The Armed Forces Tribunal Act 2007 created the forum that serving and retired members of the Army, the Navy and the Air Force had lacked for half a century: a statutory body that can decide a service grievance on its merits and can sit in appeal over a court martial. The Armed Forces Tribunal decides disputes about pay, pension, promotion and tenure, and it rehears the gravest product of military discipline, the finding and sentence of a court martial, with power to substitute its own conclusion. Yet the statute that built this forum withheld the two features that ordinarily secure the independence of an adjudicating body. The Tribunal is administered by the Ministry of Defence, the department that appears against the applicant in the great majority of cases before it. And an appeal from the Tribunal to the Supreme Court does not lie as of right in the ordinary case: it lies only where the Tribunal itself certifies that a point of law of general public importance is involved, or where the Supreme Court grants leave. A member of the armed forces who finishes this article should know what the Tribunal can decide, what it cannot, where its orders can be taken, and why its design remains contested.

Why a separate forum had to be created

The starting point is the structure of military justice itself. Persons subject to the Army Act 1950, the Navy Act 1957 and the Air Force Act 1950 live under a separate disciplinary code, tried by courts martial whose findings and sentences were, for most of the period after Independence, subject to confirmation and to petitions within the service hierarchy rather than to appeal before an independent judicial body. The Judge Advocate General’s branch reviewed court martial proceedings inside the services, and the Central Government and the Chiefs of Staff held powers of pardon, remission and review under the service Acts, but none of that machinery amounted to an appeal heard by a body outside the chain of command. A person convicted by a court martial who believed the finding was wrong in law or unsupported by evidence had, in substance, one external remedy: a petition to a High Court under Article 226 of the Constitution, a jurisdiction that corrects jurisdictional error and denial of natural justice but does not reappreciate evidence as an appeal court does.

The Supreme Court identified the gap directly in Lt Col Prithi Pal Singh Bedi v Union of India, decided in 1982 and reported at (1982) 3 SCC 140. The Court upheld the constitutional validity of the court martial system, accepting that Article 33 of the Constitution permits Parliament to restrict the fundamental rights of armed forces personnel in the interest of discipline. But the judgment recorded that persons subject to the Army Act had no right of appeal against a court martial, that the absence of an appellate forum was a defect in the system, and that the matter called for legislative attention. The observation mattered because it came from the Court that had just sustained the system: the discipline code was valid, but its lack of an appeal was an institutional shortcoming that only Parliament could cure.

Parliament did not act for more than two decades. In the interval the service grievance load moved to the High Courts, which entertained writ petitions on seniority, promotion, pension and dismissal from personnel across the country, deciding military service questions without the assistance of members who understood service organisation. The Law Commission of India took up the service Acts in its 169th Report of 1999, on the amendment of the Army, Navy and Air Force Acts, and recommended an Armed Forces Appellate Tribunal to entertain appeals against the findings, orders and sentences of courts martial. The Commission proposed a body presided over by a retired judge of the Supreme Court or a retired Chief Justice of a High Court, sitting with a senior retired officer and a retired Judge Advocate General, and it recorded the views of serving and retired personnel who asked that the forum also take service disputes, since hundreds of writ petitions on conditions of service were being filed in the High Courts each year. The statute that eventually emerged adopted that wider design: not an appellate tribunal alone, but a tribunal of first instance for service matters and an appellate tribunal for courts martial combined.

The Armed Forces Tribunal Bill was introduced in Parliament in 2005 by the Defence Minister, Pranab Mukherjee, and examined by the Parliamentary Standing Committee on Defence, which supported a dedicated forum while pressing for safeguards on the exclusions from jurisdiction. The Bill was passed by the Rajya Sabha on 3 December 2007 and by the Lok Sabha on 6 December 2007, and received the assent of the President on 25 December 2007. The wider story of how India built specialised forums alongside the ordinary courts, and where this Tribunal sits within the court system as a whole, is traced in the series guide to the laws governing the judicial system in India.

When did the Armed Forces Tribunal begin functioning?

The Armed Forces Tribunal Act 2007 is Act 55 of 2007. It received Presidential assent on 25 December 2007 and was brought into force on 15 June 2008. The Tribunal itself was inaugurated on 8 August 2009, when its Principal Bench at New Delhi began work, with regional benches following at the places notified by the Central Government.

The interval between enactment and inauguration is itself instructive. A tribunal statute does not produce a functioning forum on the day of assent: members must be selected, benches constituted, premises and registries established and procedure rules framed. The Armed Forces Tribunal (Procedure) Rules 2008 supplied the procedural skeleton, and the establishment notification under Section 4 of the Act fixed the appointed day from which the Tribunal would exercise its jurisdiction. Regional benches were constituted at Chandigarh, Lucknow, Kolkata, Guwahati, Chennai, Kochi, Mumbai and Jaipur, with a bench for Srinagar added later, so that personnel and pensioners would not have to carry every grievance to Delhi. The geography was a deliberate part of the design, because the population the Tribunal serves is dispersed across cantonments, naval stations, air bases and the home districts to which personnel retire.

The statute in outline

The Act is organised around two jurisdictional grants and one appellate channel. Chapter III contains the grants. Section 14 gives the Tribunal jurisdiction over service matters of persons subject to the three service Acts. Section 15 gives it appellate jurisdiction over courts martial. Around those grants the Act builds the ordinary machinery of a tribunal: establishment and benches in Chapter II, composition and qualifications in Sections 5 and 6, appointment by the President after consultation with the Chief Justice of India under Section 7, a term of four years under Section 8, procedure and powers in Chapter IV, execution of orders under Section 29, and the appeal to the Supreme Court in Chapter V, in Sections 30 and 31. Section 2 applies the Act to all persons subject to the Army Act 1950, the Navy Act 1957 and the Air Force Act 1950, and extends it, for service matters, to retired personnel and to their dependants, heirs and successors. That extension is what makes the Tribunal, in practice, a pensioners’ forum as much as a serving personnel’s forum.

Section 33 states the exclusion that frames the whole scheme: no court other than the Supreme Court or a High Court exercising jurisdiction under Articles 226 and 227 of the Constitution may exercise jurisdiction in relation to matters within the Tribunal’s remit. The reservation of the writ jurisdiction inside the exclusion clause is significant, and the case law on it is addressed in a later section of this article. Section 34 transferred to the Tribunal the service and court martial matters pending before the High Courts and other courts on the appointed day, so the new forum inherited a working docket from its first day.

The Act belongs to the larger movement of tribunalisation that has moved whole categories of disputes out of the ordinary courts into specialised bodies, a movement whose company law instance, examined elsewhere in this series, created the National Company Law Tribunal and its appellate tribunal in the article on the NCLT and NCLAT. The Armed Forces Tribunal differs from most members of that family in one respect that shapes everything else about it: the disputes it hears arise inside a disciplined force, and many of them carry consequences for command that have no parallel in tax, company or service law generally. That is the premise on which its defenders rest its design, and the premise its critics say has been stretched further than it can bear.

Service matters: the jurisdiction under Section 14

Section 14(1) provides that, save as otherwise expressly provided, the Tribunal shall exercise, on and from the appointed day, all the jurisdiction, powers and authority exercisable in relation to service matters of persons subject to the three service Acts. Section 14(2) entitles a person aggrieved by an order pertaining to a service matter to apply to the Tribunal. The operative definition is in Section 3(o), and it repays close reading, because the Tribunal’s reach is fixed by the definition’s inclusions and by its exclusions, and most disappointed applicants are defeated by the exclusions rather than by the merits of their case.

The definition opens with a general phrase: service matters mean all matters relating to the conditions of service of the persons concerned. It then includes, in terms, remuneration including allowances, pension and other retirement benefits; and tenure, including commission, appointment, enrolment, probation, confirmation, seniority, training, promotion, reversion, premature retirement, superannuation, termination of service and penal deductions. It includes summary disposal and trials where the punishment of dismissal is awarded. It then adds a residuary inclusion of any other matter whatsoever, before turning to the exclusions. The architecture is that of a wide grant deliberately cut back: everything about conditions of service is within the Tribunal unless the definition or the Act takes it out.

The practical content of the grant is best understood through the categories that fill the Tribunal’s lists. Pay and allowances disputes concern fixation on promotion, the pay band attached to a rank, arrears following a pay commission’s implementation, and recovery of alleged overpayments, including penal deductions from pay. Promotion and seniority disputes concern selection boards, confidential reports, the criteria recorded in promotion policy, and the relative seniority from which promotion prospects flow. Premature retirement and compulsory retirement orders are challenged on the grounds open in service law: that the order rests on materials that do not support it, that it is punitive in substance while wearing a non punitive form, or that the procedure the service’s own rules required was not followed. Termination and dismissal arising from disciplinary proceedings other than courts martial, including summary disposals where dismissal is awarded, come within the grant. Pension disputes, addressed in their own section of this article, form the largest single stream.

Section 14(4) and the provisions around it equip the Tribunal to do more than declare the law. On an application the Tribunal may pass such orders as it thinks fit, and its decisions on questions of law and fact are both within its competence, since Section 14(5) provides that the Tribunal shall decide both questions of law and questions of fact raised before it. This is the feature that separates the Tribunal from the writ court it displaced. A High Court under Article 226 asks whether the decision maker acted lawfully, within jurisdiction, fairly and rationally. The Tribunal under Section 14 asks the prior question as well: whether the order was right. For a personnel system in which the operative facts sit inside confidential reports, board proceedings and medical categorisation, that merits jurisdiction is the substance of the reform the Law Commission recommended and the personnel who met the Commission demanded.

Two limits on Section 14 deserve emphasis because they are often missed. The first is that the Tribunal’s service jurisdiction runs with the person and the service Acts: it is a forum for those subject to the Army, Navy and Air Force Acts and, for service matters, for those who have retired from that subjection. It is not a general military grievance forum for every uniformed force in India, a point developed in the questions at the end of this article. The second is that the definition is exhaustive of the route: a matter that falls outside Section 3(o) cannot be brought inside Section 14 by describing it as a service grievance in general terms, however closely it touches the applicant’s service life. The exclusions are therefore the first thing an adviser must check.

The exclusions: what Section 3(o) keeps out

The definition excludes four categories. It excludes matters relating to orders issued under Section 18 of the Army Act 1950, Section 15(1) of the Navy Act 1957 and Section 18 of the Air Force Act 1950, the provisions under which the President’s pleasure operates in relation to tenure, so dismissals resting on the pleasure doctrine in that specific statutory form are kept outside. It excludes transfers and postings, including the change of place or unit on posting, whether individually or as part of a unit, formation or ship. It excludes leave of any kind. And it excludes summary court martial, except where the punishment is of dismissal or imprisonment for more than three months.

Each exclusion has a rationale stated in the legislative history, and each has been criticised. Transfers and postings were excluded on the ground that the movement of personnel is an operational and administrative function of command, and that subjecting each posting to litigation would burden units and impair operational preparedness. The Parliamentary Standing Committee on Defence accepted the exclusion while recommending that the services frame clear transfer policies, so that the exclusion would not become a shelter for arbitrary movement. Leave was excluded on a similar apprehension, pressed by the Ministry of Defence during the Bill’s examination, that leave disputes arising in every unit would flood the forum; critics of the exclusion pointed out that no empirical study supported the apprehension, and that leave can carry pay and service consequences as real as any other condition of service. The summary court martial exclusion reflects the summary character of that forum: a summary court martial is a swift disciplinary proceeding conducted by a commanding officer, and Parliament drew the line at the point where its punishments become grave, dismissal or imprisonment beyond three months, leaving lesser punishments to the internal remedies of the service Acts.

The exclusions generate a characteristic pattern in the Tribunal’s practice. An applicant whose real complaint is a posting will frame the case as a challenge to the order that followed the posting, or to a consequential entry in the service record, and the Tribunal must decide whether the substance of the grievance is the excluded matter or an included one. The case law has treated the exclusion by substance rather than by label: a challenge to a premature retirement order does not become a posting challenge because a posting preceded it, and a challenge dressed as one to a consequential order fails where the operative decision is the posting itself. For applicants the lesson is the one the brief for this article states as the payload of the artifact that follows: the exclusions are what most applicants need to know first, and they are rarely set out in one place.

