Two recent bail rulings delivered by different benches of the Supreme Court of India in early January 2026 have brought renewed attention to what critics describe as an internal inconsistency in judicial reasoning, often referred to as the problem of a “polyvocal court.” Together, these decisions have highlighted how similar legal principles can be interpreted in markedly different ways by coordinate benches, raising broader concerns about predictability and coherence in constitutional adjudication.
On January 5, a bench comprising Justice Aravind Kumar and Justice N V Anjaria declined to grant bail to student activists Sharjeel Imam and Umar Khalid, who stand accused of being part of a “larger conspiracy” linked to the 2020 North East Delhi riots. While acknowledging that both men had spent an exceptionally long period in custody and that there was little likelihood of the trial concluding in the near future, the bench held that prolonged incarceration alone did not automatically render their continued detention punitive or unconscionable. According to the judges, delay by itself could not justify bail if other statutory and evidentiary considerations weighed against the accused.
In sharp contrast, the very next day, on January 6, another two-judge bench of the Supreme Court—comprising Justice Sanjay Kumar and Justice Alok Aradhe—granted bail to Arvind Dham, the former chairman of Amtek Auto, in a case registered under the Prevention of Money Laundering Act (PMLA). Dham faces allegations related to a bank fraud amounting to nearly ₹25,000 crore. In granting bail, this bench relied squarely on the grounds of prolonged pre-trial incarceration and the absence of any realistic prospect of the trial concluding within a reasonable timeframe. The judges explicitly held that the fundamental right to a speedy trial under Article 21 of the Constitution is not overridden merely because of the seriousness or nature of the alleged offence.
At first glance, the divergence between the two verdicts has raised an obvious question: is it inherently problematic if different benches of the Supreme Court arrive at different conclusions? The answer lies in the doctrine of binding precedent, which underpins the Indian judicial system. This principle requires that a decision of the Supreme Court delivered by a bench of equal or greater strength is binding on subsequent benches, provided the legal issue and material facts are substantially similar. The doctrine exists to promote certainty, consistency, and predictability in the law—values considered essential to the rule of law itself.
Ordinarily, if a coordinate bench finds an earlier ruling to be flawed or of doubtful applicability, judicial discipline requires it to refer the matter to a larger bench rather than depart from the precedent on its own. Critics argue that the Supreme Court has increasingly deviated from this convention, resulting in different benches effectively “speaking in different voices.” This, they say, creates a perception that outcomes depend more on the composition of the bench than on settled legal principles, thereby undermining public confidence in the justice system.
What makes the situation more striking is that the same bench that denied bail in the Delhi riots case had, earlier that very day, strongly reaffirmed the importance of adhering to precedent. In a separate judgment involving a dispute over customs duty imposed on Adani Power Limited for electricity generation within a Special Economic Zone, Justices Kumar and Anjaria observed that if a later bench doubts the correctness of an earlier ruling, the only legitimate course is to refer the matter to a larger bench. They cautioned that a coordinate bench cannot dilute or sidestep an existing precedent by artificially narrowing its scope, emphasizing that judicial discipline is not merely a matter of courtesy but an institutional necessity that ensures coherence and legitimacy in adjudication.
Commenting on this apparent contradiction, senior advocate Dushyant Dave told India Today that the same bench had applied precedent in diametrically opposite ways on the same day. He argued that the Supreme Court’s primary role is to safeguard constitutional rights, particularly personal liberty, and described the riots case verdict as deeply flawed and perverse.
A broader review of Supreme Court rulings in cases under the Unlawful Activities (Prevention) Act (UAPA) and the PMLA further illustrates this uneven pattern. An analysis by the Supreme Court Observer of around 39 such verdicts shows that while some accused persons have been granted bail after long periods of incarceration, others in comparable situations have been denied relief. In one instance in February 2025, two different benches ruled on PMLA bail pleas within days of each other—granting bail in one case while refusing it in another—despite similar arguments based on prolonged pre-trial detention.
Both the UAPA and the PMLA impose stringent conditions on the grant of bail. Section 45 of the PMLA restricts bail unless the court is satisfied that the accused is not prima facie guilty and is unlikely to commit an offence while on bail. Similarly, Section 43D(5) of the UAPA bars bail if the court finds reasonable grounds to believe that the accusations are prima facie true. However, the Supreme Court has repeatedly clarified that these provisions do not amount to an absolute bar and that courts must balance statutory restrictions against the constitutional guarantee of personal liberty under Article 21.
This balance was central to the landmark 2021 judgment in K A Najeeb case, where a three-judge bench held that when a timely trial is clearly not possible and the accused has already suffered prolonged incarceration, courts would ordinarily be obliged to grant bail, even overriding statutory embargoes. Relying on this precedent, lawyers for the Delhi riots accused argued that with more than 1,000 witnesses yet to be examined, charges not even framed, and over five years already spent in jail, bail ought to have been granted.
The bench hearing the riots case, however, placed greater reliance on later two-judge bench decisions, including the 2024 Gurwinder Singh case and the 2019 Zahoor Watali case, which emphasized the gravity of the offence and the presence of prima facie evidence. According to the bench, the Najeeb ruling was not intended to operate as a universal or mechanical formula mandating bail in every case of prolonged detention. Instead, it was described as a constitutional safeguard to be applied selectively, depending on the facts and evidentiary material before the court.
Legal experts remain divided. Former Supreme Court judge Justice Abhay S Oka has stated that the Najeeb precedent ought to have been followed, particularly in cases involving long periods of pre-trial incarceration without realistic prospects of trial completion. He noted that seriousness of the offence or the maximum punishment prescribed cannot, by themselves, justify indefinite denial of bail, unless there are clear risks such as criminal antecedents or the possibility of influencing witnesses.
Dushyant Dave echoed this view more forcefully, warning that frequent departures from precedent create uncertainty and erode trust in the judiciary. Citing the famous observation of US Supreme Court Justice Benjamin Cardozo, Dave argued that changes in the composition of benches should not lead to fluctuating constitutional interpretations. When justice appears dependent on chance rather than settled principle, he said, it ceases to be certain or guaranteed—striking at the very foundation of the rule of law.
