which struck down Section 66A of the Information Technology Act or NALSA vs Union of India (2014), which affirmed the right of transgender persons to their self-identified gender. Each called for a searching analysis of the Constitution and its guarantees.

Ultimately, though, a disagreement of this kind points to a deeper fault in the Court’s architecture.

When the Court first sat, in January 1950, it had six judges (a sanctioned strength of eight judges). Article 145(3), with its minimum of five, assumed that the Court would, in the main, sit together to interpret the Constitution. Today the Court has a sanctioned strength of 38 judges. It functions largely in panels of two and spends most of its time on special leave petitions. A Constitution Bench is assembled only when the CJI can spare at least five judges from this daily churn. The result, as the scholar Nick Robinson has argued, is not one court but many, with each Bench speaking in its own voice.

To be fair, the joint order signed by Justices Datta and Sharma recognises as much. Anoop Baranwal arose out of a petition instituted in 2015, was referred to a Constitution Bench in 2018, and was decided in 2023. The reference in State of Uttar Pradesh vs Jai Bir Singh, to decide what counts as an “industry”, lay dormant for 15 years from 2002, and took nine more to reach the nine-judge Bench that finally decided it. The judges themselves describe such delay as a source of “institutional embarrassment.”

The need for clearer rules

Delay of this kind carries with it a corrosive consequence: it presents the Court with a fait accompli. Take the present challenge. By the time it is decided, Commissioners appointed under the law would have overseen multiple State elections and likely multiple general elections too. In their joint order, the judges ask the CJI to explore a permanent Constitution Bench to decide pure questions of constitutional law. The idea is far from new. The Law Commission, in its 229th report in 2009, proposed a Constitution Bench sitting in Delhi, with Benches in different regions of the country hearing ordinary appeals.

A Bench of that form would need no constitutional amendment. The CJI, as the Master of the Roster, can create one tomorrow. But therein lies another difficulty. On a court of 38 judges, the power to decide who sits on which Bench is more than an administrative power. It can, and often does, shape outcomes. We might therefore ask ourselves whether such authority ought to be concentrated in a single office. For a permanent Bench to command confidence, it must be composed through a procedure that is transparent, settled in advance, and based on bright-line rules.

There is, however, a further difficulty. India’s legal system is built on a supposed commitment to the principle of stare decisis, a promise to stand by things previously decided. Yet, there is little to suggest that the Court’s Benches, whatever their size, will consistently honour what earlier Benches have held. The present case shows as much. A Bench of two, albeit divided, has declined to apply a principle settled by a Bench of five, and has sent the case to a Constitution Bench simply because the statute under challenge is new.

The Court, in its original conception, was seen as a guardian of the Constitution that would also hear appeals. Over time, though, it has become a forum that sits predominantly as a court of appeal rather than as a constitutional court. This inversion demands urgent correction. Otherwise, the gravest questions facing the republic will continue to be answered only once the answers have ceased to matter.

Suhrith Parthasarathy is an advocate practising in the Madras High Court