Category of grievance	Within the Tribunal’s jurisdiction	Provision	Where the matter goes if outside
Pay and allowances	Yes	Sections 14 and 3(o)(i)	Not applicable
Promotion and seniority	Yes	Sections 14 and 3(o)(ii)	Not applicable
Premature retirement	Yes	Sections 14 and 3(o)(ii)	Not applicable
Disability pension	Yes	Sections 14 and 3(o)(i)	Not applicable
A court martial conviction	Yes, by appeal	Section 15	Not applicable
A transfer or posting	No	Excluded by Section 3(o)	Internal service remedies, and a writ petition in a narrow class of case
Leave of any kind	No	Excluded by Section 3(o)	Internal service remedies
A summary court martial	Only where the punishment is dismissal or imprisonment for more than three months	Section 3(o)(iv) of the exclusions	Internal remedies under the service Acts for lesser punishments

The table states the map in its simplest form. The rows marked as within jurisdiction are the Tribunal’s daily work. The excluded rows do not leave the applicant without any remedy, since the internal remedies of the service Acts and, in a narrow class of case, the writ jurisdiction remain, but they leave the applicant without this Tribunal, and the distinction between a grievance the Tribunal can decide on the merits and one it must turn away at the threshold is the distinction on which competent advice in this field begins.

Can the Tribunal release a person held in military custody?

Section 15(3) gives the Tribunal power to grant bail to a person accused of an offence who is in military custody, with or without conditions. Bail is refused where there appears reasonable ground for believing that the accused has been guilty of an offence punishable with death or imprisonment for life.

The bail power sits inside the appellate chapter because its natural moment is the period after a court martial, when a convicted person may remain in custody while an appeal is prepared, but the text is not confined to convicted persons: it speaks of a person accused of an offence and in military custody. Military custody is defined by Section 3(i) as arrest or confinement according to the usages of the service, including naval and air force custody. The power matters because the service Acts’ own custody regime is command based, and an external judicial authority able to order release on conditions introduces into military custody a check that the internal system did not supply. In practice the Tribunal applies recognisable bail considerations, the gravity of the charge, the stage the proceedings have reached, the risk of interference with witnesses who may be subordinate to or senior to the accused within the same unit, and the period of custody already undergone, adapted to a setting in which the accused and the witnesses wear the same uniform.

Appeals from courts martial under Section 15

Section 15(1) gives the Tribunal, from the appointed day, all the jurisdiction, powers and authority exercisable under the Act in relation to appeal against any order, decision, finding or sentence passed by a court martial, or any matter connected with or incidental to it. Section 15(2) entitles any person aggrieved by an order, decision, finding or sentence of a court martial to prefer an appeal in the prescribed form, manner and time. This is the provision that answers the defect identified in Prithi Pal Singh Bedi: for the first time, a court martial in India is subject to a true appeal, heard by a body outside the service, on grounds that include the sustainability of the finding itself.

Section 15(4) states when the Tribunal shall allow an appeal against conviction. It shall allow the appeal where the finding of the court martial is legally not sustainable for any reason, where the finding involves a wrong decision on a question of law, or where there was a material irregularity in the course of the trial resulting in a miscarriage of justice. In any other case it may dismiss the appeal where it considers that no miscarriage of justice is likely to be caused or has actually resulted to the appellant, and a dismissal must record reasons in writing. The structure mirrors the appellate discipline of the criminal courts while adapting it to a trial that is conducted by officers rather than by a judge: the Tribunal does not retry the case as a court martial would, but it may set aside a finding that the evidence cannot sustain, correct the law the court martial applied, and intervene where irregularity has produced injustice.

Section 15(6) arms the Tribunal with the consequential powers an appeal court needs. It may substitute for the finding of the court martial a finding of guilty of any other offence of which the offender could lawfully have been found guilty by the court martial on the evidence, and pass a sentence accordingly. It may mitigate, remit or commute a sentence, enhance it only within the limits and safeguards the provision prescribes, order a retrial by a court martial on the charges or on such of them as it specifies, and pass such further orders as the case requires, including orders on the custody and release of the appellant. The presence of an enhancement power, hemmed in by the requirement that the appellant be heard, is a marker of how complete Parliament intended the appellate substitution to be: the Tribunal steps into the place of the confirming and reviewing authorities of the old system and exceeds them, because none of those authorities could reappreciate the finding as an appeal court does.

Appeals under Section 15 also interact with the remedies the service Acts themselves preserve. The statutory petitions to the Central Government and the service chiefs, including the petition under Section 164(2) of the Army Act against a court martial finding or sentence, were not abolished by the Act, and the confirmation requirements of the service Acts continue to operate on the sentences to which they apply. The Tribunal’s appeal is the judicial layer placed over that administrative structure. An appellant’s advisers must therefore sequence the remedies: confirmation, the statutory petition where it is pursued, and the appeal to the Tribunal, each with its own limitation discipline under the Act and the Procedure Rules. The Tribunal has power to condone delay in the presentation of applications and appeals where sufficient cause is shown, and its practice on condonation in pension and court martial matters has generally recognised the circumstances of service life, distance from benches, dependence on service records that the respondent department holds, and the age of retired applicants, without treating delay as irrelevant.

Composition: judicial members and administrative members

Section 5 constitutes the Tribunal with a Chairperson and such number of Judicial and Administrative Members as the Central Government may notify, exercising jurisdiction through benches. A bench ordinarily pairs a Judicial Member with an Administrative Member, so that the two forms of knowledge the forum exists to combine are present in each hearing. Section 6 prescribes the qualifications. The Chairperson must be a retired Judge of the Supreme Court or a retired Chief Justice of a High Court. A Judicial Member must be or have been a Judge of a High Court. An Administrative Member must have held the rank of Major General or above, or the equivalent rank in the Navy or the Air Force, for a total period of at least three years, and must have served for not less than one year as Judge Advocate General in one of the services, not below the rank of Major General, Commodore or Air Commodore respectively. A serving person appointed as an Administrative Member must retire from service before assuming the appointment.

The design is the Law Commission’s 169th Report translated into statute with modifications. The Commission proposed a three member tribunal: a presiding judicial figure, a senior retired officer, and a retired Judge Advocate General. The Act separates the Judge Advocate General element into the qualification for the administrative seat, requiring every Administrative Member to carry both senior command rank and a year in the service’s legal branch leadership. The result is a bench on which the judicial member supplies adjudicative method and the administrative member supplies the institutional knowledge of how a unit, a selection board or a medical board actually works. When a case turns on what a confidential report’s grading means within a promotion system, or on how a medical board’s categorisation is arrived at, the administrative member’s experience is not decoration; it is the expertise that justified a tribunal over a court.

Section 7 provides for appointment of the Chairperson and Members by the President, and no appointment may be made except after consultation with the Chief Justice of India. Section 8 fixes a term of four years, with eligibility for reappointment, and prescribes age ceilings: a Chairperson who has been a Judge of the Supreme Court may hold office to the age of seventy years, a Chairperson who has been a Chief Justice of a High Court to sixty five, and other Members to sixty five. Section 9 governs resignation and removal, with removal on proved misbehaviour or incapacity following an inquiry by a Judge of the Supreme Court. Sections 10 to 13 deal with salaries and conditions, the financial and administrative powers of the Chairperson over the Tribunal’s own functioning, and the Tribunal’s staff. The conditions of service of members, which the Finance Act 2017 and later the Tribunals Reforms Act 2021 sought to reset across the tribunal system, are the point at which this Tribunal’s story joins the general tribunals litigation, and they are taken up in the section on reform later in this article.

The composition has never been the centre of the criticism of this Tribunal, and the brief for this article records why: the criticism is structural rather than about its members. A Chairperson drawn from the retired Supreme Court and Chief Justices, judicial members drawn from the High Courts, and administrative members who combine flag rank with Judge Advocate General service make a strong bench on paper, and the reported work of the benches, including orders against the Ministry of Defence in pension and promotion matters in large numbers, does not suggest a forum captured by the respondent department at the level of adjudication. The structural questions lie elsewhere: in who administers the institution, and in where its orders can be taken on appeal.

What is the difference between an application and an appeal before the Tribunal?

An application under Section 14 is an original proceeding in which the Tribunal decides a service matter itself, on evidence led before it. An appeal under Section 15 is an appellate proceeding against an order, finding or sentence of a court martial, in which the Tribunal reviews the court martial record on the grounds Section 15(4) prescribes.

The distinction controls the shape of the case. In an application, evidence is led before the Tribunal, affidavits are exchanged, and the Tribunal finds facts. In a court martial appeal, the record is the proceedings of the court martial, and the Tribunal’s inquiry is whether that record sustains the finding in law and on the evidence, whether the law was correctly applied, and whether any material irregularity caused a miscarriage of justice. Confusing the two produces the commonest drafting error in this jurisdiction, an appeal argued as though the Tribunal could conduct a fresh trial, or an application argued as though the Tribunal were confined to judicial review of the order under challenge. Neither is correct, and the statute’s two chapters keep the two functions apart with care.

The Tribunal’s powers: evidence, contempt and procedure

A tribunal that decides facts needs the machinery to compel them. Section 17 vests the Tribunal, for the purpose of discharging its functions, with the powers of a civil court under the Code of Civil Procedure 1908 in specified respects: summoning and enforcing the attendance of witnesses and examining them on oath, requiring the discovery and production of documents, receiving evidence on affidavit, requisitioning public records, issuing commissions for the examination of witnesses or documents, reviewing its decisions, dismissing an application for default or deciding it ex parte, and setting aside such orders. The list is the standard equipment of Indian tribunals, and in this jurisdiction it has a particular edge, because the decisive records, service documents, confidential reports, board proceedings, medical board findings, are held by the respondent department. A power to requisition the record is the difference between a merits forum and a forum that must take the department’s summary of the record on trust.

Section 19 gives the Tribunal the power to punish for contempt of itself, with the status of a High Court for that purpose under the Contempt of Courts Act framework, a power whose place in the general law of contempt is set out in the series article on the Contempt of Courts Act 1971. The contempt power does double work in this Act. It protects the Tribunal’s proceedings against obstruction and scandalising conduct, as any contempt power does. It also becomes, in practice, one of the instruments by which applicants seek to compel implementation of orders the department has not carried out, a use examined in the section on execution later in this article. Section 30(2) treats contempt differently from every other order of the Tribunal: an appeal lies to the Supreme Court as of right from an order punishing for contempt, to be filed within sixty days. The asymmetry is deliberate and revealing. Where the Tribunal’s own coercive power is used against a person, Parliament trusted the person with an appeal as of right; where the Tribunal decides a service dispute, Parliament interposed the leave filter.

The Tribunal is not bound by the procedure of the Code of Civil Procedure in its hearings; it is guided by the principles of natural justice and by the Act and the Procedure Rules, and it regulates its own procedure, including the fixing of time and place of hearings. That procedural freedom is the standard tribunal bargain, formality traded for speed and accessibility, and it is supported by the Tribunal’s power to decide matters on affidavit evidence. Applicants before the Tribunal are entitled to the assistance of the legal services machinery where they qualify for it, and the framework of legal aid institutions through which such assistance is organised is described in the series article on the Legal Services Authorities Act 1987, a route of practical importance to pensioners and to the families of deceased personnel, who form a large share of the Tribunal’s applicants and who litigate at a distance from the benches.

Execution: the order that must still be carried out

Section 29 provides that, subject to the other provisions of the Act and the rules, the order of the Tribunal disposing of an application shall be final and shall not be called in question in any court, and such order shall be executed accordingly. The sentence has two limbs, and they pull in different directions in the Tribunal’s history. Finality has been respected: the appeal channel is narrow, as the next section shows, and the writ channel is a supervisory one. Execution has been the weaker limb. The Act declares that the order shall be executed, but the executing hand belongs to the department against which the order is made, and the distance between a favourable order and its implementation, arrears calculated, pension payment orders issued, records corrected, is where a substantial part of the Tribunal’s later litigation has been generated.

