JJF v JJG [2026] SGECT 9

Preface

For many reasons, the law governing dismissal disputes in Singapore is, as I suggested in JGP v JGQ [2026] SGECT 1 (“JGP”), “a little more layered, difficult, and unclear than it perhaps needed to be” (at [3]). On reflection, however, I did not in JGP always make explicit just how beset by troubles this area of law is, or how those troubles have tangibly affected the ability of both employees and employers to plan for, respond to, and move on from an employer-initiated termination of an employment relationship.

Whether one takes a moral view of law, under which legality itself requires legal rules to be sufficiently clear to be followed (see, eg, Lon L Fuller, The Morality of Law (Yale University Press, Rev Ed, 1969) at pp 33–39 and 63–65; cf Nigel Simmonds, Law as a Moral Idea (Oxford University Press, 2007) at pp 159 and 190), or a more pragmatic view, under which law must at least be capable of guiding conduct if it is to perform its basic social function (see, eg, Joseph Raz, “The Rule of Law and Its Virtue” in The Authority of Law: Essays on Law and Morality (Oxford University Press, 2nd Ed, 2009) at pp 210 and 213–214), the conclusion is the same: it is plainly undesirable for the law of dismissal to be so uncertain that both sides may struggle to know—both before and after the event—what their rights and obligations are.

Accordingly, it is with considerable reluctance that I take up these questions again, for once the inquiry is opened, it proves very difficult to contain. What begins with the meaning of “without just cause or excuse” soon leads into why those words were enacted, the conception of industrial justice they originally expressed, the means by which dismissals were then investigated and remedied, and the consequences of transferring that responsibility to an adjudicative tribunal. I will therefore go quite some way further than I did in JGP to make those troubles explicit. The most significant of them are traceable to four related and fundamental failures in how the Employment Claims Tribunals (“ECT”) came, in 2018, to be vested with jurisdiction to adjudicate what the Employment Claims Act 2016 (“ECA”) calls “wrongful dismissal disputes” (see Employment (Amendment) Act 2018 (“2018 Amendment Act”)). Explaining those failures requires a return to the history from which the present law was inherited: to see the concerns it was meant to meet, and to understand the difficulty produced by the survival of old words after the socioeconomic conditions, assumptions, and conception of industrial justice which once animated their application have all but fallen away. Although such an exercise rarely needs to feature in reasons for an adjudicative decision, it is necessary here because, without it, the legal question posed by this case cannot be stated, still less answered, with the candour and clarity it requires. I begin, however, with what happened in this case.

Background

The claimant (“Mr C”) was dismissed pursuant to an express contractual termination clause by the payment of salary in lieu of notice. No reason was given, whether in the dismissal letter or at the meeting where the decision was communicated. Mr C then brought a claim against the respondent (“RPL”), his former employer, under s 14(2) of the Employment Act 1968 (“EA”). That provision permits a relevant employee who considers that they have been “dismissed without just cause or excuse” to bring a claim in the ECT. Because “dismiss” includes terminations “with or without notice and for cause or otherwise” (s 2(1) of the EA), the contractually neutral circumstances of Mr C’s dismissal did not prevent him from bringing that claim.

That claim came before me for trial on 9 January 2026. I heard evidence from Mr C on one side and RPL’s two witnesses on the other—its Deputy General Manager (Human Resources) (“Mr RR”) (who was also RPL’s representative at the trial), and one of its Senior Project Managers (“Mr RW”) (who was also Mr C’s supervisor for a time). On 26 January, I found that Mr C had been dismissed “without just cause or excuse” within the meaning of s 14(2) and awarded him $11,437.50 in compensation. In May, RPL obtained leave to appeal under s 23 of the ECA on the basis that my decision gives rise to a question of law. I now give the grounds of my decision in full.

The undisputed facts

Mr C commenced employment with RPL as a Senior Construction Manager on 8 July 2024 at a monthly salary of $9,000. During his three-month probation, he was assigned to a construction project for a statutory board (the “SB Project”). RPL confirmed him in employment on 8 October 2024, paid him an annual bonus of $10,011 in December 2024, and increased his basic salary to $9,150 with effect from 1 January 2025.

