Introduction
The interaction between doctrinal common law principles and constitutional provisions has long been a ground for judicial creativity in India. The question of whether a common law doctrine of discretionary forum selection can curtail the exercise of a constitutionally guaranteed writ jurisdiction is a product of such interactions. That question came before the Supreme Court of India in Baksish Ahmad v. Union of India, decided on 9 June 2026.1
The doctrine of forum non conveniens, meaning “inconvenient forum”, has a well-established pedigree in private international law. It permits a court that otherwise possesses jurisdiction to decline to exercise it if another forum is considered more appropriate for the resolution of the dispute. The doctrine serves the dual objective of judicial economy and the avoidance of unjust vexation of defendants. However, transposing a doctrine designed for the private law domain of choice of forum into the constitutional domain of writ jurisdiction raises acute theoretical difficulties.
The case arose from disciplinary proceedings against Baksish Ahmad, a constable in the Border Security Force (BSF). He was dismissed from service by the Commandant, and his statutory petition under Rule 28A of the BSF Rules, 1969 was rejected by the Inspector General, Frontier Headquarters, BSF, Jammu. He challenged both orders before the Delhi High Court. A Division Bench of the Delhi High Court declined to entertain the petition, invoking the doctrine of forum non conveniens on the ground that the relevant events had occurred in West Bengal and Jammu and Kashmir. The Supreme Court reversed this decision.
This paper proceeds in seven parts. The first part explains the forum non conveniens doctrine within its historical and theoretical context. The second examines the constitutional architecture of Article 226 and its dual-track jurisdictional framework. The third analyses the Supreme Court’s reasoning in Baksish Ahmad in depth. The fourth offers a critical analysis identifying both the judgment’s strengths and its doctrinal lacunae. The final part proposes a framework for future adjudication and concludes.
The doctrine of forum non conveniens: historical development
A. Common law and private international law origins
The doctrine of forum non conveniens has its roots in Scottish law, where courts exercised a discretion to sist (stay) proceedings where a foreign court was more appropriate. The doctrine was authoritatively established in the English decision in Spiliada Maritime Corporation v. Cansulex Ltd.2 In that landmark judgment, Lord Goff of Chieveley held that a stay should be granted where the court is satisfied that there is some other available forum, having competent jurisdiction, in which the case may be tried more suitably for the interests of all the parties and the ends of justice. This is a two-stage test: the court first identifies a prima facie more appropriate forum, and then considers whether justice nonetheless requires trial in the chosen court. It remains the foundational statement of the doctrine in common law jurisdictions.3
In the United States, the doctrine was authoritatively formulated in Gulf Oil Corp. v. Gilbert,4 where the Supreme Court identified a series of “private interest” and “public interest” factors to guide the discretionary calculus. These include the relative ease of access to sources of proof, the availability of compulsory process for obtaining the attendance of witnesses, the local interest in adjudicating disputes with local connections, and court congestion. The American formulation has been influential in the development of the doctrine in Indian commercial and arbitration law.
B. Reception in Indian civil and service law
Indian courts have recognised and applied the forum non conveniens principle primarily in the context of suits under the Code of Civil Procedure, 1908 and in writ petitions under Article 226(2). In Navinchandra N. Majithia v. State of Maharashtra,5 the Supreme Court cautioned High Courts against entertaining writ petitions where the cause of action’s connection with the forum was tenuous or contrived. The Court emphasised that a High Court should not transgress into the jurisdiction of another High Court merely because an insignificant event connected with the cause of action occurred within its territorial limits.
The treatment of Article 226(2) jurisdiction and its limits was subsequently expanded in Kusum Ingots & Alloys Ltd. v. Union of India,6 where the Supreme Court examined the scope of “cause of action” under clause (2). The Court distinguished between clause (1) and clause (2) jurisdiction and observed that even a small fraction of a cause of action accruing within the court’s limits would suffice to attract clause (2) jurisdiction, but that courts should guard against manufactured or peripheral nexuses. However, neither Navinchandra N. Majithia nor Kusum Ingots was a case involving clause (1) jurisdiction. The applicability of forum non conveniens to clause (1) cases remained undecided until Baksish Ahmad.
These applications all operate in the space of optional or discretionary jurisdiction that courts may appropriately decline when a more convenient alternative exists.
The constitutional architecture of Article 226: the jurisdictional framework
A. The structure of Article 226
Article 226 of the Constitution of India confers on every High Court the power to issue directions, orders, or writs, including writs in the nature of habeas corpus, mandamus, prohibition, quo warranto and certiorari, to any person or authority, including in appropriate cases any Government, within those territories.7 The provision is structured around two distinct jurisdictional triggers.
