Vidya Drolia and the four-fold test: the Supreme Court reorders the law of arbitrability
On 14 December 2020 a three-judge bench of the Supreme Court, in Vidya Drolia v. Durga Trading Corporation, restated and tightened the in rem / in personam taxonomy of Booz Allen into a structured four-fold test for non-arbitrability, held tenancy disputes under the Transfer of Property Act arbitrable, overruled N. Radhakrishnan on the arbitrability of fraud, and recalibrated the standard of judicial review under Sections 8 and 11 in favour of competence-competence. A close reading of Justice Sanjiv Khanna's lead judgment, Justice Ramana's concurring opinion, the doctrinal architecture and what the bar should plead in the post-Vidya Drolia world.
- Court
- Supreme Court of India
- Citation
- (2021) 2 SCC 1; 2020 SCC OnLine SC 1018
- Bench
- N.V. Ramana, J., Sanjiv Khanna, J., Krishna Murari, J.
- Decided
- 14 December 2020
Vidya Drolia v. Durga Trading Corporation is the case that turned the arbitrability inquiry from a list-based exercise into a structured test. The 2011 articulation in Booz Allen & Hamilton v. SBI Home Finance had supplied an analytical vocabulary — rights in rem versus rights in personam — and an illustrative catalogue of six non-arbitrable categories. A decade of post-Booz Allen litigation had pulled at the edges of both. Vidya Drolia tightened the analytical core, restated the catalogue as a working rule, redirected the standard of review at the referral stage, and overruled two lines that had become doctrinal embarrassments — N. Radhakrishnan v. Maestro Engineers on fraud and Himangni Enterprises v. Kamaljeet Singh Ahluwalia on tenancy.
The judgment is also institutionally interesting. Sanjiv Khanna, J. authored the main judgment for himself and Krishna Murari, J.; N.V. Ramana, J. — then second senior puisne — wrote a separate concurring opinion. The concurrence does not depart from the result; it supplies an independent route to it and adds emphasis on the competence-competence dimension that has, in turn, been the part of the judgment most relied upon by later benches.
The architecture of the question
The arbitrability question in Indian law has three loosely connected dimensions, and the confusion of the post-Booz Allen decade was largely the confusion of running them together.
The first is the subject-matter dimension: is the dispute, as a matter of its substantive content, capable of resolution by arbitration? The Booz Allen taxonomy spoke to this — rights in rem, by their nature, bind the world and are not appropriately committed to a private tribunal whose award binds only the parties.
The second is the statutory dimension: has Parliament — expressly or by necessary implication — taken a class of disputes out of the arbitral domain by conferring exclusive jurisdiction on a particular forum? The Rent Control Acts, the Industrial Disputes Act, the Family Courts Act and the Debt Recovery Tribunals architecture each illustrate this.
The third is the forum-allocation dimension at the gateway: when these objections are raised at the Section 8 (reference of a court action to arbitration) or Section 11 (court appointment of an arbitrator) stage, how deeply should the court engage? Should it conduct a full-dress arbitrability hearing, or defer to the tribunal under Section 16 (kompetenz-kompetenz)?
Vidya Drolia speaks to all three.
The factual matrix
The reference arose out of a tenancy dispute between Vidya Drolia and Durga Trading Corporation governed by the Transfer of Property Act, 1882, not by a State rent-control statute. The lessor invoked the arbitration clause; the lessee resisted on the ground that tenancy disputes were inherently non-arbitrable. The two-judge bench that first heard the matter noted a conflict in the existing authority. Booz Allen had listed eviction and tenancy under rent control as non-arbitrable; Himangni Enterprises had, in 2017, extended the non-arbitrability proposition to all tenancies, including those governed by the Transfer of Property Act, on the ground that landlord-tenant disputes inherently engage rights in rem. The two-judge bench, finding Himangni doctrinally unstable, referred the matter to a larger bench.
A connected reference required the Court to revisit N. Radhakrishnan v. Maestro Engineers (2010), which had held that disputes involving allegations of "serious fraud" were non-arbitrable — a holding that had generated a decade of confused interlocutory practice in commercial arbitrations and had been the subject of repeated qualifying gloss in A. Ayyasamy v. Paramasivam (2016) and Rashid Raza v. Sadaf Akhtar (2019).
