The Jurisprudence of Arbitrability (PART 2)

Arbitrability at Different Stages of Proceedings:

How Far Can Courts Go, and When Should They Stop?

Arbitration lives or dies on one deceptively simple question: is this dispute even allowed to go to arbitration? That is the problem of arbitrability—and it shadows the parties from the very first suit in court to the last attempt to enforce an award.

This chapter follows that journey.

Arbitrability: The Thread Running Through the Life of a Dispute

Arbitrability is about the “suitability” of a dispute for resolution by arbitration and whether any resulting award will be enforceable under national and international law . Some matters (e.g., core criminal law, certain administrative or enforcement procedures, or issues reserved to the state) are simply off-limits to arbitration in many jurisdictions.

National courts play gatekeeper at three decisive moments:

1. Pre‑reference stage – before or instead of constituting the tribunal.

2. During the arbitration – where the tribunal claims its own jurisdiction (Kompetenz‑Kompetenz).

3. At the award enforcement / setting‑aside stage – where the award meets state power and public policy.

Across these stages, the central tension is: How much should courts supervise, and how much should they stay out of the way?

Comparative Approaches to Court–Tribunal Power Sharing

A helpful way to picture the global landscape is to look at how different systems allocate the power to decide jurisdiction (Kompetenz‑Kompetenz) at pre‑award stages.

Who Decides Jurisdiction, and When?

Approach / Model	Core Idea	Example Jurisdictions & Notes
UNCITRAL / “Concurrent competence”	Courts and tribunal both have power; courts tend to defer and postpone full review until later.	Model Law states, many BRICS reforms; Russia’s 2016 reforms aim to align with this, limiting intervention in institutional arbitration.
“Tribunal first” / strong Kompetenz‑Kompetenz	Tribunal is the primary judge of its own jurisdiction; courts usually wait for a final award.	France; several jurisdictions postpone review of jurisdiction until the award stage (e.g., France, Sweden, Germany).
Hybrid / context‑specific	Sometimes court is first (e.g., validity of agreement), sometimes tribunal is first; depends on stage and grounds.	UK, Vietnam; England under the Arbitration Act 1996 allows de novo review but typically exercises restraint.

Figure 1: Comparative allocation of jurisdictional review between courts and arbitral tribunals.

Courts worldwide increasingly enforce arbitration agreements and awards, but they disagree on how soon and how deeply they may probe jurisdiction and arbitrability.

Pre‑Reference Stage: Arbitrability Before the Tribunal Exists

This is where Section 8 and Section 11 of the Indian Arbitration and Conciliation Act, 1996 sit.

• Section 8 – when a suit is filed in court despite an arbitration clause, the defendant seeks a reference to arbitration. Recent amendments require courts to send parties to arbitration if prima facie a valid arbitration agreement exists.

• Section 11 – appointment of arbitrators. The 1996 Act originally gave courts wide powers; SBP & Co v Patel Engineering (2005) treated Section 11 as a judicial, not administrative, function, inviting deeper review of jurisdiction and arbitrability. This undercut Kompetenz‑Kompetenz.

• The 2015 amendment inserted Section 11(6A), aiming to shrink this inquiry to the bare question of whether an arbitration agreement exists, restoring autonomy to the tribunal (Pratap & Kavita, 2024). Though 11(6A) was later omitted in 2019, much of the case law continued to push courts back toward a limited prima facie review rather than a full trial on arbitrability at the reference stage.

Scholars warn that broad, pre‑reference judicial scrutiny risks “judicialisation” of arbitration, particularly where courts are not consistently pro‑arbitration. In short, pre‑reference arbitrability review is where judicial restraint matters most if arbitration is to get off the ground.

During the Proceedings: Kompetenz‑Kompetenz and Judicial Non‑Interference

Once the tribunal is constituted, who decides whether it has jurisdiction? The answer is the doctrine of Kompetenz‑Kompetenz: the tribunal may rule on its own jurisdiction, including objections to the existence or validity of the arbitration agreement.

• Section 16 of the Indian 1996 Act codifies this principle, mirroring the UNCITRAL Model Law and empowering the tribunal to decide its own jurisdiction in the first instance.

• Similar provisions appear across Model Law states and many other jurisdictions, with the common theme that courts should not derail ongoing arbitrations with early, heavy‑handed review.

Nigeria offers an extreme case: under its old Arbitration and Conciliation Act and the new Arbitration and Mediation Act 2023, judicial non‑interference and Kompetenz‑Kompetenz combine to oust court jurisdiction over tribunal jurisdictional rulings, sometimes even granting them finality. This is a rare, very pro‑arbitration reading: almost all other systems preserve some later judicial review.

The logic is simple: businesses choose arbitration to avoid the delays and appeals of national courts. If every jurisdictional objection could instantly halt proceedings in court, arbitration would lose its edge. But the price of non‑interference is that some jurisdictional defects may only be remedied much later.