Clause (1) of Article 226 grounds jurisdiction in the situs of the respondent authority: the person or body against whom the writ is sought must be situated within the territorial limits of the High Court. Clause (2), inserted by the Constitution (Forty-Second Amendment) Act, 1976 and retained after the Constitution (Forty-Fourth Amendment) Act, 1978, extends jurisdiction to cases where the cause of action arises wholly or in part within the territorial limits of the High Court, irrespective of where the respondent authority is located. This was a significant expansion of the original constitutional design.
B. The significance of the clause (1) and clause (2) distinction
The duality of Article 226 creates two independent, co-existing sources of jurisdiction. A litigant may invoke a High Court’s jurisdiction under clause (1) if the authority against whom the writ is sought has its seat within that court’s territorial limits. Alternatively, a litigant may invoke jurisdiction under clause (2) if the cause of action arose within those limits.
This distinction is not merely textual; it carries profound normative implications. Clause (1) jurisdiction is, in an important sense, respondent-centric: it confers a right of access to the court proximate to the decision-making authority. Clause (2) jurisdiction is petitioner-centric or dispute-centric: it brings the forum to the cause of action. When a litigant invokes clause (1) jurisdiction, the choice reflects a deliberate invocation of the constitutional grant. The respondent authority is necessarily within the territorial ambit of the court, and the records and documents pertaining to the impugned order will be in the custody of the respondent and thus accessible to the chosen court. The theoretical justification for forum non conveniens, that the chosen court has no genuine connection to the dispute, is thereby materially weakened.
C. The constitutional status of the writ remedy
It is well established in Indian constitutional law that the power of judicial review vested in the High Courts under Articles 226 and 227 constitutes an integral and essential feature of the Constitution. In L. Chandra Kumar v. Union of India,8 a seven-judge Constitution Bench of the Supreme Court held that the jurisdiction conferred upon High Courts under Articles 226 and 227 is a part of the inviolable basic structure of the Constitution and cannot be ousted even by a constitutional amendment. The impugned provisions of the Administrative Tribunals Act, 1985 that attempted to exclude High Court jurisdiction were accordingly read down.
This constitutional elevation of the writ jurisdiction is not without consequence for the forum non conveniens analysis. A power that is part of the basic structure, and whose exercise constitutes the primary constitutional mechanism for the vindication of citizens’ rights against state action, cannot be casually declined on grounds of procedural convenience. The application of a common law discretionary doctrine to abrogate or impede the exercise of a constitutional right requires the strongest justification.
The judgment in Baksish Ahmad: an analytical study
A. Factual matrix and proceedings below
Baksish Ahmad was enrolled in the Border Security Force in December 2010 and was serving with the 44th Battalion in West Bengal.9 In 2022, allegations surfaced that he had contracted a second marriage while his first marriage was still subsisting, and without obtaining the requisite permission from the competent authority. A Staff Court of Inquiry conducted by the BSF found that Ahmad had entered into a second marriage. A show-cause notice was served on him at Narayanpur in West Bengal’s Malda district. Upon his failure to submit a reply within the prescribed period, the Commandant dismissed him from service in 2022 without any pensionary benefits.10
His statutory petition under Rule 28A of the Border Security Force Rules, 1969, though addressed to the Director General, BSF, was placed before the Inspector General, Frontier Headquarters, BSF, Jammu. After condonation of delay, the petition was rejected on merits on 22 December 2023. Ahmad then challenged both the dismissal order and the rejection of his statutory petition before the Delhi High Court under Article 226 of the Constitution, contending that the court had jurisdiction because both the Director General, BSF and the Ministry of Home Affairs are headquartered in Delhi.11
The Division Bench of the Delhi High Court declined to entertain the petition. It observed that merely because the office of the Director General, BSF and the Ministry of Home Affairs is situated at Delhi, that would not make the Delhi High Court a forum conveniens.12 It held that the proper forums would be either the Calcutta High Court, given that the show-cause notice and dismissal order were issued in West Bengal, or the High Court of Jammu and Kashmir and Ladakh, given that the statutory petition was rejected there. Liberty was granted to approach the appropriate High Court.
B. The Supreme Court’s reasoning
The Supreme Court, in a judgment authored by Justice Dipankar Datta, reversed the Delhi High Court on two related grounds.
