The Jurisprudence of Arbitrability

Why Arbitrability Is the Hidden Fault‑Line of Arbitration:

Every modern arbitration clause carries a quiet assumption: that the dispute can, in fact, be arbitrated. Yet from Booz Allen & Hamilton v. SBI Home Finance (2011) to Vidya Drolia v. Durga Trading (2020) in India, and from the New York Convention (1958) to the UNCITRAL Model Law (1985), courts and legislatures have drawn—and redrawn—the shifting boundary of what is “arbitrable”. This blog, “The Jurisprudence of Arbitrability”, is about that fault‑line.

Arbitrability is not about whether an arbitration agreement exists or is valid; it is about what the parties are legally allowed to submit to private adjudication, and who is legally capable of doing so. Even the most carefully drafted clause collapses if the subject matter, or the parties themselves, fall on the wrong side of this divide.

If arbitration is the architecture of private justice, arbitrability is its planning permission. This blog invites you to walk along the boundary walls—where contract meets statute, autonomy meets sovereignty, and private bargains meet public values—and to see how each doctrine, case, and provision quietly redraws the map.

Chapter 1: Conceptual Foundations of Arbitrability

Arbitration clauses promise a private judge. Arbitrability decides whether the law will let that promise live. From the Geneva Protocol of 1923 and the New York Convention of 1958 to the UNCITRAL Model Law of 1985, one question runs like a fault‑line through arbitration statutes: which disputes are legally fit for arbitration, and which are reserved to state courts? Arbitrability is the label lawyers use for that boundary.

What Is “Arbitrability”?

In arbitration jurisprudence, arbitrability means the legal capacity of a dispute to be resolved by arbitration rather than by public courts. Put differently: even if parties have a valid arbitration agreement, the law may still say, “This type of dispute stays with the courts.”

Modern writers distinguish two dimensions:

• Objective arbitrability – whether the subject‑matter of the dispute can be arbitrated (e.g., commercial contracts often yes; criminal offences, status and some insolvency matters usually no).

• Subjective arbitrability – whether the parties themselves are allowed or empowered to arbitrate at all (e.g., States, public bodies, minors, or certain entities may face capacity limits).

Most legal systems treat arbitrability as a matter of public order or mandatory law, not something parties can rewrite at will.

Valid Arbitration Agreement vs Arbitrability of Subject‑Matter

Arbitration law begins with the agreement: it must exist, be valid, and typically be in writing (see New York Convention, Art. II(1)–(2)).

• Validity of the arbitration agreement asks:

– Did the parties consent?

– Is the clause properly formed, not null, void, inoperative, or incapable of being performed?

• Arbitrability of subject matter asks:

– Is this type of dispute legally referable to arbitration at all, even if the clause is perfectly valid?

Article V(2)(a) of the New York Convention (1958) makes the distinction explicit: a court may refuse enforcement if “the subject matter of the difference is not capable of settlement by arbitration under the law” of the enforcing state, even if the arbitration agreement itself is sound. This separation is reinforced by the doctrine of separability: the arbitration agreement is treated as juridically independent from the main contract, so its validity is tested separately. But separability does not make every dispute arbitrable; it only preserves the agreement long enough for arbitrability to be tested.

How Arbitrability Evolved: Common Law and Civil Law

Arbitration has deep roots in both families of law, but the road to modern arbitrability doctrine has been uneven.

• Common lawEarly English law was hostile to ousting the courts’ jurisdiction, but the Arbitration Act 1889 and later the Arbitration Act 1996 progressively embraced party autonomy, while keeping certain fields—like status, criminal matters, and some public law claims—outside arbitration on public policy grounds. Courts used arbitrability to mark domains of “exclusive jurisdiction” of state courts (e.g., some corporate, real property and insolvency disputes).

• In the United States, the Federal Arbitration Act 1925 began a long arc that, after the mid‑1980s, turned strongly pro‑arbitration, extending enforceability even to standard‑form consumer and employment clauses—sparking intense debate about the erosion of public rights and open justice.