The reported position, stated with the attribution it requires, is that applicants have repeatedly had to return to the Tribunal, and in numbers of cases to invoke its contempt jurisdiction, to obtain implementation of orders in their favour, and that benches of the Tribunal have recorded disapproval of delay in implementation by the authorities in pension and pay matters. The Supreme Court has also had occasion, in appeals arising from this Tribunal, to direct time bound compliance in individual cases. The structural explanation is not mysterious. Implementation requires the respondent department to move its own record, accounts and pension authorities, and the Tribunal has no executing establishment of its own comparable to the executing courts of the civil system. Its sanctions are the contempt power and the costs and directions it can attach, instruments that work case by case.

Candour requires the counterweight. A pension system administering the records of a force of well over a million serving personnel and a retired population of comparable scale will generate implementation friction in any design, and many orders require recalculation across pay commissions and policy changes that take genuine administrative time. The criticism that carries weight is not that implementation takes time but that the applicant must litigate again to obtain what an order has already granted, transferring the cost of the department’s delay to the person the order favours. On the evidence of the Tribunal’s own later dockets, execution is the point at which the statute’s promise is least completely kept, and it is the point on which reform proposals directed at this Tribunal, including proposals for an implementation cell and for interest on delayed compliance, have concentrated.

How does the Tribunal secure compliance with its orders?

Section 29 declares the Tribunal’s order final and directs that it be executed accordingly. Behind the declaration stands the Tribunal’s contempt power under Section 19, its power to review its own decisions, and the Supreme Court’s appellate and supervisory authority. An applicant facing non implementation ordinarily returns to the Tribunal for directions or contempt proceedings.

In practice the sequence runs from an execution application or a representation to the department, through a further order of the Tribunal fixing time for compliance, to contempt where default persists. That the sequence is needed at all is the candid weakness of the scheme: the Act gives the Tribunal no executing machinery of its own, so the compliance of the respondent department is obtained by the pressure of successive orders rather than by attachment and delivery through an executing court. Applicants and their advisers should therefore treat the first order as the beginning of enforcement, diarise the period it allows, and move promptly when it expires, because the record of this jurisdiction shows that unimplemented orders do not implement themselves.

The appeal to the Supreme Court: Sections 30 and 31

Chapter V of the Act contains the provision that, with administrative control, defines this Tribunal. Section 30(1) provides that, subject to Section 31, an appeal shall lie to the Supreme Court against the final decision or order of the Tribunal, other than an order passed under Section 19, provided the appeal is preferred within ninety days. There is no appeal against an interlocutory order. Section 31 then imposes the condition that gives the scheme its character. No appeal shall lie under Section 30 unless the Tribunal certifies that the case involves a point of law of general public importance, or the Supreme Court is of the opinion that the point is one that ought to be considered by it and grants leave. The gateway to the Supreme Court therefore has two keys, and the first of them is held by the Tribunal whose order is challenged.

The consequences are considerable. A service dispute that is intensely important to the applicant, a pension wrongly computed, a promotion wrongly denied, a dismissal wrongly sustained, raises no point of law of general public importance in the ordinary case, and such an applicant’s statutory appeal is, in substance, unavailable: the Tribunal will not certify a point that is not there, and the Supreme Court’s leave under Section 31 is exercised for questions of general importance rather than for the correction of individual error. Individual grievances are thus finally determined by the Tribunal, subject to the writ jurisdiction and to the Supreme Court’s constitutional powers, which the Act does not and cannot remove. The design was a choice, not an oversight. The personnel who met the Law Commission in 1999 asked for an appeal directly to the Supreme Court, and Parliament supplied the direct route while filtering it, on the stated logic that discipline and finality in a force require service litigation to end, and that the Supreme Court’s docket cannot receive the whole output of a busy tribunal as of right.

The filter deserves to be stated precisely, because it is frequently misdescribed in both directions. It is not a bar on the Supreme Court’s jurisdiction: the Court may grant leave under Section 31 itself, and its powers under Article 136 of the Constitution stand outside the Act. It is not a certificate of the kind a litigant obtains as a formality: benches of the Tribunal have declined certificates where the case turned on its own facts, and the Supreme Court, in appeals and leave applications from this Tribunal, has repeated that a point personal to the litigant does not become a point of general public importance by being pressed hard. And it is not symmetrical with the rest of the Act’s scheme, because of the contempt exception in Section 30(2): the single appeal as of right in the Act protects the person punished by the Tribunal, not the person whose service claim the Tribunal has rejected.

Set beside the general tribunal pattern, the filter is the AFT’s most distinctive feature. The ordinary tribunal statute of the tribunalisation era routed appeals to a division bench of the High Court, preserving a judicial tier above the tribunal within the state. This Act contains no High Court appeal at all. The reasons given at the design stage were uniformity, since service law is central and a single national appellate authority avoids divergent High Court rulings on the same service rules, and speed. The cost is the one the next section examines: the High Court enters this field only through the writ jurisdiction, which the courts spent a decade defining for this Tribunal.

The writ question: self restraint, ouster, and the correction of 2023

Section 33 excludes the jurisdiction of courts other than the Supreme Court and the High Courts under Articles 226 and 227 in matters within the Tribunal. Two constitutional provisions frame what that exclusion can mean. Article 227(4) provides that the High Court’s power of superintendence does not extend to courts and tribunals constituted under laws relating to the armed forces, so the Tribunal stands outside High Court superintendence by the Constitution’s own text. Article 226 contains no such carve out, and since L Chandra Kumar v Union of India, (1997) 3 SCC 261, it has been settled that the writ jurisdiction of the High Courts and the Supreme Court’s jurisdiction under Article 32 are part of the basic structure, which no tribunal statute can exclude. Even before the tribunal era, S N Mukherjee v Union of India, (1990) 4 SCC 594, had held that court martial proceedings are amenable to judicial review on the recognised grounds. The question for this Tribunal was therefore never whether the writ jurisdiction existed, but how it should be exercised.

The Supreme Court answered in Union of India v Major General Shri Kant Sharma, (2015) 6 SCC 773. The Court summarised the position in propositions that became the working law of the field for eight years: the power of judicial review under Article 226 is a basic feature that the Act cannot curtail; the High Courts and the Supreme Court will still have due regard to the legislative intent and exercise jurisdiction consistently with the Act; where a statutory forum is created for redressal of grievances, a writ petition should not be entertained ignoring the statutory dispensation; and a High Court will not entertain an Article 226 petition where an effective alternative remedy is available or the statute contains its own redressal mechanism. Applied to this Tribunal, the propositions were widely read, and in many High Courts applied, as a direction to refuse writ petitions against the Tribunal’s orders and against service orders within its jurisdiction, remitting the litigant to the Tribunal and to the filtered appeal.

That reading did not survive. In Union of India v Parashotam Dass, 2023 SCC OnLine SC 314, decided on 21 March 2023 by a bench of Justices Sanjay Kishan Kaul, Abhay S Oka and B V Nagarathna, the Supreme Court held that the judgment in Shri Kant Sharma did not lay down the correct law and was in conflict with the Constitution Bench judgments in L Chandra Kumar, S N Mukherjee and Rojer Mathew. The Court held that there is no per se restriction on the exercise of power under Article 226 by a High Court in relation to orders of this Tribunal, that to deny the High Court the power to correct an error into which the Tribunal may fall would be against the constitutional scheme, and that judicial scrutiny under Article 226 is not an appeal: the first independent judicial scrutiny of a service decision is the Tribunal’s own, and scrutiny of the Tribunal by the High Court does not convert the writ into a second appeal. Matters that had been disposed of as not maintainable in reliance on Shri Kant Sharma were remitted to the High Courts for decision on merits.

The position that emerges, and the answer to the question applicants in this field ask most often, is therefore a qualified yes with defined limits. A High Court can be approached under Article 226 against an order of the Tribunal, and the jurisdiction is constitutional rather than statutory, so the Act neither grants nor removes it. What the writ court does is judicial review, not appeal: it corrects jurisdictional error, error apparent, denial of natural justice and denial of fundamental rights, and it does not reappreciate the evidence or substitute its view of the service merits for the Tribunal’s. The discipline of self restraint survives the overruling as a matter of practice rather than of bar, since a litigant who bypasses the Tribunal at first instance, where the Tribunal can decide the merits, will ordinarily be remitted to it, not because the High Court lacks power but because the statutory forum exists to decide the question first. The error to avoid is the one the brief for this article flags: stating that the High Court’s jurisdiction is excluded. It is not excluded; it is reserved by Section 33 itself, structured by the Constitution Bench line of authority, and exercised within the parameters of review.

This history also explains the Tribunal’s distinct place in the appellate geography of Indian tribunals. For most tribunals, L Chandra Kumar supplied the missing tier by holding that tribunal decisions are subject to scrutiny by a division bench of the High Court. For this Tribunal, the combination of Article 227(4), the absence of a statutory High Court appeal, and eight years of Shri Kant Sharma produced an interval in which the filtered Supreme Court route appeared to be the only channel, an interval Parashotam Dass closed by restoring the writ channel to its constitutional footing. The Tribunal is therefore best described as a forum of final merits determination, supervised by the High Courts on review principles and by the Supreme Court through a filtered appeal, which is a closer supervision than its design suggested and a looser one than the ordinary tribunal receives.

Who administers the Tribunal: the Ministry of Defence question

The Tribunal’s administrative home is the Ministry of Defence. Its budget is carried in that Ministry’s demands, its infrastructure, staff sanctions and administrative support run through that Ministry, and the department that appears as respondent in the great majority of proceedings before the Tribunal, the Union of India acting through the Ministry of Defence and the service headquarters, is the department that houses the forum. This is the second of the two features the brief for this article identifies as distinctive, and it is the feature on which the structural criticism of the Tribunal has concentrated.

The criticism is not that the Ministry interferes in adjudication, and it should not be stated in that form, because the record does not support it: benches of the Tribunal decide against the Ministry daily, in pension, pay and promotion matters, and no serious participant in the debate alleges direction of outcomes. The criticism is about the appearance and the machinery of independence. The judge made independent of the parties in the courtroom remains dependent on one party for the courtroom: for the sanction of posts, the filling of vacancies, the provision of registries and residences, and the processing of the Tribunal’s own administrative needs. A litigant who has served in the force is asked to accept that the forum deciding against the department is administered by the department, and the constitutional law of tribunals in India has treated that arrangement, wherever it appears, as a defect in design rather than a detail of housekeeping.

The recommendation to move the Tribunal has been made repeatedly and has not been effected. The Law Commission’s 272nd Report of 2017, which assessed the statutory frameworks of tribunals generally, recommended a National Tribunals Commission and a nodal agency structure that would take tribunal administration out of the sponsoring ministries, and parliamentary committees examining defence and tribunal administration have pressed versions of the same point for this Tribunal, favouring transfer of administrative control to the Ministry of Law and Justice. The Supreme Court’s tribunals jurisprudence, from Rojer Mathew v South Indian Bank in 2019 through the Madras Bar Association judgments to the judgment of November 2025 on the Tribunals Reforms Act 2021, has treated ministerial administration of tribunals as the central structural flaw of the tribunal system and has directed the creation of an independent commission to end it. Through all of this the Tribunal has remained where the Act’s administrative practice placed it, under the Ministry of Defence. The transfer has been recommended at every level at which the question has been examined, and it has not happened; the article records that position without embellishment, because the gap between recommendation and action is the fact.