Mr C continued working on the SB Project during the first half of 2025. His responsibilities included overseeing site operations, coordinating with supervisors and subcontractors, and attending the project’s regular meetings. In June 2025, he took medical leave. He attended a clinic on 3 and 10 June, received two days’ sick leave on each occasion, and was then referred to a psychiatrist. A doctor’s memorandum issued on 11 June stated that he had been diagnosed with major depression and anxiety, “likely precipitated by long working hours and insomnia”. A further memorandum dated 30 June stated that he was fit to work but recommended shorter hours to prevent a depressive relapse.

With effect from 1 July 2025, Mr C was moved off the SB Project and assigned to RPL’s tender team at its headquarters (“HQ”) to assist with a tender project which I will call “Project G”. He worked regular hours of 8.30am to 5.15pm and assisted with developing plans for reinforced concrete works, preparing a master schedule, and compiling construction methodology and tender documents with a senior engineer. The circumstances and basis of that move later formed the heart of the parties’ dispute.

Mr C remained with the tender team through July and August 2025. Project G’s tender proposal was submitted on 18 August. On 19 August, RPL issued Mr C a letter titled “Cessation of Employment”. It referred only to cl 2.10(b) of RPL’s “Employment Regulations”, which had been incorporated into his employment contract (the “Termination Clause”), and stated that he would receive one week’s salary in lieu of notice and, “as a gesture of goodwill”, his full salary for August. His employment ended on 20 August.

The parties’ cases and the problems therewith

This section has three tasks. I first state Mr C’s case, including the discrimination allegation and the relief he sought. I next set out RPL’s contractual and factual defence. I then identify the legal confusion shared by both positions.

The claimant’s case

Mr C’s case was that RPL’s compliance with the Termination Clause did not answer his statutory claim. He alleged that his dismissal was “without just cause or excuse” because it was directly linked to his disclosure in June 2025 of a mental health condition which RPL’s excessive working demands had precipitated. On his account, RPL discriminated against him because of that condition: its lack of support or accommodation led to his removal from the SB Project, his redeployment to an ad hoc role at HQ, and his dismissal immediately after Project G’s tender submission, in circumstances he regarded as retaliatory.

Mr C’s case rested on what were, in substance, five interlocking strands.

(a) First, the employment relationship before June 2025 disclosed no settled dissatisfaction with his work: RPL had confirmed him, paid him a bonus, and increased his salary.

(b) Second, he relied on his account of excessive working hours and the psychiatrist’s memoranda described at [7] above. He also invited me to draw an adverse inference because RPL took immediate possession of his company-issued laptop at the termination meeting; according to him, it contained attendance and work records capable of corroborating those hours.

(c) Third, he said that he disclosed his diagnosis on 13 June 2025 but remained able to work. The shift in RPL’s attitude thereafter—its instruction that he report to HQ and its later removal of him from the SB Project—suggested that it had formed an adverse view of him because of his condition, rather than any independent and pre-existing concern about his performance.

(d) Fourth, he characterised the move to HQ as the beginning of the end, rather than a bona fide accommodation. He relied on the lack of meaningful consultation, his removal from the operational role for which he had been employed, the absence of any structured accommodation notwithstanding his psychiatrist’s recommendation, and what he described as scrutiny of his medical leave and “administrative isolation” at HQ.

(e) Finally, he relied on the timing and manner of his dismissal: it followed immediately after he had assisted with Project G’s time-critical tender; there had been no warning, counselling, written performance feedback, or performance-management process; RPL had, he alleged, begun arranging a replacement before the termination meeting; and the dismissal letter gave no reason beyond the contractual clause. He inferred that RPL had withheld its real reason because that reason was discrimination, and relied on para 9 of the Tripartite Guidelines on Wrongful Dismissal (Govt Gazette, Electronic Ed, 23 July 2024) (“Tripartite Guidelines” or “Guidelines”) to submit that any presumption against wrongfulness arising from notice could not survive proof of such a reason.