First, relying on its earlier decision in Abrar Ali v. CISF,13 the Court held that the Delhi High Court possessed territorial jurisdiction under Article 226(1). The Court observed that in the case of any member of the Central Armed Police Forces (CAPF), including the BSF, who is aggrieved by an administrative order of termination of service, notwithstanding that the cause of action arose outside Delhi, the Delhi High Court would have territorial jurisdiction in light of the situs of the office of the Union of India and the Director General, BSF, under clause (1) of Article 226.14
Second, and more fundamentally, the Court addressed the applicability of forum non conveniens to writ proceedings under Article 226(1). The Court held that the doctrine of forum non conveniens had been misapplied by the Division Bench in the context of writ jurisdiction referable to Article 226 of the Constitution; that Article permits the filing of a writ petition as per the situs of the office of the respondent under clause (1) and the cause of action under clause (2), which gives the right of action; and that where the question of pursuing a constitutional remedy is involved and the invocation of writ jurisdiction is traceable to clause (1) of Article 226, the doctrine of forum non conveniens may rarely apply.15
The Court further reasoned that in proceedings for a writ of certiorari, the records of the impugned proceedings would ordinarily be in the possession of the respondent authority and could readily be produced before the court. Accordingly, directing the litigant away from a forum chosen on the basis of the respondent’s location would be self-defeating: a suitor having himself chosen the forum convenient to the respondents, the application of the doctrine of forum non conveniens could be self-defeating and likely to deny access to justice rather than advance it.16
The Court’s reasoning in Baksish Ahmad implicitly reconceives the doctrine: when clause (1) jurisdiction is invoked, the chosen forum is already the forum most closely connected to the respondent authority. Since it is the respondent’s records and institutional capacity that are at issue in certiorari proceedings, the chosen forum is by definition appropriate for the administration of those proceedings. The theoretical foundation of forum non conveniens thereby evaporates in the Article 226(1) context.
Pre-existing jurisprudence and the interpretive landscape
A. The Navinchandra line and Article 226(2)
Pre-Baksish Ahmad jurisprudence on forum non conveniens in writ proceedings was developed primarily in the Article 226(2) context. In Navinchandra N. Majithia,17 the Supreme Court, while affirming that even a small part of the cause of action arising within the court’s limits suffices to attract clause (2) jurisdiction, cautioned that High Courts must not countenance fictitious causes of action or trivial jurisdictional pegs. The relevant test, the Court held, was whether the asserted factual nexus was genuine and substantial, and not a mere pretext.
In Kusum Ingots,18 the Court went further, holding that where a parliamentary legislation applies to the entire territory of India, the situs of Parliament, that is, Delhi, does not by itself confer jurisdiction on the Delhi High Court under Article 226. The passing of legislation is not a cause of action in the Article 226(2) sense. These decisions collectively establish a rigorous approach to clause (2) jurisdiction claims, guarding against manufactured nexuses.
B. The CAPF and service law precedent
In the service law domain, particularly concerning central paramilitary forces, the trend has been towards recognising Delhi High Court jurisdiction under Article 226(1). The logic is straightforward: the Union of India and the Director General of a CAPF are the apex authorities whose orders govern service conditions, and their offices are in Delhi. In Abrar Ali v. CISF,19 the Delhi High Court had already affirmed this principle. Baksish Ahmad extends and consolidates this line of authority, specifically by addressing what happens when a High Court attempts to override the clause (1) nexus by invoking forum non conveniens.
What Baksish Ahmad does is to establish a differentiated analysis: the threshold for invoking forum non conveniens must be higher for clause (1) cases than for clause (2) cases. This differentiation is constitutionally principled, given that clause (1) jurisdiction is anchored in the location of the respondent, an anchor that directly bears on the administration of certiorari proceedings.
Critical analysis: strengths, ambiguities and gaps
A. The strengths of the judgment
The Supreme Court’s decision in Baksish Ahmad is to be welcomed on several grounds. First, it correctly identifies the constitutional dimension of the problem. By situating the analysis within the specific structure of Article 226’s two jurisdictional clauses, the Court avoids the error of treating writ jurisdiction as functionally analogous to ordinary civil jurisdiction under the Code of Civil Procedure. The constitutional character of the writ remedy, reinforced by L. Chandra Kumar,20 demands a higher threshold for any doctrine that impedes its exercise.
Second, the Court’s observation about certiorari proceedings is doctrinally acute. The writ of certiorari is, by its very nature, a record-calling writ.21 It requires the court to examine whether the impugned order was made in accordance with law, on the basis of the record before the inferior tribunal or authority. Those records are in the possession of the respondent. A forum chosen on the basis of the respondent’s location is therefore inherently well-positioned to examine those records. The premise of forum non conveniens, that a distant forum lacks access to relevant evidence and witnesses, loses much of its force in this context.