The case for the arrangement should be given in its strongest form, as the stance of this series requires. The Ministry’s defenders point out that the Tribunal’s work is unintelligible without the services: its records are service records, its applicants are serving and retired personnel whose documentation sits with the services, its benches sit in cantonment cities, and its administrative member stream comes from the force. Administrative proximity, on this view, is operational necessity rather than capture, and the Tribunal’s output, which includes a very large volume of orders against the Ministry, is the practical refutation of the capture thesis. There is force in the output point. But it answers a charge that is not the charge. The constitutional objection is not that this Tribunal decides for the Ministry; it is that no adjudicating body should depend for its functioning on a party before it, because independence in the tribunal cases is a structural condition, tested by who controls appointments, budgets and administration, and not an inference from outcomes. On that test, which the Supreme Court’s line of authority supplies, the arrangement fails, and the verdict of this article is that the administrative control of the Tribunal by the Ministry of Defence is its principal design defect, curable by the transfer that has been recommended for years and not made.

Disability pension: the largest stream

Disability pension is the largest single category of case before the Tribunal, and the brief for this article directs that the framework be described rather than the merits of individual claims. The framework begins inside the service. A member of the armed forces invalided out of service, or retiring with a disability, is assessed by a medical board, which records the disability, its percentage, and its characterisation: whether the disability is attributable to service, aggravated by service, or neither. Entitlement to disability pension under the pension regulations governing the services turns on that characterisation and on the percentage, with a service element and a disability element composing the award. The Tribunal’s jurisdiction over the dispute is the Section 14 jurisdiction: pension is expressly within Section 3(o).

The governing approach in the case law is the one the Supreme Court laid down in Dharamvir Singh v Union of India, (2013) 7 SCC 316: a member is presumed to have been in sound physical and mental condition on enrolment, the medical board’s opinion is entitled to respect but is not conclusive against the record, and where a disability arises during service and was not recorded at entry, the benefit of the attribution question is approached liberally in favour of the member, since the conditions of service are themselves capable of causing or aggravating conditions that present later. The Tribunal applies that framework to the records of individual boards: whether the board gave reasons, whether it addressed the service conditions the applicant describes, whether a constitutional condition unrelated to service has been correctly identified as such. A substantial part of the Tribunal’s disability docket consists of cases in which a board recorded a disability as neither attributable nor aggravated without engaging with the applicant’s service history, and the Tribunal’s corrective role in those cases is among the most consequential work it does, because the sums are modest and the applicants are, in the main, other ranks and their widows, for whom the pension is a principal income.

Two cautions belong in any account of this stream. The first is that the Tribunal decides on records: it does not conduct medical examinations, and its intervention is directed at the legality and rationality of the board’s characterisation, informed by the administrative member’s understanding of how boards work. The second is that the framework’s generosity at the level of principle coexists with demanding proof at the level of the individual file, and applicants who treat a long service record as self proving misunderstand what the Tribunal needs. This article is legal information about the framework, not advice on any individual claim, and nothing in it should be read as a prediction of the outcome of a case that turns on its own medical record.

Reform, conditions of service and vacancies

The Tribunal has not stood outside the general litigation over tribunal reform. The Finance Act 2017, in Part XIV, empowered the Central Government to prescribe qualifications, appointments, term and conditions of service for a schedule of tribunals that included this one, and the rules framed under it were struck down in Rojer Mathew. The Tribunals Reforms (Rationalisation and Conditions of Service) Ordinance and then the Tribunals Reforms Act 2021 reset the conditions of service for members of tribunals generally, including the members of this Tribunal, and the fate of those provisions is the subject the courts cluster of this series owns: the Supreme Court’s judgment of November 2025, reported at 2025 INSC 1330, struck down Sections 3 to 7 of the 2021 Act while leaving its abolitions untouched, and the conditions of members of this Tribunal stand governed by that settlement rather than by any distinct dispensation. Appointment and service condition litigation concerning this Tribunal is therefore best read as part of that general line, and this article keeps it at that level deliberately.

Vacancies are the operational face of the same problem. The Tribunal’s sanctioned strength is fixed by notification, and benches cannot sit without members. The position reported over the years, in answers given in Parliament and in the reports of the committees that examine the defence establishment, is of sanctioned posts lying vacant for substantial periods and of benches functioning below strength, with the predictable consequence that lists lengthen in the busiest registries. This article attributes the vacancy position in that form, to parliamentary answers and committee reports across the period, because the precise number of vacancies is a moving figure that changes with each appointment round, and a figure printed without its date would mislead. What does not move is the structure that produces the vacancies: appointments require a process the executive controls, the administrative home of the Tribunal is a respondent ministry, and the gap between a member’s retirement and a successor’s appointment is absorbed by litigants as delay.

The interaction between vacancies and the appeal filter deserves notice, because it compounds. A litigant whose case waits behind an under strength bench, and who then faces a leave filter at the Supreme Court and a supervisory rather than appellate writ jurisdiction, experiences the system as a series of narrow gates. Each gate has its rationale: specialisation, finality, discipline, docket control. The series thesis that runs through this cluster of articles is that a tribunal’s independence and accessibility are determined less by the quality of its members than by its administrative arrangements and its appellate route, and this Tribunal is the instance in which both are configured against accessibility while the membership is strong. That is why the criticism of the Armed Forces Tribunal, when it is made carefully, is structural rather than personal, and why the remedies proposed for it, transfer of administrative control, a functioning appointments pipeline, and reconsideration of the leave filter, are institutional rather than disciplinary.

Why was a specialised tribunal preferred to the ordinary courts?

The case made to the Law Commission, and accepted in its 169th Report, was that service disputes were decided in writ jurisdiction by a remedy that could not examine the merits, in slow proceedings, before judges without service experience. A specialised tribunal promised merits adjudication, service expertise on the bench, and one national forum for one national service law.

The counter consideration, which the design’s critics press, is that specialisation purchased by separation carries its own price: the forum must be equipped with the independence the ordinary courts possess by constitutional design, or the litigant has traded a slow court for a dependent tribunal. The history traced in this article suggests that the first half of the bargain was delivered, the forum exists, sits across the country and decides merits, and the second half, independence of administration and an appellate structure that does not gatekeep itself, remains the unfinished portion of the settlement of 2007.

The court martial system the Tribunal reviews

To understand the appellate jurisdiction, the institution appealed from must be understood. A court martial is the criminal trial of the military system, constituted under the service Acts for offences those Acts create or adopt. The Army Act 1950 knows four kinds: the general court martial, with the widest sentencing powers, extending in the gravest cases to death; the district court martial, with limited powers; the summary general court martial, designed for use on active service and in places where a general court martial cannot conveniently assemble; and the summary court martial, held by a commanding officer, swift and limited in punishment, which Section 3(o) of the 2007 Act largely keeps away from the Tribunal. The Navy Act works through disciplinary courts and courts martial of its own pattern, and the Air Force Act follows the Army model closely. In each system the trial is conducted by service officers as members of the court, advised on the law by a Judge Advocate, prosecuted and defended within the service’s legal arrangements, and recorded in proceedings that become the record on appeal.

Over that trial the service Acts built layers of internal correction. Findings and sentences of the graver courts martial required confirmation by a confirming authority, which could withhold confirmation, refer the finding or sentence back for revision, mitigate, remit or commute. The Judge Advocate General’s branch examined the proceedings for legality and advised the confirming authority. The convicted person could petition: pre confirmation and post confirmation petitions under the Army Act, including the petition under Section 164(2) to the Central Government, the Chief of the Army Staff or a prescribed officer, asking for annulment or mitigation on grounds including illegality and injustice. None of these layers, singly or together, was an appeal. Each was an exercise of executive or command discretion within the same institution that had convened the trial, advised in law by the service’s own legal branch. The sentence could be softened; the finding could be disturbed for illegality; but no independent body re heard the case as of right, and the person convicted had no entitlement to have the evidence re weighed by anyone outside the force.

The Constitution framed this structure without displacing it. Article 33 empowers Parliament to restrict or abrogate the fundamental rights of armed forces personnel to the extent necessary for discipline and duty, and the service Acts rest on it. Article 136(2) excludes the Supreme Court’s special leave jurisdiction over judgments of courts and tribunals constituted under laws relating to the armed forces, which is why, before 2007, the court martial convict’s constitutional routes were the writ under Article 226 and the petition under Article 32, both supervisory and neither appellate. When the Supreme Court in Prithi Pal Singh Bedi sustained the system while deploring the absence of an appeal, it was describing this exact architecture: a valid disciplinary code, internally reviewed, externally supervised on narrow grounds, and nowhere appealed.

Section 15 changed the architecture at its centre. The Tribunal is not another layer of the same kind. Its judicial members are judges, its proceedings are public adjudication, its grounds in Section 15(4) include the sustainability of the finding itself, and its Section 15(6) powers include substitution of a finding, which no confirming authority possessed in that form. The internal layers survive beside it, and the interaction is a matter of sequencing rather than election in most cases: confirmation operates on the sentence, the statutory petition may be pursued, and the appeal goes to the Tribunal. What the appellant gains is the thing the old system structurally withheld, a hearing on the merits of the conviction before a body that owes nothing to the chain of command in its adjudicative capacity.

One feature of the appellate jurisdiction is regularly misunderstood and should be corrected. The Tribunal on a Section 15 appeal is not a second court martial. It does not assemble witnesses afresh as a matter of course or conduct the trial again; it works on the record of the proceedings, testing the finding against the evidence recorded, the law against the Judge Advocate’s summing up and the court’s directions, and the procedure against the service Act and rules under which the trial was held. Where the record cannot sustain the finding, the appeal succeeds on Section 15(4)(a). Where the law was misstated, it succeeds on Section 15(4)(b). Where irregularity has caused a miscarriage, it succeeds on Section 15(4)(c). The discipline of those three grounds, applied to a paper record by a bench that includes an officer who has run courts martial from the legal branch’s side, is the closest Indian military justice comes to the criminal appeal familiar in the civil courts, and it is materially more than the writ review that was the whole of the external remedy before 2009.

Benches, registries and the geography of access

The Tribunal sits where its litigants are. The Principal Bench sits at New Delhi, and regional benches sit at Chandigarh, Lucknow, Kolkata, Guwahati, Chennai, Kochi, Mumbai and Jaipur, with the bench for Srinagar serving the personnel and pensioners of Jammu and Kashmir and Ladakh, whose service litigation is heavy in proportion to population because of the scale of recruitment and deployment from the region. Circuit sittings have been used to carry benches to stations within a region. The distribution is not a courtesy; it is a condition of the forum’s usefulness. A pensioner in Kerala or a dismissed soldier in Rajasthan cannot litigate in Delhi, and a tribunal that required Delhi would reproduce, in a new form, the access problem the writ system presented when service cases were scattered across High Courts of uneven distance from the cantonments.

Each bench is served by a registry that receives applications and appeals, scrutinises them for the requirements of the Procedure Rules, lists them before benches and maintains the record. The Chairperson holds the financial and administrative powers of the Tribunal under Section 12, including the constitution of benches and the distribution of business, and Section 28 provides for decision by majority where members of a bench differ, with a reference to the Chairperson on a point of difference where the bench is equally divided. Transfers of cases between benches are provided for where the interests of justice or the location of the record and the parties require them, a power of practical significance where an applicant has retired far from the station at which the service events occurred.

The geography also explains the Tribunal’s docket profile. Chandigarh, serving Punjab, Haryana and Himachal Pradesh, from which recruitment is deep and pensioner populations are large, has carried one of the heaviest lists in the system, dominated by pension, disability and pay fixation matters. Lucknow serves the largest state and a dense cantonment network. The southern benches carry the pension litigation of the Madras Regiment’s recruiting ground and of the naval establishment on both coasts. This regional distribution of subject matter, invisible in the statute, is the lived shape of the jurisdiction, and it is why the vacancy problem bites unevenly: an unfilled seat in a heavy registry delays more justice than the same vacancy in a light one.

Procedure: how a case is brought and heard

A service matter begins as an original application to the bench within whose territory the cause of action arises or as the rules of allocation direct, in the form the Armed Forces Tribunal (Procedure) Rules 2008 prescribe, supported by the documents on which the applicant relies and the order under challenge. A court martial appeal is presented in the prescribed form and within the prescribed time, accompanied by the record or by the application through which the record is requisitioned from the service authorities. The respondent is ordinarily the Union of India through the Ministry of Defence and the service headquarters, the record office and the pension sanctioning authorities, as the case requires. Interim orders are available: stay of a recovery, stay of the operation of an order of discharge or compulsory retirement, and in the appellate jurisdiction the bail power under Section 15(3) and suspension of sentence in the circumstances the Act contemplates.