A third strand concerned the arbitrability of debt-recovery disputes that would otherwise have been within the exclusive jurisdiction of the Debt Recovery Tribunals under the Recovery of Debts and Bankruptcy Act, 1993. The Delhi High Court's Full Bench in HDFC Bank v. Satpal Singh Bakshi (2013) had held such disputes arbitrable; the correctness of that view was in question.
The three-judge bench took up the three strands together. The judgment is structured around them, but its doctrinal contribution is the cross-cutting four-fold test.
The four-fold test
Justice Khanna's articulation is the doctrinal centre of the case. A dispute is non-arbitrable when:
(i) the cause of action and subject-matter of the dispute relate to actions in rem that do not pertain to subordinate rights in personam arising from rights in rem;
(ii) the cause of action and subject-matter of the dispute affect third-party rights, have erga omnes effect, or require centralised adjudication where mutual adjudication would not be appropriate or enforceable;
(iii) the cause of action and subject-matter of the dispute relate to inalienable sovereign and public-interest functions of the State — and accordingly, mutual adjudication would be unenforceable;
(iv) the subject-matter of the dispute is expressly or by necessary implication non-arbitrable under a mandatory statute.
The four prongs are not watertight. They overlap — a class of disputes might fail (iii) and (iv) simultaneously, and a class might satisfy (i) but be saved by the in personam subordinate-rights qualification. But the test does two important things together. It restates Booz Allen's in rem / in personam axis as the first prong, retaining the doctrinal continuity that two decades of judicial work had built up. And it adds the second, third and fourth prongs to capture categories that the binary in rem / in personam axis did not comfortably hold — third-party effect, sovereign function, statutory exclusion.
The bench applied the test category by category.
Tenancy under the Transfer of Property Act — arbitrable. The dispute is between landlord and tenant inter se, it does not bind the world, and the Transfer of Property Act does not contain an exclusive-forum clause of the kind that would engage the fourth prong. Himangni Enterprises — which had reasoned to the contrary — was overruled to the extent it held the contrary.
Tenancy under rent-control legislation — non-arbitrable. The Rent Control Acts, in conferring exclusive jurisdiction on specially designated forums and in granting protections to tenants that engage public-interest considerations, satisfy the fourth prong; the Booz Allen listing on this point stands.
Fraud — arbitrable. N. Radhakrishnan was overruled. The Court held that allegations of fraud — including allegations of "serious fraud" — do not, of themselves, take a dispute out of arbitration. Where the allegation has a public-law dimension (criminal proceedings, fraud on the State) the public-law remedies remain available; the existence of those remedies does not denude the arbitral tribunal of jurisdiction over the contractual dispute between the parties.
Debt-recovery disputes within the DRT's exclusive jurisdiction — non-arbitrable. The Recovery of Debts and Bankruptcy Act vests exclusive jurisdiction in the Tribunal and supplies a specialised remedial architecture (including third-party rights, asset attachment and a sui generis enforcement scheme). The fourth prong is satisfied. The Delhi HC Full Bench in HDFC Bank v. Satpal Singh Bakshi was overruled.
The standard of review at the gateway
The second pillar of the judgment is the recalibration of the standard at the Section 8 and Section 11 stage. The pre-2015 jurisprudence — anchored in SBP & Co. v. Patel Engineering (2005) and National Insurance v. Boghara Polyfab (2009) — had treated the referral court's jurisdiction as appellate in some respects, with the court routinely conducting a full inquiry into the existence and validity of the arbitration agreement. The 2015 Amendment had inserted Section 11(6A) (since deleted in 2019, but the underlying policy direction remains) which confined the court to "examination of the existence of an arbitration agreement."
Vidya Drolia aligns the arbitrability inquiry with that policy direction. The referral court should adopt a prima facie standard. Where the arbitrability objection is "manifest" — where no reasonable doubt exists that the subject-matter is non-arbitrable — the court may refuse reference. Where the objection requires a finely-balanced assessment, the court should refer the dispute to the tribunal and leave the determination to be made under Section 16.
Justice Ramana's concurring opinion presses this point particularly hard. The competence-competence principle — that the tribunal has jurisdiction to rule on its own jurisdiction — is statutorily entrenched in Section 16 and should not be subverted at the gateway by a referral court that takes upon itself the very determinations the tribunal is statutorily empowered to make. The concurrence formulates a "when in doubt, refer" working rule that has, in the subsequent jurisprudence, become the practical lodestar.
The reasoning that runs through
Three doctrinal threads weave through the judgment.
