[This is a guest post by Bhushan Raut.]
In the second episode of the second season of the popular Netflix series “Maamla Legal Hai”, a newly appointed judge, casually smiles at a familiar advocate and is immediately branded as biased. Though exaggerated, it nonetheless captures an enduring anxiety within legal systems where justice must not only be done, but must also be seen to be done.
This blog has already examined the factual matrix and offered a preliminary critique of the judgement. Given the breadth of the issues involved, no single piece can exhaust its doctrinal implications. This post therefore focuses on additional lines of critique that remain underexplored.
Critical Analysis of the Judgement’s holdings
I. Prima Facie Observations as a Ground for Recusal
One of the substantial grounds argued by the applicant was the March 9, 2026 order. It was passed ex parte, and described the trial court’s 600-page discharge order as “prima facie erroneous” after an extremely brief hearing. This, as per Kejriwal, created a reasonable apprehension that the court had already formed a view. The Court’s response to this ground is legally orthodox but analytically thin.
The Court is correct in holding that prima facie observations are inherently tentative and do not by themselves give rise to a reasonable apprehension of bias or pre-judgement (Indore Development Authority v. Manohar Lal and AIIMS v. Prof. Kaushal K. Verma). Retired Justice Abhay Oka in an interview argues that sometimes a strongly worded prima facie observation is to invite effective counter-argument from the affected party.
However, the difficulty lies in the application and absence of an engagement with the contextual dimension of the prima facie observations. By treating the “prima facie” label as dispositive, it avoids the more difficult inquiry mandated by recusal jurisprudence. The Court fails to evaluate the nature, context and manner in which such prima facie observation was made. The March 9 order was expressed ex parte, at speed, and in emphatic terms and, when taken cumulatively may create a perception that the court has committed itself to a particular view. The omission, therefore, is not doctrinal but analytical where the Court failed to apply the standard to the facts.
II. The Kanaklata Distinction
The Supreme Court in Kanaklata v. State of NCT Delhi held that there can be situations where the nature of the observations made by the court can be so strongly worded that it can create a reasonable apprehension in the mind of the litigant that the court has committed to a thought process and cannot be retraced back and give a fair and non-partisan view in that matter.
Negating the reliance placed by Kejriwal on Kanaklata, the Court tries to distinguish it from the present case on the ground that the observations in Kanaklata were made in the same proceeding (discharge order), whereas the present observations were made at the stage of bail and were confined to the material then available. This distinction is based on a critical question: does Kanaklata require that the prior observations arise from the same proceeding, or does it extend to observations made in related proceedings?
Applying Wambaugh’s inversion test, the “same proceeding” limitation cannot form part of the ratio. If we reverse the proposition – that prior observations need not arise from the same proceeding – the outcome in Kanaklata would not have changed since it did not turn on such a distinction at all. This means the limitation to same-proceeding observations is neither confirmed as ratio nor excluded as obiter; it is simply absent from the ratio. Therefore, the Court’s reading of the same proceeding limitation is not a legitimate reading of the ratio but an interpolation of a restriction the ratio of Kanaklata never imposed.
When we apply the Cross and Harris test, where ratio decidendi is the rule of law which the judge treated as a necessary step in reaching the conclusion, then confining Kanaklata to same-proceeding observations collapses into a trivial proposition. Its doctrinal force lies precisely in recognising that the effect of strongly worded observations travels across proceedings. In Kanaklata, the operative concern was not the procedural stage at which such observations were made but their strength and tenor and whether they would create a reasonable apprehension in the mind of a reasonable observer that the Court committed itself to a particular view.
In the Sisodia bail order, the Court does not merely find a prima facie case but holds that the applicant “was actively involved… thereby committing the offence of money laundering.” Similarly, in the Kejriwal order, the court on prima facie grounds held the applicant guilty at that stage for committing the said offence. While both orders expressly couch their findings in terms such as “prima facie” and “at this stage,” the reasoning, at points, employs language that appears to approach a conclusive assessment of culpability.
From the perspective of legal theory, these are the “strongly worded” observations that trigger the Kanaklata ratio. Because the Court has, in prior proceedings, recorded strongly worded prima facie findings on the applicant’s role in a related order, a litigant would have a “reasonable apprehension” that the judge has a “preconceived notion” that cannot be “retraced.”
III. The ABAP Ground: Attendance at Events of an Ideologically Affiliated Organisation
The Court’s reasoning on the participation of Justice Sharma in the Akhil Bharatiya Adhivakta Parishad’s (ABAP) events rests on a critical conflation. The Court argues that the events attended by Justice Sharma, although organised by a political organisation, were not political but rather institutional/academic engagements with the members of the Bar. The applicable normative framework is provided in The Restatement of Values of Judicial Life and the Bangalore Principles of Judicial Conduct centring the standard of perceived impartiality. The Bangalore Principles urge judges to ensure that their conduct out of the courtrooms to maintain and enhance the confidence of the public and their impartiality and similarly in The Restatement of Values of Judicial Life (¶ 6 and 16). On a holistic reading these sources discourage the engagement of judges with organisations whose defining institutional character includes political affiliation.