Time discipline in this jurisdiction has a human shape. The applicants include serving personnel posted in field areas, pensioners of advanced age, and widows pursuing family pension and disability claims on the papers of a deceased husband or son. The Tribunal’s power to condone delay, exercised on sufficient cause, has been applied with those realities in view, and delay explained by the pace of departmental replies, the time taken to obtain service records, or the applicant’s circumstances has ordinarily been treated differently from delay that is unexplained or that has altered the position of others. At the same time, service law values finality with unusual intensity, because seniority and promotion are relative: an order unsettling a seniority list after years disturbs the settled expectations of every officer below and above the applicant, and the Tribunal’s case law reflects that systemic interest when it measures delay and laches in promotion and seniority matters.

Hearings before the Tribunal are shorter and less formal than in the civil courts, conducted on affidavit evidence in most service matters, with oral argument concentrated on the legal questions the record presents. The administrative member’s presence alters the hearing’s character in a way applicants notice: explanations of service practice that a civil court would receive as expert evidence, the meaning of a grading profile, the working of a selection board, the sequence of a medical categorisation, are matters the bench can test from knowledge, and counsel address them accordingly. The Tribunal is not bound by the Code of Civil Procedure’s procedure, being guided by natural justice, but its Section 17 powers give it the civil court’s coercive equipment where evidence must be compelled, and its orders are reasoned orders, subject to the appellate and writ channels this article has described.

The ordinary docket: pay, pension and promotion beyond disability

Disability pension is the largest single stream, but the Tribunal’s daily work is wider, and its character is set by three further streams. The first is pay and pension fixation. Each pay commission’s implementation generates a cohort of disputes: the fitment of a rank, the treatment of a special allowance, the reckoning of qualifying service, the effect of a break in service or of a period of extraordinary leave, the revision of pension following a court martial sentence of forfeiture or a dismissal converted on representation. Pension is a right and not a bounty, on principles the Supreme Court settled long before this Tribunal existed, and the Tribunal’s role is to apply the pension regulations to the individual record, correcting the sanctioning authority where the regulation has been misread or the record misread. Errors in this stream are usually small in money and large in consequence, because a pensioner’s income is the pension, which is why the stream’s volume measures administrative error rates as much as litigiousness.

The second stream is promotion and selection. Challenges to selection board outcomes turn on the confidential report profile, the criteria the board was directed to apply, comparative merit within the batch, and the integrity of the process: whether a report was shown to be actuated by bias, whether an adverse entry was communicated where communication was required, whether the board considered material the policy excluded or ignored material it required. The Tribunal’s approach follows the settled restraint of service law, that a court does not sit as a selection board, combined with the merits jurisdiction that Section 14 confers: it does not re grade officers, but it examines whether the process the service prescribed for itself was followed and whether the profile considered was the profile the officer had actually earned. Non empanelment and supersession cases, among the most bitterly felt in an officer’s career because they usually end it, are decided in that register, and the administrative member’s knowledge of how boards actually deliberate is brought to bear on whether the record supports the process claimed for it.

The third stream is tenure and exit: premature retirement, compulsory retirement, release and discharge, resignation refused, extension denied, and the conversion of one form of exit into another with pension consequences. Here the Tribunal applies the public law of punitive and non punitive orders to the service context, asking whether an order wearing the form of a simpliciter retirement rests in substance on misconduct that required an inquiry, whether the material before the authority bore on the officer’s utility and integrity in the sense the rules intend, and whether a discharge following summary disposal was within the disposal power. Each of these questions arrives at the Tribunal framed by the service Acts and the rules and instructions under them, and each is decided on the record, which returns the article to its recurring theme: in this jurisdiction the record is the case, and the side that holds the record, the department, holds it subject to the Tribunal’s power to requisition it and to draw the consequences of its non production.

Pension litigation on questions of general application also reaches the Supreme Court from this Tribunal despite the filter, because some questions are points of general public importance in the true sense. The One Rank One Pension litigation, in which the Indian Ex Servicemen Movement challenged aspects of the scheme’s implementation, was decided by the Supreme Court in March 2022 and illustrates the route: a question affecting the entire pensioner body is certifiable and is the kind of question the Section 31 filter was designed to admit, while the individual computation disputes that surround it remain with the Tribunal. The distribution is the statute working as designed, and it shows the filter’s real character, which is not hostility to appeal but a sorting of appeals by generality.

A word should be said about the standard of the department’s own decision making that this docket reveals. The Tribunal’s orders, read across the streams, correct a recurring set of administrative failures: the order that gives no reasons where the service’s own instructions require them, the board that does not record the material it considered, the sanctioning authority that applies a later policy to an earlier entitlement or an earlier policy to a later one, the record office that treats its summary of the record as the record. None of these failures requires a tribunal to detect; each could be prevented inside the department by the disciplines the Tribunal’s case law repeats. The persistence of the streams is therefore evidence about the administration as well as about the forum, and it qualifies the flood argument that supported the exclusions in 2005 and 2006: a department whose first instance decisions were more carefully made would face a smaller docket in any forum, and part of the Tribunal’s systemic value is the pressure its published orders place on the quality of the decisions that precede it. That pressure, applied order by order across two decades of reported decisions, is the quiet mechanism by which a tribunal improves the administration it reviews, and it operates in this jurisdiction with unusual directness because the respondent in nearly every case is the same department, able to read the pattern of its own reversals and act on it.

The limits of the Tribunal’s power: policy and review

A merits tribunal needs its limits stated as carefully as its powers, and two limits define this Tribunal’s practice. The first is policy. The Tribunal decides disputes under the law and the policy that governs them; it does not frame the policy. The Supreme Court made the point in Union of India v Air Commodore N K Sharma, decided in December 2023, setting aside a direction of the Tribunal that required the government to frame a policy for filling a Judge Advocate General branch vacancy: the determination of the age of superannuation and the structure of a promotion policy lie within the executive’s domain, and a tribunal that grants individual relief within its jurisdiction exceeds that jurisdiction when it directs the content of policy for the service at large. The distinction runs through the case law: the Tribunal may hold that an officer was wrongly denied the benefit of a policy, that a policy was applied unequally, or that a policy as framed violates the service Acts or the Constitution and must be read accordingly in the case before it, but the remedy stops at the parties and the point decided.

The second limit is internal. The Tribunal may review its decisions under Section 17’s inclusion of the civil court’s review power, on the narrow grounds review permits, error apparent and the other recognised heads, and it may correct clerical and arithmetical slips. Review is not a rehearing, and the benches have resisted its use as an appeal in disguise, for the same reason the appellate discipline of Section 15(4) is kept narrow: a system serving a disciplined force, in which seniority and sentence both demand settled outcomes, cannot treat every order as provisional. The litigant dissatisfied with the substance of an order has the channels described in this article, certificate and leave to the Supreme Court, or judicial review under Article 226, and review before the Tribunal is the instrument for the order that is wrong on its face, not for the order that is merely debatable.

A comparative glance: military appeal elsewhere

Other democracies that retain courts martial have faced the same design problem, and their solutions show the Indian choices in relief. The United Kingdom, from whose military law the Indian service Acts descend, created the Courts Martial Appeal Court in 1951, staffed by judges of the Court of Appeal, hearing appeals against court martial convictions on grounds closely resembling the criminal appeal, with a further appeal to the Supreme Court of the United Kingdom on a certified point of law of general public importance. The resemblance to Section 31 is not accidental: the certificate formula is the Commonwealth’s standard device for filtering military appeals at the final tier. The British tribunal of first instance for service grievances, by contrast, does not exist in the AFT’s form; service complaints run through an ombudsman model, the Service Complaints Ombudsman, whose recommendations are not adjudication. India chose adjudication for grievances as well as appeal for convictions, and in that respect the AFT is the more complete forum.

The United States maintains the sharpest separation of the three systems. Courts martial under the Uniform Code of Military Justice are reviewed within each service by a Court of Criminal Appeals with fact finding power, then by the Court of Appeals for the Armed Forces, a civilian court of five judges appointed by the President with the Senate’s consent, and thereafter by certiorari to the Supreme Court of the United States in a restricted class of case. The American design places the final military appeal court at a civilian distance from the services that the Indian design does not attempt: the AFT’s administrative members are retired officers of the very services whose cases are heard, a feature American practice would treat with suspicion and Indian practice treats as the forum’s defining expertise. Canada stands between the two, with a Court Martial Appeal Court drawn from the federal and superior court judiciary. The comparison yields no verdict by itself, since each system answers its own constitutional history, but it locates the AFT accurately: a combined grievance and appeal forum, mixed in composition, filtered at the apex, of a kind no other major military power replicates in exactly this combination.

Appointments: the search committee era and this Tribunal

The method of choosing members connects this Tribunal to the decade of litigation that Article 280 of this series recounts. As enacted, Section 7 required appointment by the President after consultation with the Chief Justice of India, a formula that gave the judiciary a voice without a committee structure. The Finance Act 2017 sought to move the tribunal system to rule based qualifications and search cum selection committees under executive framed rules, and the rules of 2017 were struck down in Rojer Mathew v South Indian Bank in 2019 for, among other defects, executive dominance of the selection process. Parliament responded in the Tribunals Reforms Act 2021 with a statutory committee pattern, and the Supreme Court’s successive Madras Bar Association judgments, concluding in the judgment of November 2025 reported at 2025 INSC 1330, struck down the core conditions provisions of that Act, Sections 3 to 7, while leaving its tribunal abolitions standing. For this Tribunal the practical consequence is that appointments proceed under the 2007 Act’s own provisions as overlaid by the principles the Court has laid down for all tribunals, and the recurring vacancies described earlier in this article are the local cost of a national appointments contest that has been running, unresolved in legislation, for the better part of a decade.

It is worth stating what the litigation is and is not about, because the point is often lost. No judgment in the line questions the Tribunal’s existence, its jurisdiction or its mixed composition. The contest is over who selects, for how long members serve, and under whose administrative control they work, the three levers by which an executive that is also the principal litigant could, in design, weaken a forum without ever touching a case. The Court’s answers have consistently protected tenure long enough for independence, selection processes with judicial participation, and administration separated from the sponsoring ministry. Applied to this Tribunal, those answers condemn the Ministry of Defence’s administrative control by the same reasoning that condemns the four year term’s insecurity in the general scheme, which is why this article treats administrative control, rather than composition or jurisdiction, as the defect on which reform should concentrate.

The discipline argument, stated fairly

The strongest argument for the design as it stands is discipline, and it deserves its full force. A fighting force is not a civil department. Its members can be ordered into danger, its commanders must be obeyed in circumstances where disobedience kills, and its personnel decisions, postings, selection, the composition of a unit before an operation, are taken under operational pressure that no court shares. On this view, the exclusions of transfers, postings and leave are not gaps but the point: they mark the territory in which command must be answerable upward through the chain, not outward to a tribunal. The leave filter similarly protects the finality without which a commander’s authority, once second guessed at leisure by a distant forum, would soften in the field. Even the administrative location has its operational defence: the Ministry that runs the force runs the forum because the forum’s inputs, records, members and litigants all come from the force.

The answer is not that discipline is unimportant; Article 33 of the Constitution is the framers’ own recognition that it is a constitutional value in this field. The answer is that the argument proves too much at the points where the design is weakest. Administrative control of the Tribunal does not protect discipline in the field; it protects the department’s convenience in litigation, and no operational commander has ever suggested that a posting decision is safer because the Tribunal’s budget sits in the Ministry. The leave filter’s discipline rationale applies to the appeal in a court martial case, where finality of conviction matters to the unit, and applies far more weakly to a pension computation or a pay fixation, which are accounting questions wearing a uniform. A design that sorted its restrictions by the strength of the discipline interest, strict for courts martial and operational matters, ordinary for money and record matters, would answer the discipline argument better than the present design does, and the absence of any such sorting is a legitimate ground of criticism that does not depend on doubting the sincerity of the discipline case.