Mr C sought $18,000 in compensation, which he described as a conservative sum kept within the ECT’s monetary jurisdiction. He said it reflected both the financial consequences of losing his employment and the harm caused by the allegedly discriminatory and retaliatory manner and motive of his dismissal.

Mr C quantified that sum principally by reference to an anticipated period of unemployment, contending that his dismissal in the final quarter of the year made it harder to obtain work. At trial, he said that he had made 47 job applications, obtained only two interviews—around mid-September and late December—and remained unemployed for several months. His full loss, including salary, Central Provident Fund contributions, and bonus, would in his estimation exceed $18,000, but he confined the claim to that sum as a compromise within the ECT’s jurisdiction rather than accounting for every asserted head of loss.

Mr C also sought two forms of non-monetary relief, though he identified no legal basis for either. The first was a written apology “from the Managing Director of [RPL], stating that the dismissal was wrongful and discriminatory, and expressing regret for the harm caused”, which he regarded as “a necessary non-financial vindication of [his] rights”. The second was “[a] mutually agreed upon statement that the termination was without prejudice … and [that] no negative reasons for dismissal (eg, poor performance) [would be] disclosed to future employers”.

The respondent’s defence

RPL’s defence had three limbs:

(a) Mr C had been dismissed pursuant to an express contractual right, with salary in lieu and without any stated reason; under the Guidelines, such a dismissal was presumed not to be wrongful unless he substantiated a wrongful reason.

(b) In any event, the surrounding evidence disclosed business, suitability, and operational considerations, rather than the mental-health discrimination or retaliation alleged.

(c) Even if the dismissal were wrongful, Mr C had received more than his contractual entitlement, his monetary claims were speculative or legally irrecoverable, and the ECT could not compel the non-monetary relief sought.

On the first limb, RPL relied on the Termination Clause, which allowed either party to terminate the employment contract by giving one week’s notice or paying salary in lieu once Mr C had served 26 weeks but less than two years. RPL had paid him a full month’s salary for August although, on its case, he was entitled only to one week’s salary in lieu from 19 August 2025.

RPL therefore characterised the dismissal as “in accordance with the law” and “not considered wrongful”. It also relied on: (a) s 11(1) of the EA, which permits termination without notice by payment of salary in lieu; (b) Dong Wei v Shell Eastern Trading (Pte) Ltd and another [2022] 1 SLR 1318 (“Dong Wei”) at [21], where the Appellate Division stated that “employers may terminate employees without cause, in accordance with an express right of termination, so long as sufficient written notice or pay in lieu of notice is given”; and (c) paras 3 and 7 of the Tripartite Guidelines, which recognise the parties’ contractual right to terminate their relationship and presume dismissals with notice not to be wrongful.

RPL emphasised that, as Mr C acknowledged, it had given no reason for the dismissal, whether orally or in writing. Relying on Illustration 3 in the Guidelines, it submitted that the dismissal could not be wrongful unless Mr C established the discrimination which he alleged. The illustration states:

The second limb was that the surrounding circumstances did not support discrimination. RPL said that its decision had been driven by business, suitability, and operational considerations, while its conduct after Mr C disclosed his condition was inconsistent with the motive alleged.

RPL first contended that it had concerns about Mr C’s behaviour, attitude, performance, fit, and suitability before his disclosure, and had conveyed them in the ordinary course of employment. His confirmation, bonus, and increment did not prevent concerns from arising or persisting thereafter, nor did they undermine RPL’s eventual assessment that he could not meet the expectations of the roles assigned to him.

Through Mr RW, RPL identified two matters which had arisen on the SB Project. First, on or around 25 or 26 April 2025, Mr C allegedly abandoned lean concrete casting works because of a supply shortage, failed to contact an alternative supplier or ensure the concrete met the required level, and left the site before completion without clear follow-up instructions. Second, the SB Project’s Resident Engineer had reportedly told Mr RW that Mr C lacked initiative, could not lead the site team or engage effectively on technical issues, and was unable to conduct daily subcontractor coordination meetings as expected of a Senior Construction Manager. RPL said that Mr RW conveyed these concerns at a lunch on 23 May 2025, where he also reminded Mr C to attend key meetings because his absence could explain why he lacked information needed to perform his work. RPL saw no significant improvement thereafter.

