How courts choose arbitrators

by Animesh Bordoloi and Prashant Narang.

A court hearing an application under section 11 of the Arbitration and Conciliation Act 1996 must answer two questions. Should an arbitrator be appointed? If so, who should it be? Indian arbitration law has paid far more attention to the first than to the second.

Our new working paper, How courts choose arbitrators: Evidence from the Bombay and Madras High Courts, examines the second question. It codes 372 reported and retrievable section 11 orders from 2016 to 2024, naming 387 appointments among 201 individuals, and draws on 36 semi-structured interviews.

A question the Act leaves open

The Act regulates the person once in view, covering agreed qualifications, independence and impartiality, disclosure, and fees. It says little about how candidates enter consideration, whether alternatives are weighed, or why one eligible person is preferred over another. Beyond qualifications agreed by the parties, the Act prescribes no positive qualification or accreditation criteria. Those norms are left to regulations by the Arbitration Council of India, which the Government told Parliament in February 2026 had still not been constituted.

This creates an observability problem. Judges may draw on professional knowledge, party suggestions or informal inquiries that never appear in the order. The record shows who was appointed and which steps were recorded. It cannot necessarily reveal how the choice was made.

What the orders record

The process is largely court-driven. In 315 of 372 orders (84.7%), the court named the arbitrator without recording both parties’ agreement to that person; such agreement appeared in 50 (13.4%). Consent to arbitration, or to “a retired judge”, is not agreement on who should hear the dispute.

Retired judges received 63.3% of named appointments, and advocates and senior advocates 33.9%. Technically qualified persons received six appointments across the two courts over nine years; former bureaucrats or senior officials, five.

Orders rarely explained the choice. Only 18 (4.8%) recorded any reason bearing on it, mostly a category of person such as a retired judge or legal expert. Six (1.6%) gave a ground tied to the individual appointed, and none compared candidates.

Disclosure practice differed sharply. Bombay directed a section 12 disclosure in 115 of 123 orders; Madras in four of 249. Section 11(8) places disclosure before appointment, yet 111 of the 119 directions came after the arbitrator was named. Some appointments may have remained conditional on disclosure, and an order’s silence does not prove none was obtained.

The orders recorded fee frameworks rather than the amounts eventually charged. High Court rules appeared in 51.1% of orders; 22.8% left the fee to the arbitrator, and 15.9% recorded no basis. Interviewees described the difficulty of negotiating remuneration with a tribunal already chosen to decide one’s dispute.

Repetition was substantial, with 186 of 387 named appointments going to someone the same court had already appointed. The ten most frequently appointed individuals received 31.6% of Bombay’s appointments and 34.3% of Madras’s. In Madras, most repeats came from the same judge appointing the same person, however they are counted.

What the record cannot show

These patterns do not establish favouritism. Repetition may reflect expertise, availability, reputation or a thin alternative pool. But professional knowledge and professional familiarity look alike from outside, and the orders rarely contain enough to tell them apart. The limit cuts both ways, since an arbitrator appointed twenty times on merit has no published record capable of showing it.

The study covers two High Courts and reported orders only. It cannot say how other courts appoint, or whether institutions do better. Its argument is about information. Identifying candidates, checking conflicts, confirming availability and settling fees need records that persist across cases. An institution with a continuing caseload can keep them; a rotating section 11 roster does not itself provide that continuity. That favours parties choosing an institution, appointment method and fee framework when they contract, with institutional performance still open to scrutiny.

Where courts continue to appoint, a narrower step is available. High Courts could publish periodic aggregate figures on appointment numbers, professional backgrounds, unique appointees, repeat appointments and concentration, all compiled from the orders themselves. That would constrain no individual appointment. It would make a public function observable, and let competing explanations be tested rather than asserted.

References

Animesh Bordoloi and Prashant Narang, How courts choose arbitrators: Evidence from the Bombay and Madras High Courts, TrustBridge Rule of Law Foundation Working Paper No. 25, 13 October 2026.

Prashant Narang and Vishnu Suresh, The Blind Spot in Indian Arbitration: Fees, Power, and Structural Oversights, The Leap Blog, 27 February 2025.

Animesh Bordoloi is Senior International Case Counsel at the Asian International Arbitration Centre, Malaysia. Prashant Narang is a researcher at the TrustBridge Rule of Law Foundation, New Delhi. We thank Renuka Sane for comments on earlier drafts.

Source: https://blog.theleapjournal.org/2026/11/how-courts-choose-arbitrators.html