The in rem / in personam distinction, refined. Justice Khanna does not abandon the Booz Allen taxonomy; he qualifies it. Many disputes that have a notional in rem dimension nonetheless turn, at the level of the cause of action between the parties, on subordinate rights in personam. A mortgage suit under Section 67 of the Transfer of Property Act binds the property and is in rem; a dispute about the rate of interest on the loan is in personam. Vidya Drolia signals that the first prong is to be read with that subordinate-rights qualification expressly built in.
Statutory exclusivity and necessary implication. The fourth prong requires the court to ask whether Parliament has, by conferring exclusive forum jurisdiction and supplying a specialised remedial scheme, taken a class of disputes out of the arbitral domain. The test is not whether the statute mentions arbitration; the test is whether the legislative architecture is incompatible with private adjudication. The DRT framework was held incompatible; the Transfer of Property Act framework was held compatible.
The pro-arbitration default at the gateway. The third thread is the most consequential for daily practice. Where the arbitrability objection is not manifest, the doubt is resolved in favour of reference. The reasoning is institutional — the tribunal is the statutory adjudicator of its own jurisdiction, the court's residual review survives at the post-award stage under Section 34, and the policy of the 1996 Act is to minimise judicial intervention.
The Bench's working rule, paraphrased from the reasoning of Khanna, J.: the referral court is to interfere only where it is manifest that the dispute is non-arbitrable; in cases of doubt, the matter must be referred to arbitration, leaving the tribunal to decide arbitrability under Section 16.
The doctrinal contribution
Vidya Drolia is the moment Indian arbitrability law moved from a list to a test. Booz Allen had supplied vocabulary and a catalogue; Vidya Drolia supplied the analytical structure under which fresh categories could be assessed. The four prongs are now the working rubric in every Section 8 and Section 11 arbitrability contest.
The fraud-arbitrability overruling matters because it closes a particularly destructive interlocutory channel. Pre-Vidya Drolia, a respondent in a commercial arbitration could destabilise the proceedings by alleging fraud and pressing the N. Radhakrishnan line; the tribunal would often be obliged to grapple with the issue at the threshold; the arbitration would lose momentum and the case would migrate, in substance if not in form, into the courts. The post-Vidya Drolia rule treats the fraud allegation as a substantive contention to be adjudicated within the arbitration unless the allegation has an independent public-law dimension.
The tenancy clarification matters at the granular level — a substantial class of landlord-tenant disputes that had been held non-arbitrable post-Himangni returned to the arbitral domain. The State rent-control distinction holds: the Rent Control Acts continue to oust arbitration.
The DRT clarification matters in a more limited domain — banking and financial-creditor recovery actions — but the analytical move (statutory framework as exclusionary by necessary implication) has been extended in subsequent cases.
The competence-competence emphasis matters most of all. The bar's experience after Vidya Drolia has been that referral courts cite the judgment for the proposition that, in doubtful cases, the dispute should go to the tribunal. The Section 11 docket has, by perceptible measure, become quicker.
What the Court did not decide
A few matters were left open or only obliquely addressed.
The exact contours of "manifest" non-arbitrability. The judgment does not supply a definitional test for what makes an objection manifest. Later benches have, in practice, calibrated the standard around two questions — whether the non-arbitrability proposition rests on a single, undisputed legal proposition, and whether the underlying facts that engage the legal proposition are themselves undisputed.
The interaction with stamping. The judgment does not address whether an unstamped or insufficiently stamped arbitration agreement is part of the Section 11 gateway inquiry or of the tribunal's Section 16 inquiry. That question was answered first by N.N. Global in 2023 (in the affirmative, holding the agreement could not be acted upon) and then re-answered by the seven-judge bench in In Re Interplay in December 2023 (in the negative, holding stamping is for the tribunal and the agreement is enforceable in the interim). The full arc is traced in the stamping-arc digest.
The arbitrability of consumer disputes and intellectual-property disputes. The judgment touches each only in passing. Consumer disputes had already been held non-arbitrable in Emaar MGF Land v. Aftab Singh (2019) — Vidya Drolia leaves that intact, and the Emaar result fits comfortably within the fourth prong (the Consumer Protection Act being read as conferring exclusive jurisdiction). IP disputes — the arbitrability of trademark infringement, of patent validity, of copyright enforcement — remain a live area, with the analytical structure of Vidya Drolia applicable but the prong-by-prong analysis case-by-case.