RPL next denied that Mr C’s disclosure in June 2025 caused adverse treatment. After he took medical leave on 3, 4, 10, and 11 June, RPL convened a meeting on 12 June, attended by its Senior Project Director and Mr RW, to understand the operational implications of his absence and make contingency plans for the time and resource-intensive construction work. RPL said that Mr C first disclosed his condition at that meeting, which took it by surprise, and it asked for a doctor’s memorandum so that it could respond appropriately. After Mr C submitted the memorandum on 13 June, Mr RR arranged an engagement session on 16 June with RPL’s General Manager (Engineering). Mr C was told to rest and encouraged to consider counselling through RPL’s Employee Assistance Programme. RPL did not question the diagnosis, honoured his sick leave from 16 to 30 June, and did not press him to return early. When Mr C returned on 1 July with the second memorandum recommending shorter hours, RPL’s Managing Director met him and assigned him to Project G at HQ, where he worked from 8.30am to 5.15pm and was not instructed to work beyond those hours. RPL relied on these steps as contradicting his allegation that it had failed to support or accommodate him.

RPL’s third proposition was that the move to HQ did not cure its concerns, which were reinforced by Mr C’s work on Project G. It alleged that:

(a) His plans for reinforced concrete works, prepared between 5 and 31 July 2025, were too granular for a tender, departed from instructions, adopted an impractical sequence and inadequate resource allocation, and were not used in the tender submission.

(b) His master schedule, prepared between 21 July and 10 August, was unusable and inconsistent with the expected timeframe despite a baseline programme, at least three rounds of feedback, and time for revision.

(c) His contribution to gathering methodology and compiling tender documents during the same period was negligible beside the work of RPL’s Project Manager (“Mr PM”) and Senior Engineer (“Mr SE”).

RPL further alleged that Mr PM and its Technical Director repeatedly gave feedback without visible improvement. It also relied on missed deadlines and occasions when Mr C raised his voice at Mr PM and insisted on his own approach, which it regarded as showing a lack of professionalism and difficulty accepting senior guidance.

On this basis, RPL denied that Mr C’s mental health condition had caused his dismissal. It was careful, however, to say that the performance and attitude evidence was advanced only to rebut his allegation and “dispel any notion” that discrimination supplied the only possible explanation. Its position remained that business, suitability, and operational considerations underlay its decision to part ways, although the dismissal itself had been effected under the contractual clause and “not for any particular reason”.

RPL accordingly submitted that Mr C’s claim failed because he had not substantiated a wrongful reason of the kind required by paras 9 and 10 of the Tripartite Guidelines:

The third limb of RPL’s defence concerned relief. Since Mr C had received his full contractual entitlement and, indeed, a full month’s salary instead of one week’s salary in lieu, RPL submitted that no further payment was due. RPL relied on the common law measure of damages for wrongful dismissal: subject to mitigation, the sum which the employee would have received had the contract been lawfully terminated by notice or salary in lieu thereof. In Wee Kim San Lawrence Bernard v Robinson & Co (Singapore) Pte Ltd [2014] 4 SLR 357 (“Wee Kim San”) at [25] and [38], the Court of Appeal observed, in substance, that it would be inconsistent with an agreed right to terminate on notice to award damages extending beyond the notice-pay entitlement. RPL further submitted that Dong Wei at [20]–[22] reaffirmed that reasoning and showed that an employee who had received full salary and the agreed payment in lieu suffered no further loss in law.

In any case, RPL also answered Mr C’s asserted bases for compensation:

(a) His claim that work was harder to obtain in the final quarter of 2025 was speculative, and he had produced no evidence of any job application.