Post‑Award Stage: Arbitrability at Enforcement and Set‑Aside

At the end of the journey, an arbitral award has to meet a national court—either where it was made (set‑aside / vacatur) or where it is to be enforced.

Under the New York Convention (NYC) 1958, arbitrability and public policy are classic grounds to refuse recognition or enforcement of an award.

How Often Do Courts Actually Enforce?

Empirical research of over 1,000 decisions from 2010–2020 shows that national courts:

• Enforced foreign awards in 73% of cases,

• Vacated awards in only 23% of set‑aside actions,

with relatively little variation across jurisdictions. This suggests a strong global pro‑enforcement bias—at least at the aggregate level.

Divergent National Practices

Yet underneath these numbers, there is striking diversity:

• UK & Australia – courts show flexibility. In cases such as Dallah (UK) and IMC Aviation v Altain Khuder(Australia), enforcing courts allowed parties to raise substantive jurisdiction and arbitrability challenges at the enforcement stage, even if no challenge was brought at the seat (Rosli et al., 2024). This protects fairness (no one should be bound without consent), but risks undermining finality.

• Malaysia – tends to defer strongly to the supervisory court at the seat, often treating unchallenged awards as effectively final and limiting re‑litigation at the enforcement stage (Rosli et al., 2024).

• Transnational trend – scholarship notes a gradual extension of arbitrable subject-matter, and argues that non‑arbitrability or public policy should bar recognition only in truly exceptional cases, ideally raised by the opposing party, not ex officio.

India’s journey has been turbulent, but recent Supreme Court decisions and statutory amendments reflect a consolidating pro‑enforcement stance, especially on “public policy” as a ground for setting aside or refusing enforcement, confining it to serious defects such as fundamental procedural unfairness or core illegality.

Prima Facie Review vs Final Determination

Behind all of this lies a methodological choice: When courts look at arbitrability and jurisdiction, how deep should they go, and when?

• Prima facie review – a quick, surface check: Is there ostensibly an arbitration agreement? Is the dispute facially arbitrable? If yes, send the matter to arbitration and leave detailed analysis to the tribunal, subject to later review. This approach is now common at the pre‑reference stage in many jurisdictions and recommended for Sections 8 and 11 in India.

• Final determination – a full merits‑level examination by the court (witnesses, documents, detailed law) that settles jurisdiction or arbitrability conclusively. This is usually reserved for the set‑aside or enforcement stage, although some hybrid systems allow earlier final rulings in specific contexts.

Comparative work shows three broad patterns:

1. Courts defer heavily and postpone final determination to the end.

2. Courts retain broad power, reviewing jurisdiction even at the outset.

3. Emerging “middle paths” that insist on prima facie review at the start and full review only once the award is made, to protect both efficiency and legality .

   
   

Judicial Restraint vs Judicial Supervision: The Core Analytical Tension

Every jurisdiction must answer the same question differently:

How much court intervention is enough to safeguard legitimacy, and how much is too much, destroying arbitral autonomy?

The Case for Judicial Restraint

• Arbitration depends on speed, finality, and party autonomy; constant or intrusive court interference drives parties back to litigation.

• Countries seeking to brand themselves “arbitration‑friendly” (e.g., London, Singapore, Hong Kong) cultivate a reputation for minimal interference, narrow public policy, and near‑automatic enforcement.

• A transnational trend pushes non‑arbitrability and public policy to the margins, reserving them for objectively serious issues that threaten fundamental norms.

The Case for Judicial Supervision

• Courts are the ultimate guardians of public policy, due process and non‑arbitrable subject‑matter—they cannot simply rubber‑stamp anything labelled an “award”.

• In systems where judicial culture is not uniformly pro‑arbitration, broad court review is often defended as necessary to correct errors, control fraud, and ensure that crucial sectors (e.g., consumer rights, administrative acts, some regulatory matters) are not privatized through arbitration.

• Some jurisdictions, like Pakistan historically, have leaned heavily into supervision, though at the cost of uncertainty and criticism for undermining international commitments.

The deepest contemporary work in this field cautions that full merits review of arbitral awards by courts is undesirable and corrosive, proposing instead a narrow, procedure‑focused, and public‑policy‑based review, with clear statutory limits and short time frames for challenge.

Where the Debate Seems to Be Heading

Comparative analysis across India, UK, Nigeria, BRICS states and others points to a converging model:

• Pre‑reference: courts apply prima facie review; doubts go to the tribunal.

• During proceedings: Kompetenz‑Kompetenz reigns; courts assist but do not decide jurisdiction (save in exceptional cases).

• Post‑award: courts perform a limited, but real, final determination focused on procedure, jurisdiction, arbitrability and narrow public policy.

In that model, arbitrability is not a one‑time hurdle; it is a moving checkpoint—but each checkpoint is carefully calibrated so that a dispute can actually travel from clause to award, without being endlessly pulled back into court.