The verdict this article reaches is the one its structure has prepared. The Tribunal was a necessary creation, fifty years late against the Supreme Court’s first prompting, and its adjudicative work, merits review of service decisions and true appeals from courts martial by a bench that understands both law and service, is a permanent gain in the legal position of the Indian soldier, sailor and airman. Its unfinished business is institutional: administration by the respondent ministry, an appointments pipeline that leaves benches short, execution that depends on the respondent’s pace, and an appellate filter that the restoration of writ review in 2023 has mitigated but not replaced. Those are defects of structure, and they are curable by legislation that Parliament has been recommended to enact, in one form or another, for most of the Tribunal’s life.

Transition: the docket the Tribunal inherited

Section 34’s transfer of pending proceedings deserves more than the sentence it usually receives, because the transfer shaped the Tribunal’s early character. On the appointed day, service and court martial matters pending before the High Courts across India, and before other courts and authorities within the Act’s field, stood transferred to the Tribunal and continued from the stage reached. The transferred docket was not a sample of the jurisdiction; it was the jurisdiction’s backlog, accumulated over years in forums that could only review and not decide. Cases arrived at every stage: writ petitions admitted and awaiting hearing, matters part heard, records called for and not produced. The benches had to convert review proceedings into merits proceedings, call for the service records on which decision depended, and apply Section 14 and Section 15 to disputes whose facts were, in some cases, a decade old.

The transition also carried a human transfer that the statute does not mention. Litigants who had briefed counsel in the High Courts found their cases before a new forum in a different city, under new forms, with a registry building its practice as it went. The Procedure Rules 2008 and the Practice Rules that followed had to be written for a litigant population that includes serving personnel under posting orders, and the Tribunal’s early practice on transfers between benches, on exemption from personal appearance, and on the production of records by the services was formed under that pressure. The significance for a reader of the Act is that the Tribunal’s institutional habits, its insistence on the complete service record, its condonation practice, its execution follow up, were set in the transition years by the condition of the docket it inherited, and they have persisted because the docket’s condition has persisted: the Tribunal receives, in the main, disputes that the administrative system has already decided once and defended once before they arrive.

The dependant applicant: family pension and succession

Section 2’s extension of the Act to dependants, heirs and successors makes the Tribunal a forum for the families of personnel to a degree unusual among service tribunals. The characteristic case is family pension. On the death of a serving member or a pensioner, entitlement passes through the categories the pension regulations prescribe: the widow or widower, children within the prescribed conditions, dependent parents in their turn, and the disabled child for life in the circumstances the regulations recognise. Disputes arise over the recording of the family in the service documents, over marriages disputed after death, over the division between a widow and the children of an earlier marriage, over the date from which family pension runs where the death was itself the subject of a court of inquiry, and over the characterisation of a death as attributable to service, which carries enhanced rates. Each of these is a Section 14 service matter, and the applicant is frequently a person who has never dealt with a court, litigating from a village against the record office’s file.

The Tribunal’s handling of this stream shows the merits jurisdiction at its most valuable. The record office’s view of its own documents is not the end of the matter: the Tribunal examines the service record, the nomination, the court of inquiry proceedings, the medical record in death cases, and decides entitlement on the whole material, with the administrative member’s knowledge of how service documents are made and maintained informing the weight given to each entry. Cases in this stream also generate a disproportionate share of the execution difficulty described earlier in this article, because the applicant’s need is immediate income and the sanctioning chain for family pension runs through several offices. An article on this Tribunal that confined itself to officers’ promotion disputes would misdescribe the institution; a large part of its moral weight sits in this stream, where the amounts are small, the applicants are unrepresented or assisted by the legal services machinery, and the order, when it comes, is the difference between a pension paid and a pension withheld on a point of record.

Sentence, forfeiture and the pension consequences of a court martial

A court martial sentence reaches further than the punishment it names, and the Tribunal’s appellate work cannot be understood without the consequences that travel with conviction. Dismissal from service carries the loss of pension and gratuity that would otherwise have accrued, subject to the competent authority’s power under the pension regulations to grant compassionate allowance in a proportion of cases, and a sentence of imprisonment interacts with service and pension reckoning in ways the regulations specify. Cashiering and dismissal differ in their incidents. Forfeiture of seniority and of service for promotion and pension are themselves punishments the courts martial may award, and each is appealable matter under Section 15 like the principal finding. The appellant before the Tribunal is therefore rarely contesting liberty alone: the appeal decides the family’s income for a generation, which is one reason the Tribunal’s power to substitute a lesser finding, to mitigate and commute, and to mould the consequential orders matters as much in practice as its power to quash a conviction outright.

The same reach explains the care of Section 15(4)’s drafting. The grounds are conviction focused: sustainability of the finding, error of law, material irregularity causing miscarriage. Sentence appeals travel with them, the Tribunal examining proportionality against the service’s own sentencing practice, the appellant’s record, and the parity between co accused where a single court martial tried several persons. The bail power in Section 15(3) belongs to this picture: a convicted person whose sentence is short may serve it entirely before an appeal is heard unless the Tribunal intervenes, and the power to release on conditions, withheld only where there is reasonable ground to believe the offence is one punishable with death or life imprisonment, is the mechanism that keeps the appeal real for the appellant in custody. Practitioners in this jurisdiction treat the bail application as a substantive part of the appeal, not a formality, and benches decide it on the appeal’s apparent grounds as well as on custody considerations, because the two are connected by the sentence’s clock.

What the Tribunal cannot do: a consolidated account

The boundaries of the jurisdiction are scattered through this article, and applicants are served by seeing them together. The Tribunal cannot entertain the excluded categories of Section 3(o): the specified pleasure doctrine orders, transfers and postings, leave of any kind, and summary courts martial below the dismissal and three month imprisonment threshold. It cannot take up grievances of personnel of the central armed police forces or of other uniformed services outside the three service Acts, whose members must use their own remedies. It cannot decide a matter that is not a service matter or a court martial appeal, however closely the applicant’s service is involved: a dispute over government accommodation after retirement in its purely civil character, a criminal prosecution in the ordinary courts for an offence the service Acts do not govern, a claim against a regimental fund or a service institute in its private capacity, all lie elsewhere. It cannot direct the framing of policy, as the Supreme Court confirmed in the Judge Advocate General promotion litigation in December 2023. It cannot execute its own orders through machinery of its own, having to rely on the respondent’s compliance pressed by successive orders and the contempt power. And it cannot be appealed from to a High Court, the channels upward being the filtered appeal to the Supreme Court and judicial review under Article 226.

These limits are not a counsel of despair; they are the map an adviser uses. A grievance outside Section 3(o) may still be a writ case of the narrow kind the High Courts entertain, or an internal remedy with real prospects, since a large share of service grievances are corrected on representation when the record is put right. A court martial grievance below the Tribunal’s summary threshold may still be petitioned under the service Acts. A policy grievance may be a challenge to the policy’s legality in the writ court rather than a service application. The statute’s exclusions channel cases; they do not end them, and the quality the Tribunal’s case law most rewards is the accurate identification, at the outset, of which channel a grievance belongs to.

The Srinagar bench and the litigation of a conflict region

One bench illustrates the jurisdiction’s reach into the country’s hardest service conditions. The bench serving Jammu and Kashmir and Ladakh, sitting at Srinagar with sittings at Jammu, hears the service litigation of a region in which the Army has been deployed in aid of civil authority and in counter insurgency roles for decades, and from which recruitment into the forces is heavy. Its docket carries the characteristic cases of such a deployment: disability claims arising from operations and from the conditions of high altitude and field service, pension claims of the families of personnel killed in action and the disputes over the characterisation of such deaths for the enhanced family pension categories, and service matters of the territorial and home and hearth formations recruited locally. The bench’s existence is the geography principle of Section 5 carried to its logical end: the litigants least able to travel, the families of the fallen and the disabled in a conflict region, are given a forum in their own region, staffed by a bench that includes members who know what the service conditions in the pleadings actually were.

The point generalises. A tribunal’s design is tested at its edges, in the cases of the applicants with the least money, the least education in law, and the most completely department held records. The Armed Forces Tribunal’s edges are the Srinagar family pension applicant, the Kochi disability pensioner, the Guwahati bench serving the North East’s recruits, and the dismissed soldier at Jaipur whose whole case is a summary proceeding. The statute’s inclusions reach all of them; its exclusions and its execution weakness fail some of them; and any honest account of the institution must hold both halves of that sentence together, which is what this article has attempted.

The making of the Act: from the Bill of 2005 to the statute

The legislative path explains several of the Act’s settled features. The Armed Forces Tribunal Bill 2005 was introduced in the Rajya Sabha in December 2005, carrying the Law Commission’s core design: a tribunal for service matters and court martial appeals, presided over by senior judicial figures, with benches across the country. The Parliamentary Standing Committee on Defence examined the Bill through 2006 and reported in its favour with observations that left marks on the final text. The Committee considered the exclusions from the service matter definition and accepted the exclusion of transfers and postings while recommending clear transfer policies in the services. It considered the position of leave, on which the Ministry of Defence pressed the apprehension of flood and operational burden, and the exclusion survived. It examined the appellate design and the proposal that the appeal lie to the Supreme Court on the Tribunal’s certificate, a design the personnel’s representatives had themselves sought in the consultations the Law Commission recorded, preferring one national appeal to the divergent rulings of many High Courts.

The Bill’s passage in December 2007 was not the product of party contest in any serious degree; military justice reform of this kind draws support across the House, and the debates concentrated on the welfare of personnel, the pensioner population the Tribunal would serve, and the speed with which benches would be established. The two year gap between assent in December 2007 and inauguration in August 2009, with commencement of the Act in June 2008 in between, records the practical labour of constituting a national tribunal: selection of the first Chairperson and members, notification of benches, framing of the Procedure Rules 2008, and the transfer machinery of Section 34. The first Chairperson was a retired Judge of the Supreme Court, and the Principal Bench’s early orders, transferring in the High Court dockets and settling the Tribunal’s practice on records and interim relief, set lines the institution has followed since.

Understanding the making matters for interpretation. When the Tribunal or a court construes the exclusions, the certificate filter or the bail power, the materials of the making, the 169th Report, the Committee’s report, the statements of the Defence Minister on introduction, are the record of what the design was for. The exclusions were debated choices. The filter was a requested feature, adopted with knowledge of its cost to individual appellants. The mixed bench was the Law Commission’s answer to a genuine dilemma, that a purely civilian forum might misread the service and a purely service forum could not be independent of it. Criticism of the Act is most useful when it engages those choices as choices, rather than treating the statute as an accident, and the same discipline applies to this article’s own verdicts.

Confidential reports and the evidentiary core of service litigation

Most service disputes are decided on documents the applicant has never seen in full. The confidential report, the annual assessment by reporting and reviewing officers through which an officer’s career is graded, sits at the centre of promotion, selection, premature retirement and extension litigation. The Tribunal’s practice in this territory is a study in how a merits forum handles a closed record. The reports are produced to the Tribunal by the respondent under its Section 17 powers, examined by the bench, and the applicant’s challenge is framed on the grounds service law permits: that an entry was adverse and uncommunicated where the rules required communication, that a reporting officer acted from bias or on material outside the period of report, that the profile the selection board considered contained an entry later expunged or a grading inconsistent with the record of the same period, that the pen picture and the numerical grading diverge in ways the service’s own moderation process should have caught.

The Tribunal does not publish the file to the applicant as a matter of course, and the balance it strikes, between the confidentiality on which candid reporting depends and the fairness an applicant needs to plead a case, is one of the working compromises of the jurisdiction. What the applicant receives is the substance necessary to answer: the adverse entry relied on, the board’s criteria, the comparative position stated in the terms the record allows. The bench’s administrative member, who has written and been written up in the same system, tests the department’s account of its own reporting practice against experience of how that practice runs in units and headquarters. It is in these cases, invisible in the statute’s text, that the tribunal form earns its keep over the writ court, which would have seen the same file only through the keyhole of judicial review, asking whether the process was lawful rather than whether the profile was the officer’s own.

