A curious reference at the Allahabad High Court

Article 141 of the Constitution provides that ‘[t]he law declared by the Supreme Court shall be binding on all courts within the territory of India.’ Supplementing this, Article 144 further confirms that ‘All authorities, civil and judicial, in the territory of India shall act in aid of the Supreme Court.’ One can correctly read these provisions as obliging all High Courts to follow the law declared by the Supreme Court. 

A single-judge Bench order of the Allahabad High Court has uniquely tested the limits of this principle. On 27 May 2025, in Shashank Gupta v State of UP, Justice Arun Kumar Singh Deshwal referred a question of criminal procedure to a nine-judge Bench of the same High Court. 

He noted that while a seven-judge Bench had previously ruled on the question, the Supreme Court had decided the opposite way in subsequent cases. So while the High Court’s precedent had turned ‘obsolete’, it had not been formally overruled within the High Court. “In the interest of judicial discipline,” Justice Deshwal called for a nine-judge Bench to be constituted to explicitly overrule the seven-judge Bench decision. 

In this piece, I examine whether the single judge could have decided the case without referring it to a larger Bench. 

‘Savings’ and grace 

Section 482 of the Code of Criminal Procedure 1972 (CrPC) was a savings provision. It preserved the ‘inherent powers’ of the High Courts to ‘secure the ends of justice.’ The provision allowed the High Court to make a variety of orders to prevent abuse of law and process. Section 482 of the CrPC is now Section 528 of the Bharatiya Nagarik Suraksha Sanhita, 2023.

In a line of decisions from the 1940s to the 1960s, beginning from the Privy Council in Emperor v Khwaja Nazir Ahmad (1944) to the Supreme Court in State of West Bengal v SN Basak (1962), it was settled that this section could not be used to quash a first information report (FIR) or prevent investigations by the police. 

In Ram Lal Yadav v State of UP (1989), a seven-judge Bench of the Allahabad High Court affirmed this view, holding that where a cognisable offence was concerned, the High Court did not have the power to interfere or quash the investigation before the chargesheet had been filed. In cases of non-cognisable offences, the High Court could use its powers under Article 226 to quash proceedings, but Section 482 was still toothless.

However, a couple of years later, the Supreme Court overturned this position in State of Haryana v Bhajan Lal (1990). It held that in cases where no cognisable offence could be made out from the FIR, the High Court could use Section 482 to quash the FIR or stop the investigation. 

For your reference 

This brings us to Justice Deshwal’s order in Shashank Gupta. The accused had applied for quashing the investigation against him under Section 528 of the BNSS. Recognising the binding nature of Bhajan Lal, the judge observed that “the evolution of circumstances and legal principle as articulated in the Apex Court’s judgement render the law established by Ramlal Yadav (supra) obsolete.” 

Despite this, Justice Deshwal referred the question of Ram Lal Yadav’s correctness to a nine-judge Bench so that it may be formally overruled within the Allahabad High Court. One can see why Justice Deshwal chose the route. Though the Allahabad High Court is bound by Supreme Court precedent, it must also ensure that its own large-Bench set precedents are not effaced or overruled by smaller Benches. The reference of this matter to a larger Bench gets the best of both worlds—the High Court does not rule against the Supreme Court or its own larger Bench. 

But did Justice Deshwal have to refer the matter to a larger Bench for judicial discipline reasons? 

Speaking as the High Court 

There are three reasons why the Shashank Gupta Bench should have followed Bhajan Lal without referring the matter and waiting for Ram Lal Yadav to be overruled by a larger Bench. Bhajan Lal is binding on the single-judge Bench of the High Court just as much as it would be to the nine-judge Bench. Even when sitting alone, Justice Deshwal speaks as the High Court of Allahabad. Accordingly, Article 141 applies to the decisions he renders, much like it applies to any nine-judge Bench that may sit over the reference question. 

Justice Deshwal, whenever he renders a decision, must follow Bhajan Lal as a matter of constitutional obligation. If the nine-judge Bench were to hypothetically render a verdict that conflicts with Bhajan Lal, Justice Deshwal would still be obligated to disregard it and follow the top court’s line under Article 141.  

Secondly, overrulings can be explicit or implicit. Once the Supreme Court declares the law on a point, all High Court decisions to the contrary are automatically overruled. As D.D. Basu states in his seminal Commentary on the Constitution of India (LexisNexis, 9th edn, vol 9), a High Court Bench need not refer a question of a precedent’s correctness to a larger Bench “if the Supreme Court has given a decision in the matter. As soon as the Supreme Court gives its decision, all decisions of the High Court on the point are overruled.” 

Finally, let us entertain the hypothetical that Ram Lal Yadav is considered not to have lost the force of law. Perhaps this is based on the argument that Ram Lal Yadav was not clearly overruled by Bhajan Lal. One might argue that Justice Deshwal was bound by both decisions and needed clarity on what to apply. 

Here, it is relevant to note that the obligation for a smaller Bench to follow the decisions of a larger Bench is a device of judicial discipline. It is a rule of self-restraint created by courts out of public policy considerations. On the contrary, Article 141 is a constitutional provision. It has superior legal force than any such rule adopted by courts. In a conflict between the two, Article 141 would prevail. Accordingly, Justice Deshwal would be bound by Bhajan Lal regardless of whether Ram Lal Yadav was operative or not.  

In CN Rudramurthy v K Barkathulla Khan (1998), the Supreme Court affirmed the view that Article 141 is independent of any doctrine of precedents. It applies as a rule, and therefore cannot be negated on the basis that a different policy may be more conducive to Bench discipline. 

Conflict and clarity 

The ability to refer questions to a larger Bench is a discretionary power. This means that, even at its worst, Justice Deshwal’s reference was superfluous. This does not make it illegal. Ideally, the High Court would proceed with Bhajan Lal controlling the field, and the application for quashing under Section 528 would be analysed on its merits. 

That established, one can contemplate a situation where a reference such as this one would be more appropriate than directly applying Supreme Court precedent—this is when there is some ambiguity around whether a Supreme Court decision overrules a High Court verdict. 

In these cases, it would be prudent—although still not necessary—for the High Court to reference the question of the previous ruling’s validity to a larger Bench, in light of the Supreme Court precedent. This is because, even if the larger Bench finds that the Supreme Court did not overrule the High Court precedent, it does not need to rest with that conclusion. It has the option of considering the wisdom of the High Court precedent afresh and overruling it of its own accord. However, as Justice Deshwal himself notes, this was not the case here—Bhajan Lal overruled Ram Lal Yadav, if only implicitly. 

The law of precedents is a delicate terrain. It has the potential to create smooth and efficient intersections, securing effective and principled disposal of cases. Article 141 provides a schematic foundation for this. Obstacles in this process are injurious: they not only cause confusion in the case at hand but also add to the log-jam of precedent, potentially giving rise to greater uncertainty and inefficiency. In making a reference, therefore, High Courts must be cautious and conscious. 

Sarthak is an intern at the Supreme Court Observer.