Third, the judgment advances the constitutional value of access to justice embedded in Articles 14, 21, and 39A of the Constitution.22 For a BSF constable in disciplinary proceedings, forcing repeated forum-shopping across multiple High Courts, such as those at Calcutta and Jammu and Kashmir, would impose a severe burden. The Court correctly recognises that the doctrine’s application in this context would deny rather than advance justice.
B. Doctrinal ambiguities left unresolved
Notwithstanding these strengths, the judgment leaves several important questions unresolved. The Court’s formulation that the doctrine “may rarely apply” is frustratingly imprecise. When does “rarely” become “never”? When does it become “sometimes”? The judgment provides no structured test for determining when, exceptionally, forum non conveniens might justify declining clause (1) jurisdiction. The absence of such a test creates uncertainty for High Courts and litigants alike.
For example, consider a case where the Union of India is a respondent and its offices are present in Delhi by definition, but the entire factual matrix concerns events in a single remote state, all material witnesses are located there, the petitioner resides there, and the cause of action arose entirely there. Should the Delhi High Court necessarily entertain the petition merely because clause (1) jurisdiction technically attaches? The Court’s “rarely” formulation presumably contemplates that such exceptional cases might warrant forum non conveniens, but the framework for identifying them is left undeveloped.
A second ambiguity concerns the relationship between forum non conveniens and the principle of self-imposed restraint that the Supreme Court has applied in certain categories of writ petitions. The Court has occasionally suggested that High Courts should exercise restraint when entertaining petitions concerning matters that are quintessentially local in character. Whether this form of judicial self-restraint is conceptually distinct from forum non conveniens, or merely a variant of it, remains unclear after Baksish Ahmad.
Third, the judgment does not address the question of pendency and judicial congestion. The Delhi High Court has one of the highest caseloads of any High Court in India.23 If the situs of government offices invariably confers jurisdiction and forum non conveniens cannot be invoked, the Delhi High Court will absorb an enormous volume of service disputes from across India. The Court’s silence on this systemic concern is understandable given the constitutional constraints, but it is a significant omission.
C. The problem of strategic party joinder
A deeper structural problem concerns the nature of the respondent whose situs triggers clause (1) jurisdiction. The Court’s reasoning proceeds on the basis that the Union of India and the Director General, BSF are necessary parties. But the concept of necessity requires closer examination. In a service law dispute concerning a constable in a specific frontier battalion headquartered outside Delhi, is the Union of India truly a necessary party, or is it a convenient respondent joined precisely to attract Delhi High Court jurisdiction?
Indian courts have, in other contexts, been alive to the possibility of collusive or strategic joinder of parties to manufacture jurisdiction. The Aligarh Muslim University case, noted in Kusum Ingots,24 offers a cautionary example of a High Court exercising jurisdiction on a fragile jurisdictional pretext. The judgment in Baksish Ahmad does not engage with this concern, which may, in less clear-cut cases, open the door to precisely the forum-shopping the doctrine was designed to prevent.
This is not to suggest that Ahmad himself was engaged in illegitimate forum-shopping. The Delhi High Court’s jurisdiction was supported by sound precedent, and the Union of India is a proper respondent in any challenge to BSF disciplinary proceedings. The point is theoretical: the Court’s broad formulation, without qualification, could be misread in future cases as validating genuinely manipulative jurisdiction-seeking.
D. The access to justice dimension: a deeper analysis
Access to justice is both an aspect of the rule of law and a constitutional value implied in Articles 14, 21, and 39A.25 However, access to justice is not a unidirectional principle. It operates as much to protect petitioners from unreasonable jurisdictional burdens as it does to protect respondents from vexatious litigation in distant forums.
A BSF constable filing in Delhi rather than before the Calcutta High Court may find the constitutional remedy more accessible, but the respondent witnesses and officials who must appear and defend may face equal hardship if required to travel to Delhi for proceedings arising from events in West Bengal. The Court’s formulation, in prioritising the petitioner’s choice of forum under clause (1), implicitly makes a policy choice about whose access to justice is weightier. For constitutional writ proceedings, this choice may well be correct, but the reasoning would benefit from an explicit acknowledgment of this normative choice, rather than treating access to justice as pointing unambiguously in one direction.
A proposed framework and conclusion
A. A structured test for forum non conveniens in writ proceedings
The judgment in Baksish Ahmad establishes that forum non conveniens “may rarely apply” to Article 226(1) proceedings, but leaves the standard indeterminate. Drawing on the reasoning of the Court and the broader jurisprudence, this paper proposes the following structured framework for High Courts.