(b) His claim for “stress, anxiety, and depression” was legally irrecoverable because mental distress does not ordinarily constitute actionable contractual damage, even where a dismissal was allegedly humiliating (Arul Chandran v William J Gartshore and others [2000] 1 SLR(R) 436 at [13]–[15]). An employment contract was not within the limited class of contracts whose object was to provide peace of mind or freedom from distress.

(c) His alleged reputational and career damage was inconsistent with a dismissal letter which stated no adverse reason, and was unsupported by evidence that the dismissal had prejudiced any job application.

As for the non-monetary relief, RPL submitted that there was no legal basis to compel it to provide a written apology of the kind sought even if wrongful dismissal was made out.

The problems with the parties’ understanding of the law

Strictly speaking, only four dispositive issues arose from the parties’ cases. The first concerned the burden of proof under s 27(2)(b) of the ECA: did Mr C have to show that his dismissal fell afoul of s 14(2) of the EA, or did RPL have to show that it did not? Section 27(2)(b) states:

The second issue was whether, applying the appropriate burden, Mr C had proven his positive case of discrimination on grounds of his mental health condition. The third issue was whether—irrespective of my decision on the second issue—it could nevertheless be said that Mr C had been “wrongfully dismissed” in the sense that he was dismissed “without just cause or excuse”, as the phrase has been used in s 14(2). The fourth issue was how Mr C’s “wrongful dismissal” (if established, whether in the course of the second or third issues) was to be remedied.

The parties’ cases may seem legally unremarkable, and these four issues straightforward. Indeed, as framed, the cases were mostly cogent—the claimant’s case had a central thesis premised upon the Guidelines to which the ECT ought to have regard, and the respondent’s case would probably accord—at least broadly—with a notional employment lawyer’s understanding of the phrase “wrongful dismissal”, albeit recast slightly by the Guidelines. Similarly, the four questions which arose would seem to be ordinary questions of burden, fact, application, and remedy. However, as one might have inferred from my preface at [1]⁠–⁠[3] above and the title of this subsection, that impression is only possible because I have temporarily compressed the major troubles and fault lines which exist within this area of law into the simple language of those issues. Unfortunately, the unassuming facts of this case—and how the parties advanced their positions—brought many of those fault lines into view.

The principal reason is this. Mr C’s claim plainly rested on s 14(2) of the EA; this much should have been clear to both him and RPL, if not from anything else, then at least from the Claim Referral Certificate that was issued. That provision supplied both the source and shape of the right Mr C invoked: the right to seek redress where an employee has been dismissed “without just cause or excuse”. The other sources on which the parties relied—the Termination Clause, s 11(1) of the EA, the Guidelines, and the common law authorities on wrongful dismissal—therefore had to be arranged around that statutory inquiry. However, the parties failed to do so and, as a consequence, their cases became confused because those sources were invoked by rote without a clear understanding of how they related to each other and to s 14(2).

RPL’s defence was a particularly clear illustration of this. RPL began with the contract. Since the Termination Clause—or, alternatively, s 11(1) of the EA—permitted termination by notice or salary in lieu, and since Mr C had been paid more than the minimum sum due in lieu, the dismissal was said to be “in accordance with the law” and “not considered wrongful”. RPL then moved to the Guidelines. Since this was a dismissal with notice and no reason was given, RPL then said that in order to establish that the dismissal was wrongful, Mr C was required to substantiate a positively wrongful reason, such as discrimination. RPL then relied on evidence of performance, suitability, attitude, and operational concerns; but only, it said, to rebut Mr C’s allegation that the true reason was his mental health condition, and not because those matters had been the reason for dismissal. Finally, in respect of remedy, RPL turned to the common law and primarily argued that payment of notice or salary in lieu exhausted any recoverable loss.

The structure of RPL’s defence was wholly unstable and betrayed a fundamental misunderstanding of several things, but chief amongst them was how primary legislation interacts with contract and non-binding administrative guidance. It treated contractual validity as substantially answering a distinct statutory claim; the Guidelines as capable of confining that claim to positively wrongful reasons despite the words of s 14(2); performance and suitability as material to rebut discrimination but not as the asserted “just cause or excuse”; and the common law measure of damages as governing compensation for the statutory wrong. The result was conceptually untidy and, more importantly, obscured the question which s 14(2) required me to answer.