The Court erroneously puts the official functions organised by universities and law school on the same footing as that of an event, albeit educational, organised by a political organisation. Justice Oka distinguishes between a judge attending a function organised by a law school/university, even if it is run by trusts or societies headed by politicians and a function organised by a political organisation. The judgement does not meaningfully engage with this distinction. Instead, it relies on the absence of political content in her speeches and the general permissibility of judicial participation in legal events. Ideally the order ought to have engaged with the ethical framework regarding the judicial conduct and apply it to the specific context. This would have required to distinguish between events organised by educational and political organisations, and then assess from a reasonable observer’s standpoint, whether such participation would undermine perceived impartiality in the eyes of a reasonable observer. Unfortunately, the order undertakes none of these steps.
IV. The Catch-22 Reasoning: A Problematic Holding
The Court puts forth the argument that the litigant had put the Court in a catch-22 situation where any outcome would suit applicant’s narrative. If the Court recuses it would position his claim that his allegations had substance and if the Court doesn’t recuse then he could claim the outcome was predicted. The reasoning adopted by the Court is an argumentum ad consequentiam (Appeal to Consequences) which does not address the merits of the recusal application, rather it addresses the political optics of the outcome. These are different questions. Furthermore, the “catch-22” logic could be adopted to any recusal application in any high-profile political case, where the applicant will always have a narrative available regardless of outcome, which, if accepted, would render recusal doctrine functionally inapplicable in precisely those cases where it is most needed.
V. Conflict of Interest: The Family Empanelment Ground
One of the eleven grounds advanced by the litigant which was admitted even after the judgement was reserved, was the empanelment of Justice Sharma’s children by the applicant’s rival advocate i.e., the Solicitor General. The Court negates this argument citing Virbhadra Singh v. CBI and NJAC case. Moreover, it is true that the children of Justice Sharma have no involvement in the excise policy case and hence no direct nexus. However, the Court’s reasoning has partial force.
The argument raised by Kejriwal is not the one the Court answers. The conflict is not between the Justice’s children’s empanelment and the current case. It is about the conflict between the Solicitor General’s dual role as he is both a) the professional authority who determines the volume and nature of government work flowing to the judge’s children, and (b) the advocate actively seeking an outcome in proceedings before her.
VI. The Missing Engagement: Satyender Kumar Jain case
The Delhi High Court in Satyendra Kumar Jain v. Directorate of Enforcement had upheld the transfer order of the Principal District and Sessions judge. Kejriwal relied on this order and argued his facts have a stronger footing than that of the Enforcement Directorate in the Satyender Kumar Jain case. However, the Court neither substantively discusses nor differentiates the decision in Satyender Kumar Jain. As compared to the length at which the Court tries to distinguish the Kanaklata case from this case, no such engagement with the Satyender Kumar Jain case was done, where a transfer was upheld because justice must “be seen to be done” from the perspective of the “mind of the party” rather than the judge. The absence of engagement is not merely a gap but a doctrinal omission. Where a coordinate bench has upheld transfer on the basis of perceived impartiality, the failure to distinguish or reconcile that reasoning weakens the analytical coherence of the present order, especially when it was expressly relied upon by Kejriwal.
Motive for recusal
A consistent thread in the judgement is the conflation of Kejriwal’s subjective motivations with the objective legal standard for recusal. A party may seek recusal for strategic reasons and still have a legally valid apprehension of bias. Conversely, a party may seek recusal in complete good faith and have a legally invalid application. Motive is simply irrelevant to the merits of the application.
The recusal test is explicitly objective: it asks what a fair-minded and informed observer would think, not what the applicant himself thinks or why. By repeatedly directing attention to Kejriwal’s motivations and conduct, the order conflates the subjective question (is he acting in good faith?) with the objective question (does the apprehension meet the legal standard?). This is ad hominem because it attacks the person rather than the argument.
The Deeper Doctrinal Problem
Justice Sharma’s order, for all its length, ultimately rests on a single structural move: the judge who is the subject of the recusal application is also the one deciding it. Such a determination sits in evident tension with the principle of nemo judex in causa sua (no one should be a judge in their own cause). However, this is not the first instance where a recusal application is heard by the very judge whose recusal is sought. Justice Sharma acknowledges that deciding one’s own recusal is inherently awkward, but holds that the alternative — stepping aside without hearing the application — would itself set a wrong precedent. A system where the applicant for recusal must argue before the very judge whose recusal is sought, without any pre-determined alternative forum for that argument, is not a system designed to maximise confidence in its own impartiality.
This concern is compounded when viewed alongside the opacity of judicial allocation. As noted earlier, multiple judges are assigned to hear criminal revision petitions, yet there exists no articulated principle explaining why this particular matter came to be listed before this bench. When the power of roster allocation operates without transparent criteria, and recusal challenges to that allocation are adjudicated within the same closed institutional loop, the result is a form of structural circularity. The issue, therefore, is not merely one of individual bias, but of an institutional design that limits meaningful scrutiny of both allocation and adjudication. This underscores the need to revisit the discretionary and opaque exercise of the Master of the Roster power.
Conclusion
The recusal judgement presented an opportunity to become a watershed moment in the recusal jurisprudence in India. In spite of having the opportunity to decide on recusal applications on numerous occasions before constitutional courts, the recusal jurisprudence has still not crystallised. The question, therefore, is not whether recusal doctrine exists, but whether it is being operationalised in a manner that sustains public confidence in judicial impartiality. Addressing that question requires moving beyond individual cases toward a more coherent and transparent framework for both judicial allocation and recusal.