First, the High Court should identify whether jurisdiction is being invoked under clause (1) or clause (2) of Article 226. For clause (1) cases, the presumption should be strongly in favour of exercising jurisdiction, and the threshold for declination should be correspondingly high. For clause (2) cases, greater judicial discretion may be warranted, consistent with Navinchandra N. Majithia.
Second, the court should assess whether the respondent authority against whom the writ is principally sought, and not merely a nominally joined party, genuinely has its office or headquarters within the territorial jurisdiction. If so, clause (1) jurisdiction is authentic and should ordinarily be exercised.
Third, even where clause (1) jurisdiction is authentically established, the court may decline to exercise it only upon a clear demonstration of both (a) the existence of another available and adequately equipped forum with competent jurisdiction, and (b) a concrete showing that exercising jurisdiction would result in genuine and substantial inconvenience to the administration of justice, and not merely theoretical inconvenience. The mere fact that the cause of action arose elsewhere is insufficient.
Fourth, the court should assess whether the nature of the writ sought, particularly whether it is a certiorari calling for records, reinforces the jurisdictional connection. Given the record-based nature of certiorari proceedings, clause (1) jurisdiction in such cases will almost always be appropriate.
Fifth, in all cases, the court should weigh the petitioner’s constitutional interest in effective access to justice against any concrete disadvantage to the administration of justice that would result from exercising jurisdiction, bearing in mind the constitutional character of the Article 226 remedy as articulated in L. Chandra Kumar.
B. Conclusion
Baksish Ahmad v. Union of India represents a significant clarification of the relationship between the forum non conveniens doctrine and writ jurisdiction under Article 226(1) of the Constitution of India. By holding that the doctrine may only rarely apply in this context, the Supreme Court has reinforced the constitutional character of the writ remedy and protected the access of ordinary citizens, including members of the armed and paramilitary forces, to effective judicial review of state action.
The judgment leaves certain questions unresolved: it provides no clear test for when the doctrine might “rarely” justify declination; it does not address the systemic implications for the Delhi High Court’s caseload; and it does not engage with the possibility of strategic party joinder to manufacture jurisdiction.
The Supreme Court has correctly identified the constitutional direction of travel. In doing so, it must remain attentive to the fundamental insight that constitutional remedies are not merely procedural conveniences; they are the institutional embodiments of the rule of law.
*****
Footnotes
1. Baksish Ahmad v. Union of India, 2026 INSC 630, (2026) SCC OnLine SC (India) (reported at 2026 LiveLaw (SC) 616).
2. Spiliada Maritime Corp. v. Cansulex Ltd., [1987] AC 460 (HL) (appeal taken from Eng.).
3. Id. at 476.
4. Gulf Oil Corp. v. Gilbert, 330 U.S. 501 (1947).
5. Navinchandra N. Majithia v. State of Maharashtra, (2000) 7 SCC 640 (India).
6. Kusum Ingots & Alloys Ltd. v. Union of India, (2004) 6 SCC 254 (India).
7. India Const. art. 226, cl. 1.
8. L. Chandra Kumar v. Union of India, (1997) 3 SCC 261 (India).
9. Baksish Ahmad, supra note 1.
10. Id.
11. Mariya Paliwala, BSF Jawan Wins Supreme Court Relief as Court Limits Use of Forum Non Conveniens in Writ Petitions, Juris Hour (June 9, 2026), https://www.jurishour.in/other-laws/bsf-jawan-forum-non-conveniens-writ-petitions/.
12. Baksish Ahmad, supra note 1 (quoting the observation of the Division Bench of the Delhi High Court, as reproduced in the judgment).
13. Abrar Ali v. Central Industrial Sec. Force, W.P. (C) No. 1241 of 2011 (Delhi High Court) (India).
14. Baksish Ahmad, supra note 1, para. 27.
15. Id. para. 27.
16. Id.
17. Navinchandra N. Majithia, supra note 5, para. 43.
18. Kusum Ingots, supra note 6, paras. 10-15.
19. Abrar Ali, supra note 13.
20. L. Chandra Kumar, supra note 8.
21. D.D. Basu, Commentary on the Constitution of India art. 226 (9th ed., LexisNexis 2015) vol. 8.
22. Anita Kushwaha v. Pushap Sudan, (2016) 8 SCC 509 (India).
23. Nat’l Court Mgmt. Sys. Comm., Report on Arrears and Backlogs 1-4 (Supreme Court of India 2016); DAKSH India, A Quantitative Study of Case Pendency in the Delhi High Court (DAKSH 2016).
24. Aligarh Muslim Univ. v. Vinay Engg. Enters. (P) Ltd., (1994) 4 SCC 710 (India).
25. Anita Kushwaha, supra note 22; see also India Const. arts. 14, 21, 39A.