Medical records perform the same function in the disability and invaliding streams, and service records in the pay and pension streams: the sheet roll, the service book, the record of enrolment and discharge, the pay accounts. A recurring shape of case is the record that contradicts itself, a disability recorded at release but not at enrolment, a date of enrolment differing between the sheet roll and the pay record, a casualty publication that does not match the court of inquiry. The Tribunal’s Section 17 requisition power and its willingness to decide between competing entries on the whole material, rather than accepting the entry the department prefers, is the everyday substance of its independence, exercised without drama in hundreds of orders that no appeal will ever disturb.

Representation, cost and the character of the Bar

The Tribunal’s Bar is a specialist one. Practice before the benches is concentrated among advocates who work in service law, many of them retired Judge Advocate General branch officers or the sons and daughters of service families, and the department is represented by panel counsel for the Union of India instructed by the service headquarters and the record offices. The applicant’s side is frequently unrepresented or represented at modest cost, and the Tribunal’s procedure, affidavit based, form driven, registry guided, is designed to be usable without elaborate representation, though the complexity of a court martial record or a selection board challenge makes representation valuable in the cases that matter most. Legal services assistance reaches qualifying applicants through the legal services authorities, and the services themselves maintain machinery for advising personnel on the preparation of statutory complaints and representations that precede litigation.

Costs in this jurisdiction are not generally the engine of litigation strategy that they are in commercial courts, and the Tribunal’s orders on costs are used principally to mark disapproval, of a departmental stand the record did not support, of delay in producing records, of non implementation that forced a further round. The modest economics of the forum matter to its design assessment: a merits tribunal for pensioners and dismissed personnel succeeds or fails on whether an applicant of ordinary means can use it, and the evidence of the dockets, in which unrepresented and legally aided applicants obtain orders against the Ministry in substantial numbers, is part of the answer to the capture thesis that the design’s critics must absorb. The institution’s failures, on the record assembled in this article, are failures of administration, appeal structure and execution, not of access or of adjudicative willingness, and reform that addressed the first three while preserving the forum’s accessibility would keep the settlement of 2007 and complete it.

The Judge Advocate General’s branch and the Tribunal

No account of this jurisdiction is complete without the legal branch of the services themselves. The Judge Advocate General’s branch in each service advises commanders on military law, supplies the Judge Advocates who guide courts martial on the law, conducts the legal review of court martial proceedings for the confirming authority, and advises the service headquarters on the litigation the Tribunal generates. The branch is therefore present on both sides of the Tribunal’s work: its officers shaped the record the Tribunal reviews, and its standards of advice and review are among the things the Tribunal’s appellate jurisdiction disciplines. A court martial whose Judge Advocate’s summing up misstated the burden of proof, or whose proceedings the branch’s review should have caught as irregular, is corrected on appeal under Section 15(4), and the correction feeds back into the branch’s training and practice.

The 2007 Act recognised the branch in the Tribunal’s own composition by requiring every Administrative Member to have served at least a year as Judge Advocate General. The qualification creates a small and defined pool: officers who have held flag rank and have led the legal branch. Appointments from that pool give the bench a member who has personally run the machinery under review, which is the design’s strength and its most debated feature in one. The strength is obvious in hearing, where the member can identify what a record should contain and where in the service’s process a step was missed. The debate is the appearance question in a particular form: a member drawn from the service’s legal leadership reviews decisions of a system he himself served. The answers the design gives are retirement before appointment, the judicial member’s presence on every bench, and appointment by the President after consultation with the Chief Justice of India. Whether those answers suffice is part of the larger independence question this article has already judged, but the composition itself, as the series record puts it, has not been the centre of the criticism, and the branch qualification is best understood as Parliament choosing expertise with safeguards over distance without knowledge.

The branch also stands behind the largest practical question in the appellate jurisdiction, the quality of the record. Courts martial are tried by officers whose profession is arms, and the record they produce, evidence recorded in summary, exhibits, the Judge Advocate’s advice, the court’s findings without reasons of the kind a civil court gives, is the whole material of the appeal. The Tribunal’s repeated insistence, in its orders, on complete proceedings, on the production of the original record rather than extracts, and on reasons where the Act requires them, has had the systemic effect of raising the standard of the record across the services, an effect that does not appear in any appellate statistic but is among the Tribunal’s real contributions to military justice. A reform whose appellate tier improves the trial tier’s habits has done more than correct individual cases, and the credit belongs in part to the branch, which receives the corrections and institutionalises them.

Interim relief: the order that decides the case before it is heard

In service litigation the interim order is often the substance. A posting is excluded from the jurisdiction, but an order of discharge, compulsory retirement, reversion or recovery operates on a date, and if it operates, the applicant litigates from outside the service, on a pension or without one, while the case proceeds. The Tribunal’s interim powers, exercised under its general authority over the applications before it and the Procedure Rules, therefore carry the weight of the final relief in a class of case the statute nowhere names: stay of discharge until the application is heard, stay of a recovery of alleged overpayment from a pensioner’s monthly pension, maintenance of the seniority position pending decision, and in the disability stream, interim directions that have the effect of continuing a pension the sanctioning authority proposed to stop. Benches grant such relief on the familiar triad of prima facie case, balance of convenience and irreparable injury, translated into service terms, where the injury is frequently the loss of the service status itself and cannot be undone by arrears if the applicant succeeds years later.

The appellate jurisdiction has its own interim centre of gravity in bail and suspension. Section 15(3) supplies bail in custody cases. Suspension of a sentence pending appeal, and the Tribunal’s power under Section 30(3)’s proviso pattern to keep an appellant at liberty where an appeal is intended, complete a set of tools by which the Tribunal controls the position while the appeal is pending. The discipline of these powers is symmetrical: the department may also seek stay of the Tribunal’s orders in the channels above it, and the spectacle of a favourable order stayed in the writ court or the Supreme Court while implementation waits is a further chapter of the execution story this article has told. The applicant’s advisers therefore plan the interim battle as carefully as the final one, because in this jurisdiction time is not neutral: it runs in favour of whichever side the interim order favours, and both sides know it.

The filter under challenge: the road to Parashotam Dass

The human cost of the leave filter, and of the Shri Kant Sharma years, is recorded in the proceedings that produced the correction. Among the matters decided with Parashotam Dass was a writ petition under Article 32, Gurbux Singh Dhindsa v Union of India, filed by the father of an Air Force officer killed in an operational area in Jammu and Kashmir. The Tribunal had granted relief, and the father’s remaining grievance concerned interest on what had been granted. A direct appeal to the Supreme Court was unavailable to him in substance, because the matter was personal to the litigant and raised no point of law of general public importance, and the High Courts were not entertaining petitions against the Tribunal in reliance on Shri Kant Sharma. The petition therefore challenged the vires of Sections 30 and 31 themselves. When the Court held that the writ jurisdiction under Article 226 is maintainable without per se restriction, the challenge to the sections was not pressed, and the matter was disposed of with liberty to approach the High Court. The sequence is the filter’s biography in miniature: a bereaved father’s interest claim, a small sum by the standards of constitutional litigation, was able to reach a constitutional court only by attacking the statute, because both ordinary channels, the filtered appeal and the writ as then understood, were closed to it.

The episode disciplines any assessment of the filter. Defended as a sorting device for points of general importance, it works; the One Rank One Pension class of question passes through it, as it should. Experienced by the individual litigant whose case is fact bound, it is a wall, and the wall stood higher between 2015 and 2023 because the supervisory channel that the Constitution keeps open had been read down by a judgment that a three judge bench later held to be in conflict with the Constitution Benches. The restoration in Parashotam Dass does not convert the writ into the appeal the statute withheld, and the individual litigant’s merits remain finally decided by the Tribunal, but it restores the correcting hand for error of law, jurisdiction and fairness, which is the minimum supervision the constitutional scheme promises any tribunal. An account of this Act written after March 2023 that described the High Court’s jurisdiction as excluded, or the filter as the only upward route, would be stating the law of the interval, not the law.

There is a further lesson in the manner of the correction. The Supreme Court did not strike down Sections 30 and 31, did not create a High Court appeal the statute omits, and did not disturb the certificate design. It corrected the reading of its own earlier judgment and remitted the individual matters to the High Courts for decision on merits. Reform of the filter, if it comes, must therefore come from Parliament, and the question for Parliament is the narrow one the Dhindsa petition posed and withdrew: whether a litigant with a grievance personal in character and small in amount should have, somewhere in the system, one full appeal on the merits beyond the Tribunal, or whether the Tribunal’s merits decision, supervised for legality, is a sufficient terminus. The personnel who asked the Law Commission for this forum in 1999 wanted finality close to the ground and expertise on the bench; they also wanted a remedy better than the writ. They received a forum that is better than the writ at first instance and, on the merits, final. Whether that exchange satisfies the promise of the reform is the question the next amendment of this Act will have to answer.

Who counts as subject to the service Acts

Section 2’s phrase, persons subject to the service Acts, carries more content than its brevity suggests, because each service Act defines its own subjection and the 2007 Act borrows the result. The Army Act applies to officers, junior commissioned officers, warrant officers and other enrolled persons of the regular Army, to the reserve forces when called up, to persons belonging to the Indian Reserve Forces and the Territorial Army when subject to the Act in the circumstances it specifies, to the Defence Security Corps and like formations governed by it, and, on active service, to persons not otherwise subject to military law who are employed by, in the service of, or accompanying the forces. The Navy Act and the Air Force Act draw their corresponding circles, the naval one including the reserve and volunteer reserve when embodied and the disciplinary court structure peculiar to the sea service. The Tribunal’s personal jurisdiction follows those circles exactly: a person’s route to the Tribunal is to show subjection to one of the three Acts at the material time, or retirement from such subjection for a service matter, or status as a dependant, heir or successor within Section 2.

The edges generate cases. Civilians employed by the services, including the large civilian workforce of the ordnance factories, the Military Engineer Services and the defence accounts and audit establishments, are not subject to the Army Act by employment alone, and their service disputes belong to the Central Administrative Tribunal and the writ courts, a boundary that surprises civilian employees who work beside uniformed colleagues on the same station. Personnel of the Assam Rifles occupy a debated margin, the force being administered by the Ministry of Home Affairs and officered substantially by the Army, with its own Act of 2006; its members’ grievances have required case by case attention to the statute governing the individual at the material time rather than a blanket rule. Cadets, recruits and persons under training are within the service Acts from enrolment or appointment as those Acts provide, and disputes over discharge during training have reached the Tribunal on that footing. The precision matters for the same reason the exclusions matter: the Tribunal’s threshold questions, subjection and subject matter, decide more cases than any question of merits, and they are decided on the statutes’ text rather than on the uniform the applicant wore.

Records, proof and the protection of sensitive material

Cases before the Tribunal are built on paper: the service record, the confidential reports, the medical board proceedings, the court martial transcript. Applicants and their families frequently hold copies of material that is sensitive in the plain sense, identity documents, medical records, service papers of deceased personnel, and the preservation and controlled sharing of that record is a practical part of preparing a case. The discipline of keeping sensitive documents in a controlled store, and disclosing them only to the advisers and forums that need them, is the same discipline described in the account of how journalists use a document vault to protect their sources, and it applies with equal force to a pensioner’s file: the record is the case, and the record must be kept complete, ordered and safe from the moment a grievance arises, because the Tribunal decides on the record the parties place before it and cannot reconstruct a file that has been lost.

Conclusion: a strong bench inside an unfinished structure

The Armed Forces Tribunal Act 2007, Act 55 of 2007, did the thing the Supreme Court had asked Parliament to do in 1982 and the Law Commission had designed in 1999. It gave persons subject to the Army Act 1950, the Navy Act 1957 and the Air Force Act 1950 a forum that decides service matters on their merits under Section 14, hears appeals from courts martial under Section 15 with powers to grant bail, substitute findings and reshape sentences, wields the powers of a civil court in the respects Section 17 specifies, and punishes contempt under Section 19. Its benches combine High Court judges with flag rank officers who have led the Judge Advocate General’s branch, a composition that has not been the target of the criticism the institution attracts.