Mr C’s case reflected the same difficulty from the opposite direction. Because the Guidelines speak of an employee needing to substantiate a wrongful reason where dismissal with notice is effected without stated reasons, he concentrated on proving discrimination. Section 14(2), however, is framed by the absence of a “just cause or excuse”, which points to the lack of a sufficient basis for dismissal and is not confined to proof of an impermissible motive. Of course, if Mr C could prove that RPL had in fact positively dismissed him on discriminatory grounds, that would almost certainly have been enough to establish that the dismissal was “without just cause or excuse”. But it did not follow that, if discrimination could not be proved, the dismissal was ipso facto supported by just cause or excuse.

The same point may be put more formally to avoid any suggestion that I have compressed the logic. Let D denote the proposition that RPL dismissed Mr C on discriminatory grounds, and W the proposition that Mr C was dismissed without just cause or excuse. Proof of D would almost certainly establish W: D→W. However, it does not follow from a failure to establish D that W is false. That would be to reason: D→W, ¬D, ∴¬W. But that is a non sequitur which commits the familiar formal fallacy of denying the antecedent (see, eg, Irving M Copi et al, Introduction to Logic (Routledge, 15th Ed, 2019) at pp 308–309; for a work written for those in the legal profession, see Douglas Lind, “Basic Categories of Argumentation in Legal Reasoning” in Handbook for Judicial Officers (Judicial Commission of New South Wales, 2021) at pp 462–463). D→W establishes that D is sufficient for W, not that it is necessary. Put more formally: (D→W), ¬D ⊬ ¬W. The contrary reasoning treats that implication as the biconditional D↔W and wrongly excludes every other route to a finding that the dismissal was without just cause or excuse.

To be fair to the parties, the legal materials invite some of this confusion. The ECA calls a claim arising from s 14(2) a “wrongful dismissal dispute”, although “wrongful dismissal” already had an established meaning at common law: an action concerned with breach of the employment contract—ordinarily, a failure to give proper notice or salary in lieu, or a summary dismissal without adequate contractual grounds (see Ravi Chandran, Employment Law in Singapore (LexisNexis, 6th Ed, 2019) (“Chandran (Lexis)”) at paras 6.91–6.92 and 6.160–6.168). As I said in JGP at [3], it is somewhat unfortunate that this polyseme was created, and this case makes especially clear why. Once that label entered the ECA, contractual termination clauses and ss 10(1) and 11(1) of the EA naturally appeared significant. After all, if “wrongful dismissal” turns on notice, and both the contract and statute permit termination by either notice or salary in lieu, it is easy to suppose that giving or paying notice answers any claim bearing that same name.

The Guidelines then add a further layer of confusion. As a starting point, they too use the language of “wrongful dismissal” even though—by their own description—they are only concerned with dismissals impugnable under the EA (see para 2), and have nothing to do with the common law action. Having used an established common law label for what is obviously a different statutory inquiry, they then go on to reinforce the apparent importance of notice by stating that dismissals with notice are “presumed not to be wrongful”. Yet, they also suggest that such a dismissal—even one effected with full notice and without any reason being stated—may nevertheless be “wrongful” if the employee substantiates a wrongful reason. That, as I just explained, does not correspond to the language of s 14(2), which asks whether the employee was dismissed “without just cause or excuse”.

However, neither is it the common law conception of wrongful dismissal, which does not ordinarily make an employer’s reason for terminating with notice or salary in lieu a source of liability (see, eg, Latham Scott v Credit Suisse First Boston [2000] 2 SLR(R) 30 at [44]; Dong Wei at [83]–[93]; Deborah Barker with KhattarWong, A Guide to Termination of Employment in Singapore (LexisNexis, 2nd Ed, 2008) (“Barker”) at pp 3–4). As the point was so vividly put in D’Cruz v Seafield Amalgamated Rubber Co Ltd [1963] 1 MLJ 154 (“D’Cruz”), “where a contract provides for the services of an employee to be terminated on a month’s notice, the employer can dismiss the servant by giving him a month’s notice without stating any reasons for doing so, without having any reason for doing so or indeed for the most disreputable and wicked reasons” [emphasis added] (at 156D).