• Civil law (Romano‑Germanic systems):

  France, Germany, Sweden, Finland and others codified arbitrability more explicitly: typically permitting arbitration of pecuniary/private rights while excluding matters involving status, public registers, or exclusive court powers. German law, for example, allows “any property claim” to be arbitrated, provided it is not within the exclusive jurisdiction of state courts and there is a valid arbitration agreement—mirroring objective and subjective arbitrability. French law has evolved into one of the most “pro‑arbitration” regimes in Europe, especially in international arbitration, narrowing non‑arbitrable zones largely to core public policy issues.

  
  

  Across both traditions, the 1958 New York Convention and the 1985 UNCITRAL Model Law pushed convergence: arbitration is broadly favored (“favor arbitrati”), but each state reserves a non‑arbitrable core anchored in its own public policy.

  
  

  Why Limit Arbitrability at All?

  If arbitration is contractual, why not simply let consenting adults arbitrate everything?

  Three classic justifications recur:

• Public policy and public order: Certain disputes engage values the state insists on controlling—criminal liability, family status, some competition law, or high‑impact regulatory questions. Many systems treat public policy, good morals, and public security as structural limits on arbitrability.

• Sovereign functions and public power: Acts where the State exercises sovereign authority, or disputes that shape core institutional frameworks (e.g., some administrative or constitutional issues), are viewed as non‑delegable to private tribunals.

• Inalienable or status‑based rights: Matters such as marriage, guardianship, legal capacity, or some succession rights are regarded as so closely tied to personal or collective status that parties cannot contract them into private adjudication—even if they wish to.

  
  

  These justifications surface in statutory bars (lists of non‑arbitrable disputes) and in open‑textured judicial tests invoking “public order” or “good morals”. The result is a moving frontier, not a closed list.

  
  

• Consent‑Based vs Law‑Based Exclusion

  Exclusions from arbitration arise in two very different ways:

• Consent‑based exclusion: parties decide, by contract, that certain issues or remedies are not to go to arbitration, even though the law would allow it. This is essentially about the scope of the arbitration agreement—an issue of party autonomy.

• Law‑based exclusion: legislation or case law declares that certain disputes or parties cannot be arbitrated, regardless of what the contract says. This is where arbitrability bites: the state draws a red line that consent cannot cross.

• Modern debates over arbitrability often turn on whether a limitation is properly characterized as a question of consent/scope or as a mandatory law barrier—a distinction that affects who decides (court or tribunal) and how much judicial review is available.

Objective vs Subjective Arbitrability

Bringing these strands together, contemporary doctrine frames arbitrability along two axes.

1. Objective arbitrability (subject‑matter):

• Is the category of dispute legally amenable to arbitration?

• Typically excluded: criminal offences, family status, many bankruptcy and winding‑up proceedings, some public law and regulatory matters, and certain real‑property or corporate disputes in specific jurisdictions.

• Governed by the lex fori or lex arbitri, and crucial at the enforcement stage under Art. V(2)(a) of the New York Convention.

  
  

  2. Subjective arbitrability (capacity of parties):

• Are these specific parties legally capable of entering a binding arbitration agreement?

• Some systems restrict the capacity of States, public entities, minors or persons under disability; others require special authorization for State‑owned entities, or limit arbitration of certain public assets.

In civil law systems, this duality is explicit; in common law, it is often embedded in case law and statutory structure rather than in definitions.

Why Some Disputes Remain Non‑Arbitrable Despite Consent

At the heart of arbitrability lies a basic tension: arbitration is built on consent, but constrained by public law. Even where parties:

• validly agree in writing,

• clearly define the subject matter, and

• choose an arbitral institution and seat,

courts may still refuse to compel arbitration or enforce an award if:

• the dispute touches a reserved domain (e.g., criminal law, core insolvency, status, certain corporate or property issues) or

• one party lacks capacity under domestic law to arbitrate (e.g., public bodies without statutory authority) or

• enforcing the award would violate public policy or international public order.

