The making of the Indian Constitution was not only a laborious and democratic process, but also one punctuated with stories and conversations that explain why our framers designed institutions the way they did. Many of these appear in a note prepared by Sir Benegal Narsing Rau, the Adviser to the Constituent Assembly, where he recorded his interactions with judges, jurists, and statesmen across the world during the drafting process. Rau’s meeting with Justice Felix Frankfurter is usually remembered for Frankfurter’s caution against adopting the American “due process” standard under Article 21. But in that same discussion, Frankfurter offered another piece of advice to Rau, remarking that the Constitution must ensure that the Supreme Court sits en banc, so that every judge shares collective responsibility for every constitutional decision. Frankfurter was of course drawing inspiration from the American model, where the court sits together and speaks through a single authoritative voice.

This suggestion was unsuitable to India, particularly because unlike the U.S. Supreme Court, our Court was designed to function not merely as a constitutional court but as a general appellate body with far wider powers of review. It was expected to hear a vast and diverse docket, making en banc sittings wholly impracticable. The Constitution, therefore, adopted a compromise. Article 145 empowered the Court to determine the minimum number of judges who would sit for different classes of cases, thereby permitting single-judge and division benches. But Clause 3 created an important exception: when a case involved a “substantial question of law as to the interpretation of the Constitution,” the Court had to sit in a bench of at least five judges. At the time, the Court had eight judges in total; hence, a five-judge bench was arguably meant to ensure that constitutional questions were answered by a majority of the Court, preserving coherence and institutional consistency.

This design served two purposes. First, it ensured that constitutional interpretation reflected the collective wisdom of a larger bench. Second, it ensured that smaller benches could rely on clear constitutional precedent established by constitution benches, thus maintaining internal consistency.

Whether this aspiration continues to hold has been the subject of unusual introspection from within the Court. Over the past month, two judges—Justice Dipankar Datta and Justice B.V. Nagarathna—have expressed concern about how easily Supreme Court judgments are now reopened or set aside by subsequent benches.

Justice Datta, writing in a bail matter, spoke of a “growing trend” of verdicts being reconsidered or overturned by differently constituted benches, sometimes in quick succession. Article 141, he observed, exists so that a pronouncement of law “should settle the controversy” and guide future courts. Allowing litigants to reopen settled questions merely because a differently composed bench might see the matter differently undermines judicial authority. He cautioned that only errors “grossly erroneous on the face of the record” should justify reconsideration—through review or curative jurisdiction, not fresh litigation. Justice Nagarathna, in a separate instance, framed the concern in terms of judicial independence. A judgment, she said, must remain “written in ink and not in sand.” Respect for precedent is not only an internal judicial ethic but a constitutional value shared across the legal system. A decision cannot be discarded simply because “the faces have changed.”

These observations do not target any particular judgment; nor do they suggest that reconsideration is inherently problematic. Rather, they invite us to examine the institutional factors that make such revisions increasingly common. Public commentary often attributes this to ideological shifts or bench-hunting, but these account only for some instances. The deeper structural issue—and the one most relevant to constitutional adjudication—is the gradual dilution of Article 145(3).

Although Article 145(3) requires a minimum five-judge bench for any case involving a substantial question of constitutional interpretation, in practice benches of two or three judges often decide such matters. These include questions of federalism, legislative procedure, fundamental rights, and the powers of constitutional offices. The recent dispute between the Tamil Nadu government and the Governor is illustrative of this practice, as a two-judge bench introduced the doctrine of “deemed assent” to address prolonged gubernatorial inaction, effectively introducing an additional requirement in the text of the Constitution. Within months, a Constitution Bench while answering a Presidential Reference set aside the doctrine. In previous columns, I have explained the constitutional infirmities with the Presidential Reference opinion and argued that the division bench’s reasoning was not entirely without basis. But for present purposes, the deeper issue is different: the division bench should not have decided a question that, by constitutional design, falls squarely within the domain of a Constitution Bench.

A similar pattern emerged in the recent stray-dog matter, where a division bench was presented with an issue involving a conflict between the fundamental rights of human beings and the rights accorded to animals, and how the two should be reconciled—unquestionably a matter of constitutional significance. Yet instead of referring the case to a Constitution Bench, the division bench issued directions with wide-ranging implications, which were subsequently modified by a larger bench in quick succession. Notably, the Presidential Reference in the Governor’s case had identified, as one of its key questions, whether benches must first determine if Article 145(3) is attracted before deciding such matters. The Constitution Bench, however, declined to address this issue.

This is not to suggest that the erosion of Article 145(3) is the sole explanation for the trend highlighted by Justices Datta and Nagarathna. The revisiting of judgments occurs across many areas of law, and for varied reasons. However, the treatment of Article 145(3) highlights a predictable pattern: when substantial constitutional questions are decided by benches not institutionally designed to decide them, later reconsideration becomes much more likely.  Division benches, by virtue of their size, may not always be able to offer the breadth of deliberation or institutional finality that a Constitution Bench is designed to provide. Their decisions may leave open questions, generate differing lines of authority across benches, or prompt calls for clarification—developments that tend to reduce stability in areas where the Constitution anticipates a more authoritative resolution. Beyond Article 145(3), a number of broader institutional factors also contribute to this practice. Important matters are sometimes disposed of through brief or non-speaking orders, which lack doctrinal depth and are more susceptible to later refinement. In some cases, affected parties are not fully heard, leaving judgments procedurally fragile. The Supreme Court frequently sits in more than a dozen benches, making doctrinal coherence challenging.

This highlights a deeper concern: the dilution of predictability and stability—values central to the rule of law. Articles 141 and 145(3) were incorporated into the Constitution precisely to safeguard these virtues. The constitutional scheme recognises that judicial decisions may need reconsideration, but it channels such evolution through disciplined mechanisms: review, curative petitions, or references to larger benches. What it does not envisage is ad hoc revisiting of settled law by differently composed benches.

Reaffirming the discipline of Article 145(3) is therefore an important first step. If substantial constitutional questions are placed before Constitution Benches at the outset, fewer decisions will require subsequent correction. This would not only strengthen stare decisis, but also reinforce the institutional coherence that constitutional adjudication demands.

Swapnil Tripathi leads Charkha, the Constitutional Law Centre at the Vidhi Centre for Legal Policy.

Views are personal.