The criticism attaches to the two features with which this article began. The Tribunal is administered by the Ministry of Defence, the respondent in most of its cases, and the transfer of that administration to the law ministry or to an independent commission has been recommended repeatedly and not effected. The appeal to the Supreme Court is filtered by Section 31, so that the Tribunal’s own certificate or the Supreme Court’s leave stands between the litigant and the only statutory appeal, and the High Court’s correcting hand operates through judicial review under Article 226 as restored to its constitutional scope by Parashotam Dass in 2023, not through an appeal. Execution of the Tribunal’s orders, declared final and executable by Section 29, depends on the respondent department’s compliance and has generated its own litigation. A serving or retired member who knows the inclusions and exclusions of Section 3(o), the grounds of a court martial appeal, the bail power, the leave filter and the writ position knows what this forum can and cannot do, which is the knowledge the statute’s structure makes necessary.

Frequently Asked Questions

Q: What does the Armed Forces Tribunal decide?

The Tribunal decides two classes of case. Under Section 14 it decides service matters of persons subject to the Army Act 1950, the Navy Act 1957 and the Air Force Act 1950, including pay, pension, promotion, seniority, tenure and termination, on the merits of the dispute. Under Section 15 it hears appeals against the orders, findings and sentences of courts martial. It decides both questions of law and questions of fact in service matters, which separates it from a writ court confined to reviewing the legality of a decision.

Q: Which matters are outside the tribunal’s jurisdiction?

Section 3(o) excludes four categories from the definition of service matters: orders issued under Section 18 of the Army Act, Section 15(1) of the Navy Act and Section 18 of the Air Force Act; transfers and postings, including a change of place or unit; leave of any kind; and summary court martial, except where the punishment is dismissal or imprisonment for more than three months. A grievance inside an excluded category cannot be brought before the Tribunal however it is framed, and the applicant must use the internal remedies of the service Acts or, in a narrow class of case, the writ jurisdiction.

Q: How is the Armed Forces Tribunal composed?

The Tribunal consists of a Chairperson and Judicial and Administrative Members, sitting in benches that ordinarily pair one of each. The Chairperson must be a retired Judge of the Supreme Court or a retired Chief Justice of a High Court. Judicial Members must be or have been Judges of a High Court. Administrative Members must have held the rank of Major General or the equivalent for at least three years in total and must have served at least one year as Judge Advocate General of a service. Appointments are made by the President after consultation with the Chief Justice of India.

Q: Can a court martial finding be appealed?

Yes. Section 15(2) entitles a person aggrieved by an order, decision, finding or sentence of a court martial to appeal to the Tribunal. Section 15(4) requires the Tribunal to allow the appeal where the finding is legally not sustainable, where it involves a wrong decision on a question of law, or where a material irregularity in the trial caused a miscarriage of justice. On allowing an appeal the Tribunal may substitute a finding of guilty of another offence the court martial could have returned, mitigate, remit or commute the sentence, or order a retrial on specified charges.

Q: Where does an appeal from the Armed Forces Tribunal go?

An appeal lies to the Supreme Court under Section 30, within ninety days of the final decision or order, and there is no appeal to any High Court. Section 31 conditions the appeal: it lies only if the Tribunal certifies that the case involves a point of law of general public importance, or the Supreme Court grants leave on being of the opinion that the point ought to be considered by it. An appeal against an order punishing for contempt is the exception, lying to the Supreme Court as of right within sixty days under Section 30(2).

Q: Can a High Court be approached instead?

A High Court cannot be approached instead of the Tribunal for a merits decision, because Section 14 gives the merits jurisdiction to the Tribunal. It can be approached under Article 226 of the Constitution, a jurisdiction Section 33 expressly preserves and which the Supreme Court in Parashotam Dass, decided in 2023, held is subject to no per se restriction in relation to the Tribunal’s orders. The writ court conducts judicial review rather than an appeal: it corrects jurisdictional error, error apparent and denial of natural justice, and it does not reappreciate evidence or substitute its own view of the service merits.

Q: Who administers the tribunal and why is that criticised?

The Tribunal is administered by the Ministry of Defence, which carries its budget and controls its administrative support, staff sanctions and infrastructure. The criticism is structural: the Ministry is the respondent in the great majority of proceedings before the Tribunal, so the body that decides against the department is administered by the department. Transfer of administrative control to the Ministry of Law and Justice has been recommended repeatedly, including in the tribunal reform debate following the Law Commission’s 272nd Report of 2017, and the Supreme Court’s tribunals jurisprudence treats ministerial administration of tribunals as a design defect. The transfer has not been effected.

Q: How have tribunal reforms affected the Armed Forces Tribunal?

The conditions of service of the Tribunal’s members have been drawn into the general tribunal reform litigation. The Finance Act 2017 empowered the Centre to prescribe qualifications and conditions for scheduled tribunals including this one, and the rules under it were struck down in Rojer Mathew. The Tribunals Reforms Act 2021 reset conditions for tribunal members generally, and the Supreme Court’s judgment of November 2025, reported at 2025 INSC 1330, struck down Sections 3 to 7 of that Act. The Tribunal’s own jurisdiction and composition under the 2007 Act were not abolished or restructured by that litigation.

Q: Who can approach the Armed Forces Tribunal?

Any person subject to the Army Act 1950, the Navy Act 1957 or the Air Force Act 1950 who is aggrieved by an order in a service matter may apply under Section 14. Section 2 extends the Act, for service matters, to retired personnel and to their dependants, heirs and successors, which is why pensioners and the widows and families of deceased personnel form so large a share of the applicants. A person aggrieved by a court martial order, finding or sentence may appeal under Section 15 whether serving or retired, since the conviction and its consequences survive retirement.

Q: What counts as a service matter under Section 3(o) of the Act?

Section 3(o) defines service matters as all matters relating to the conditions of service of persons subject to the three service Acts. The definition expressly includes remuneration including allowances, pension and other retirement benefits, and tenure in its many forms: commission, appointment, enrolment, probation, confirmation, seniority, training, promotion, reversion, premature retirement, superannuation, termination of service and penal deductions. It also includes summary disposal and trials where dismissal is awarded, and any other matter whatsoever, subject to the four exclusions for specified pleasure doctrine orders, transfers and postings, leave, and most summary courts martial.

Q: What can the Tribunal do with a court martial sentence when it allows an appeal?

Under Section 15(6) the Tribunal may substitute a finding of guilty of any other offence of which the court martial could lawfully have found the appellant guilty on the evidence, and pass a sentence for that offence. It may mitigate the punishment, remit it in whole or part, or commute it to a lesser punishment the service Acts permit, and within the safeguards the section prescribes it may enhance a sentence after hearing the appellant. It may also order a retrial by a court martial on the charges or on specified charges, and may make the consequential orders, including on custody, that the disposal requires.

Q: What is the time limit for an appeal to the Supreme Court from the Tribunal?

Section 30(1) requires an appeal to the Supreme Court to be preferred within ninety days of the Tribunal’s final decision or order. The period is subject to the Section 31 condition that the Tribunal certify a point of law of general public importance or the Supreme Court grant leave, and the application for the certificate or leave must be pursued within the discipline the Act and the Supreme Court’s practice prescribe. An appeal against a contempt order under Section 30(2) carries a shorter period of sixty days. No appeal lies against an interlocutory order of the Tribunal.

Q: Is an appeal against a contempt order of the Tribunal treated differently?

Yes. Section 30(1) excludes orders passed under Section 19, the contempt power, from the ordinary appeal route, because Section 30(2) gives a separate and stronger one: an appeal lies to the Supreme Court as of right from any order or decision of the Tribunal in the exercise of its jurisdiction to punish for contempt, to be filed within sixty days. Pending such an appeal the Supreme Court may suspend the punishment or release the appellant on bail, and the Tribunal itself may exercise those powers where the appellant shows an intention to appeal. No certificate under Section 31 is needed.

Q: Does the Tribunal cover personnel of the BSF, CRPF and other paramilitary forces?

No. Section 2 applies the Act to persons subject to the Army Act 1950, the Navy Act 1957 and the Air Force Act 1950, and to retired personnel of those forces and their dependants, heirs and successors for service matters. The central armed police forces, including the Border Security Force, the Central Reserve Police Force, the Central Industrial Security Force and the Indo Tibetan Border Police, are governed by their own service Acts and are outside the Act. Their personnel pursue service grievances through their own departmental remedies and the writ jurisdiction of the High Courts and the service jurisdiction of the Central Administrative Tribunal where it applies.

Q: What happened to service cases pending in the High Courts when the Tribunal was established?

Section 34 transferred the pending proceedings to the Tribunal. Every suit, appeal or other proceeding pending before a High Court or another court or authority immediately before the appointed day, which fell within the jurisdiction the Act confers, stood transferred to the Tribunal and was continued from the stage it had reached. The transfer is why the Tribunal began with a working docket of service and court martial matters from across the country, and it is also why records of older grievances are found in the Tribunal’s registries rather than in the High Courts where the litigation began.

Q: How does the Tribunal approach disability pension claims?

A disability pension claim turns on the medical board’s assessment of the disability, its percentage and its characterisation as attributable to or aggravated by service. The Tribunal reviews that characterisation on the record under its Section 14 jurisdiction, applying the approach in Dharamvir Singh v Union of India, (2013) 7 SCC 316: the member is presumed sound on enrolment, the board’s opinion is respected but not conclusive, and disabilities arising in service are approached liberally on attribution. The Tribunal examines whether the board gave reasons and addressed the service conditions recorded, rather than conducting a fresh medical examination of its own.

Q: Who appoints the Chairperson and Members of the Tribunal?

Section 7 provides that the Chairperson and the Members are appointed by the President of India. No appointment may be made except after consultation with the Chief Justice of India. The President may appoint one or more Members as Vice Chairpersons. The qualifications are prescribed by Section 6: the Chairperson must be a retired Judge of the Supreme Court or a retired Chief Justice of a High Court, Judicial Members must be or have been High Court Judges, and Administrative Members must combine the prescribed senior rank with at least one year as a Judge Advocate General.

Q: Why does the Tribunal include Administrative Members drawn from the services?

The design follows the Law Commission’s 169th Report of 1999, which reasoned that a purely civilian appellate body might not be conducive to the discipline of the armed forces and that service questions require knowledge of military organisation. An Administrative Member brings to the bench direct experience of command, of how selection boards, confidential reports and medical boards operate, and of the Judge Advocate General’s branch, in which every Administrative Member must have served for at least a year. Paired with a Judicial Member on each bench, the arrangement is meant to combine adjudicative method with institutional knowledge of the force whose disputes are being decided.

Q: What powers of a civil court does the Tribunal possess?

Section 17 gives the Tribunal the powers of a civil court under the Code of Civil Procedure 1908 in specified respects while discharging its functions: summoning and enforcing the attendance of witnesses and examining them on oath, requiring discovery and production of documents, receiving evidence on affidavit, requisitioning public records from any court or office, issuing commissions for examining witnesses or documents, reviewing its decisions, dismissing applications for default or deciding them ex parte, and setting aside orders of those kinds. The document and record powers matter most in practice, because the decisive service records are held by the respondent department.

Q: How long do the Chairperson and Members of the Tribunal serve?

Section 8 fixes a term of four years from the date on which the Chairperson or a Member enters upon office, and the holder is eligible for reappointment. Age ceilings qualify the term: a Chairperson who has been a Judge of the Supreme Court may not hold office after attaining seventy years, a Chairperson who has been a Chief Justice of a High Court may not hold office after sixty five years, and no other Member may hold office after attaining sixty five years. Resignation is by writing addressed to the President under Section 9, and removal requires proved misbehaviour or incapacity after an inquiry by a Judge of the Supreme Court.