Thus arbitrability polices the outer boundary of party autonomy. It ensures that private justice does not displace the state’s monopoly over certain rights, remedies, and public interests.

In the chapters that follow, this blog will trace how courts and legislatures—especially in India and leading Model Law jurisdictions—have drawn, redrawn, and sometimes blurred that boundary between what parties want to arbitrate and what the law will actually let them arbitrate.

Chapter 2: Objective and Subjective Arbitrability: Doctrinal Distinctions

Arbitration is built on consent. Arbitrability is about its limits. This chapter asks a deceptively simple question: can parties arbitrate everything they agree to? Modern jurisprudence answers: no—and the reasons why define the doctrine of arbitrability.

• Objective Arbitrability: What Kind of Dispute Can Be Arbitrated?

Objective arbitrability (subject‑matter arbitrability) concerns what can be sent to arbitration.

The classic starting point is Art. II & V(2)(a) of the New York Convention, 1958, which allow States to exclude certain disputes from arbitration. The UNCITRAL Model Law, 1985 (Art. 34(2)(b), 36(1)(b)) echoes this, tying non‑arbitrability to public policy.

Courts typically exclude disputes that:

• affect status (e.g., marriage, divorce, custody)

• concern public enforcement regimes (criminal law, competition, securities, insolvency)

• involve sovereign functions

In Mitsubishi Motors v. Soler Chrysler‑Plymouth, 473 U.S. 614 (1985), the U.S. Supreme Court famously allowed antitrust claims to be arbitrated, signalling a pro‑arbitration turn, but left a “public policy escape” at the enforcement stage.

In India, Booz Allen & Hamilton Inc. v. SBI Home Finance Ltd., (2011) 5 SCC 532 crystallised the test: disputes involving rights in rem are generally non‑arbitrable; rights in personam are usually arbitrable, with exceptions.

• Subjective Arbitrability: Who Can Validly Arbitrate?

Subjective arbitrability addresses who can be a valid party to arbitration.

Key dimensions:

1. Capacity of Parties

   • Minors / persons under disability: Most jurisdictions require representation by guardian / legal representative; some systems treat agreements purporting to bind minors as void or voidable (e.g., Indian Contract Act, 1872, ss. 11–12 read with Arbitration and Conciliation Act, 1996, s. 7).

   • Insolvency administrators: Once insolvency/winding‑up commences, the insolvency estate is subject to collective procedures; the trustee/administrator may be bound or may override arbitration in favour of insolvency courts (e.g., IB Code, 2016, ss. 14, 238 in India; U.S. Bankruptcy Code, automatic stay, 11 U.S.C. § 362).

   • Government bodies / States: Today, States widely arbitrate commercial disputes, yet certain matters—tax, criminal prosecution, core regulatory functions—remain non‑arbitrable. ICSID Convention (1965) introduced investor–State arbitration, pushing the frontier of subjective arbitrability.

2. Immunity & Public Law EntitiesState immunity doctrines previously insulated States; the modern shift to restrictive immunity allows arbitration of jure gestionis (commercial) acts, while jure imperii (sovereign) acts remain non‑arbitrable.

• Arbitrability in Multi‑Party and Composite Transactions

Real‑world disputes rarely follow the neat lines of a single contract between two parties.

• Multi‑party disputes (shareholders, subsidiaries, guarantors, sub‑contractors)

• Composite transactions (web of connected contracts: EPC, finance, guarantees, assignments)

  Key question: can an arbitral tribunal assume jurisdiction over non‑signatories or “connected” contracts?

  Different legal systems have developed doctrines:

• Group of companies doctrine (e.g., Dow Chemical v. Isover Saint Gobain, ICC Award 1982)

• Estoppel, implied consent, alter ego, assumption under U.S. law

• Indian law’s nuanced approach in Chloro Controls India Pvt. Ltd. v. Severn Trent Water Purification Inc., (2013) 1 SCC 641 and further refined in Cox & Kings Ltd. v. SAP India (2023, 7‑Judge Bench) on the group‑of‑companies doctrine.