The treatment of Section 9 relief. The judgment is concerned with the referral inquiry. The arbitrability question as it arises at the Section 9 stage — where the court is asked for interim relief before or during arbitration — was not directly addressed. The conventional position remains that Section 9 relief is available wherever the arbitration is itself a permissible mode of dispute resolution, and the bar has tended to argue Vidya Drolia as supportive of a similar deferential standard at the Section 9 gateway.
The doctrinal arc
The arbitrability line in Indian arbitration law runs through three authorities. Booz Allen & Hamilton v. SBI Home Finance (2011) supplied the vocabulary and the catalogue. Vidya Drolia (2020) restated Booz Allen as a structured four-fold test and recalibrated the gateway standard. Cox & Kings Ltd v. SAP India (2023, Constitution Bench) extended the analytical work to the related but distinct question of who is bound by the arbitration agreement, re-anchoring the group-of-companies doctrine in the consent-based definition of "party" under Section 2(1)(h) read with Section 7.
The stamping arc — N.N. Global (April 2023, 5-judge CB, 3:2) → In Re Interplay between Arbitration Agreements under the Arbitration and Conciliation Act, 1996 and the Indian Stamp Act, 1899 (December 2023, 7-judge CB, unanimous overruling) — operates on a parallel plane. The 2023 seven-judge bench's reaffirmation of the separability of the arbitration agreement and of kompetenz-kompetenz draws explicitly on Vidya Drolia's gateway-discipline reasoning.
The standard-of-review thread in Vidya Drolia has, in turn, fed into the discussion around the draft Arbitration & Conciliation (Amendment) Bill 2024 and into the Court's Section 11 docket-management practice — referral courts now routinely cite the "prima facie" formulation as their working standard.
The practitioner's take
For the party resisting reference. The post-Vidya Drolia rule is unfavourable. The threshold for non-arbitrability is manifest; doubtful objections will be referred. Pleading discipline: assemble the non-arbitrability argument so that it rests on (a) a single, dispositive legal proposition that admits no reasonable counter-argument, and (b) undisputed underlying facts. A general invocation of fraud, or a generalised challenge based on third-party effect, will be sent to the tribunal.
For the party seeking reference. Press the "manifest" threshold. Where the resisting party invokes Booz Allen's six-category list, isolate the relevant category and run the four-fold test through it. Tenancy under TPA — arbitrable; tenancy under rent control — non-arbitrable; the distinction is now textually anchored.
For the tribunal. Vidya Drolia trades the Section 11 gateway burden for Section 16 tribunal burden. Tribunals need to be procedurally serious about hearing arbitrability objections early in the proceedings, recording a reasoned interim award on the point, and proceeding only on the basis of a clear jurisdictional ruling. A tribunal that defers the arbitrability question to the merits hearing risks producing an award that is exposed at the Section 34 stage.
For Section 11 practice. The "when in doubt, refer" working rule means that Section 11 pleadings should be lean. Arguments that require the court to read evidence, to construe contested clauses, or to make findings on disputed facts are very unlikely to succeed at the gateway. Save those arguments for the tribunal.
For drafting. A well-drafted arbitration clause should be alert to two arbitrability traps. First, the underlying dispute may engage a statute that has been read as conferring exclusive forum jurisdiction (Rent Control, Consumer Protection, DRT, Insolvency); arbitration in such areas is futile and should be ruled out at the drafting stage. Second, the clause should be drafted to survive the Section 16 tribunal-level scrutiny — separability, governing law and seat clauses should be express and unambiguous.
Related editorial pieces
- Booz Allen and the in rem / in personam taxonomy: the foundation of Indian arbitrability
- Cox & Kings and the group-of-companies doctrine: a Constitution Bench re-anchors the law of non-signatories
- N.N. Global and In Re Interplay: the eight-month arc on stamping and the arbitration agreement
- Silence is consent: the Supreme Court's reading of Section 29A waiver in arbitration
Related reading
Booz Allen v. SBI Home Finance: the foundational taxonomy of arbitrability and its six-category illustrative list
TRF v. Energo: the Supreme Court's first articulation of ineligibility-cannot-nominate
Perkins Eastman v. HSCC: the extension from ineligible-cannot-nominate to ineligible-cannot-unilaterally-appoint
Trace how this proposition has been treated across Indian courts — citations, bench strength, and subsequent history — in one workspace built for litigators.