Service Law Research: CAT, Tribunals, and the Government Employee's Case

Service law is the practice area where fact patterns repeat most exactly. Two employees in different departments, in different States, two decades apart, and the dispute is the same: a promotion given to a junior, a regularisation refused after fifteen years of daily wage work, a punishment of removal for a lapse the employee says warranted a censure. That repetition is what makes service research tractable. It is also why ordinary keyword searching fails at it more often than in any other field, because the words differ and the situation does not.

The second defining feature is that service law is rule law. The principle you are arguing about is usually settled. What decides the matter is a recruitment rule, a seniority rule, a pay fixation order, a departmental circular or a scheme, and those documents differ by department and by State. An authority interpreting one set of rules is not automatically transferable to another set, however similar the grievance.

Put those two features together and you get the discipline of the practice: find the closest factual match, on the closest rule, from the forum whose view binds you. This piece sets out how the forums are arranged, what recurs, and a method that works.

The Forum Map: Who Hears a Service Matter

The Administrative Tribunals Act 1985 created the Central Administrative Tribunal, which hears service disputes of central government employees and of employees of such other authorities as have been brought within its ambit. It functions through benches located across the country, and the bench that hears a matter is determined by where the applicant was posted or where the cause of action arose.

The same Act contemplates State administrative tribunals, and several States have constituted them for their own employees. Where a State has no such tribunal, or where one was constituted and later abolished, service matters go to the High Court directly under Article 226. That is why a practitioner in one State works almost entirely before a tribunal and a practitioner in a neighbouring State works almost entirely in writ.

Crucially, the tribunal is not the final word. In L. Chandra Kumar (1997) the Supreme Court held that the jurisdiction of the High Courts under Articles 226 and 227 over the decisions of tribunals could not be excluded, and that such decisions are subject to scrutiny before a Division Bench of the High Court to which the tribunal is territorially subordinate. The practical effect is a stack: tribunal, then a Division Bench of the High Court, then the Supreme Court. Research has to be aimed at the right level of that stack, because a tribunal decision that reads beautifully may already have been set aside above.

What Recurs, and Why That Matters

A small number of categories account for the overwhelming majority of service litigation. Recognising which one you are in tells you immediately what the controlling doctrine is and what facts you will have to establish.

Two doctrines that run through every category

Two doctrines cut across all of these and deserve special attention because they are argued in almost every matter. The first is parity: employees similarly situated should be treated alike, which is the engine of a great many service claims and also the point at which they most often fail, because the department shows that the comparator was not similarly situated after all. The second is delay and laches. Service claims have no uniform limitation, but a stale claim is not revived by a later judgment obtained by someone else, and a claim slept on for years attracts a hard question about why. Research the delay point before your opponent does.

Why Keyword Search Fails Here More Than Anywhere

Service judgments describe the same situation in unstable vocabulary. One order speaks of ad hoc service, another of officiating service, a third of stop gap arrangement, and all three may be about the same question of whether that service counts for seniority. One speaks of regularisation, another of absorption, another of conferment of temporary status. One speaks of proportionality of punishment, another of the penalty being shockingly disproportionate, another simply of interference with the quantum.

At the same time the department names differ, the post designations differ, the rules are cited by different short titles, and the employee's grievance is described in whatever words counsel used in that matter. A keyword search rewards you only when your words match the bench's words, and in this field they routinely do not.

This is why service law is, of all Indian litigation, the strongest case for searching by situation rather than by phrase. What you want to retrieve is not documents containing a string. It is decisions about an employee in materially the same position as yours: same category of appointment, same nature of grievance, same stage of service, ideally the same or a materially identical rule. That is a similarity question, and it is answered by describing the fact pattern in a sentence or two rather than by guessing the vocabulary.

• Ad hoc, officiating, stop gap, temporary and provisional are all used, sometimes interchangeably and sometimes with real distinctions, for service that may or may not count towards seniority.
• Regularisation, absorption, conferment of temporary status and screening describe overlapping but not identical routes out of irregular employment.
• Disproportionate, shockingly disproportionate, excessive and not commensurate all appear in orders interfering with the quantum of punishment.
• Seniority, inter se position, gradation list and panel are used for the same underlying dispute about who stands where.
• Compassionate appointment appears alongside appointment on compassionate grounds and dependant employment, depending on the scheme's own vocabulary.

Disciplinary Matters: A Narrow Window, Researched Precisely

Disciplinary matters deserve separate treatment because the scope of interference is narrower than in any other category, and because that narrowness is routinely misunderstood by the employee whose case it is.

A tribunal or court exercising judicial review of a disciplinary proceeding is not sitting in appeal over the finding of misconduct. It does not reweigh the evidence or substitute its own view of what the witnesses established. What it examines is whether the process was fair and whether the conclusion is one that could reasonably be reached: whether the charge was clearly framed, whether the employee was given the documents and the opportunity to cross examine, whether the inquiry officer's report was furnished before the punishment was imposed, whether the punishing authority applied its mind, and whether there is any evidence at all to support the finding.

Quantum of punishment is a separate and equally narrow enquiry. Interference is available where the penalty is so disproportionate to the misconduct as to shock the conscience, and the reported authority on what has and has not crossed that line is extensive and highly fact specific. This is precisely the material where a close factual match is worth more than a statement of principle: a decision reducing removal to compulsory retirement for a comparable lapse by a comparable employee is a far stronger submission than another citation on proportionality.

The research method follows the structure of the enquiry. Build a procedural chronology of the inquiry from the record, checkpoint by checkpoint, and identify every departure. Then research each departure separately, because the consequences differ: some breaches vitiate the proceeding, while others require the employee to show prejudice. Only then research quantum, and research it on facts rather than on doctrine.

A Method: From Rule to Match to Authority

In service law the winning authority is rarely the most famous one. It is the one where an employee in the same position, under the same rule, was granted the relief you are asking for.

Where CourtMesh Fits, and Where It Does Not

Service research is the clearest use case for retrieval by meaning rather than by phrase. CourtMesh runs one unified search across the Supreme Court, all 25 High Courts, District Courts and Tribunals, sourced from official government portals, with roughly 310 million records keyword-searchable and roughly 2 million covered by semantic retrieval. That semantic layer is what lets you describe the situation, an employee regularised after long daily wage service against a post that was never sanctioned, and retrieve decisions that never use the words you typed.

Related-case discovery is useful in the same way, surfacing matters that resemble the one you are reading rather than matters that share its keywords. Filters for court, year, acts and sections, and disposition then narrow the field to the forum that binds you. Citation relationships help you see whether an order has been followed or upset, across the subset of the corpus that carries deep analysis, which is expanding and is not the whole of it.