These cases expand subjective arbitrability by bringing non‑signatories within the fold, but only where consent (express or inferred) and transactional unity are convincingly shown.

• Statutory Restrictions on Party Autonomy

Party autonomy is not absolute. Legislatures draw bright‑line exclusions.

Examples:

• Insolvency & bankruptcy: Collective, court‑supervised processes. Many courts treat core insolvency issues (admission of claims, distribution, resolution plans) as non‑arbitrable. Indian Supreme Court: Delhi Airport Metro Express v. DMRC, (2022) 1 SCC 131 and earlier IBC line of cases confirm the primacy of insolvency fora.

• Matrimonial, guardianship, testamentary disputes: Status‑oriented, requiring public adjudication.

• Criminal liability: Not privately compromisable (though civil consequences arising from fraud/breach may be arbitrable—see below).

• Competition / antitrust: Once considered off‑limits, now increasingly arbitrable post‑Mitsubishi, subject to public law safeguards at enforcement.

Statutes may either:

• explicitly bar arbitration (e.g., “no civil court shall have jurisdiction and matters shall be decided only by…”), or

• implicitly exclude it by creating specialised tribunals with exclusive jurisdiction.

• Judicial Tests to Determine Arbitrability

Courts deploy a range of tests, often overlapping:

1. Rights in rem vs rights in personam

   • Rights in rem (status, property against the world) → typically non‑arbitrable.

   • Rights in personam (inter‑partes obligations) → presumptively arbitrable.

     
     

     Developed strongly in India: Booz Allen (2011), A. Ayyasamy v. A. Paramasivam, (2016) 10 SCC 386.

2. Public Policy / Public Interest TestDoes the dispute touch non‑delegable public interests, sovereign functions, or regulatory structures? If yes, courts tend to withhold it from arbitration.

3. Competence of Tribunal vs Institutional DesignIs the matter one that the legislature intended to be handled only by specialised courts/tribunals? If so, arbitration is ousted. E.g., securities regulation, core labour adjudication in many jurisdictions.

4. Severability and “Core vs Peripheral” AnalysisSome courts carve out non‑arbitrable core issues (e.g., validity of IP registration) while allowing arbitration of contractual or licensing disputes surrounding those rights.

Internationally, arbitration‑friendly seats such as France, Switzerland, England, Singapore adopt a narrow, exceptional list of non‑arbitrable matters, and broadly favour arbitration subject to overriding public policy.

• Can Parties Arbitrate Everything They Consent To?

Consent is necessary—but not sufficient. Modern jurisprudence filters party autonomy through three main lenses:

1. Structural Public InterestsSome disputes are instruments of public order, not private bargains: criminal law, family status, core insolvency, constitutional rights. These are structurally non‑arbitrable, irrespective of consent.

2. Regulatory ArchitectureWhere the legislature centralises enforcement in specialised fora (competition, securities, certain labour and consumer disputes), arbitration may be limited or subordinated to regulatory oversight.

3. Capacity and RepresentationConsent is meaningful only if given by those with legal capacity and within their authority: minors, persons under disability, sovereigns and public entities all raise subject‑matter and capacity questions that courts scrutinise closely.

   
   

The global trajectory, from 1958 (New York Convention) to today’s pro‑arbitration jurisprudence, reveals a clear trend:

• Presumption in favour of arbitrability for commercial, inter‑partes disputes.

• Targeted, principled exclusions where public values, institutional design, or legal capacity demand it.

The doctrine of arbitrability, therefore, is not an attack on party autonomy. It is its constitutional boundary: ensuring that private justice thrives, but never at the expense of the legal system’s foundational commitments to public order, vulnerable parties, and sovereign functions.

In the next chapter, this tension resurfaces in a different form—when and by whom arbitrability is examined: pre‑reference courts, arbitral tribunals (Kompetenz‑Kompetenz), or enforcement judges.