The resulting legal position is, by any measure, confusing. It was therefore unsurprising—and, indeed, somewhat understandable—that, without clear guidance as to how these mismatched pieces fit together, the parties’ cases slid between contractual termination, common law “wrongful dismissal”, the Guidelines, and s 14(2). Nevertheless, whether understandable or not, the parties’ cases were still premised on error: one that lay in their failure to appreciate the specific legal question posed by s 14(2) and how the sources and concepts they referenced bore on that question.

I would also add here that the propensity to make this error is not confined to those without legal training. In another earlier case that came before me, a paralegal brought a claim under s 14(2) against his former employer, a reputable law firm, which was represented by one of its partners. Like RPL, the partner submitted that “[t]he Claimant ha[d] been paid more than he [was] entitled to under the Employment Contract”. Accordingly, it was said, there was “simply no further basis for any claim against the Respondent”. This submission proceeded as though the contractual position exhausted the statutory inquiry when it clearly did not. A different but also conceptual error was made by the respondent’s representatives in JGP—its in-house counsel—who submitted that, under the Guidelines, the ECT’s task under s 14(2) was simply to decide whether the stated reason for dismissal was false, and nothing more (see JGP at [47], [72]–[74], [A.88]). That approach treated the formulation in the Guidelines as exhaustive of the inquiry without any regard for the legislative text of s 14(2). That was methodologically surprising, to say the least, since one would assume the basic precept that primary legislation stands above administrative guidance to be well ingrained in every legal professional’s mind. I am not, to be clear, singling out these cases to criticise those who made these submissions. The point is simply to illustrate that the confusions which I just described are capable of affecting legally trained persons no less than laypersons.

For present purposes, the consequence of these confusions is simple. At this point, I cannot yet go on to explain how I resolved the four dispositive issues which arose in this case. Before I can do so properly, I first need to clarify what the law is, what it is not, and where the legal goalposts have been placed. Otherwise, my analysis would do little more than reproduce the same confusions which had tainted the parties’ cases.

A roadmap for the sections that follow

Near the end of these admittedly lengthy grounds, I will ultimately return to the four issues which were actually dispositive of this dispute, as identified at [31]⁠–⁠[32] above. Before then, however, for the reasons given, I must explain the major troubles which beset the present law, how those troubles arose, and how far the law can be reconstructed so as to minimise them without obscuring what the enacted scheme has become.

I will begin by eliciting the three major troubles which beset the law on s 14(2) claims. This will be an ordinary legal-analytic exercise and the troubles I intend to bring forth concern: (a) the juridical foundations of s 14(2); (b) the proof of its breach; and (c) the consequences which follow. I first examine each trouble individually. Thereafter, I will also need to analyse them together because the difficulties in respect of proof and remedies—once they are more fully stated—require us to re-examine the bearing they have, if any, upon the juridical character of the protection conferred by s 14(2). This re-examination is needed because issues of proof considerably affect the circumstances in which a breach of s 14(2) can practically be established, and—even if jurisprudentially orthodox—there may be a limit to how far one can maintain the view that the practical enforceability of a right does not bear on its normative basis. A similar point may be made in respect of the latter because the ordinary remedial response to an infringed legal right tends to expose the interest protected by that right and, therefore, what its normative concern is.

From there, I then turn to the history, concepts, and institutional arrangements which produced those troubles. That account begins with the inherited settlement from which s 14(2) emerged and proceeds through: (a) the faded conception of industrial justice which once gave that settlement content; (b) the difficulty created when the words chosen to express that settlement—including those of s 14(2)—remained unchanged despite structural changes to the statutory scheme in which they operated; (c) the further confusion introduced by the polysemous label “wrongful dismissal”; and (d) the consequences of transferring these disputes to the ECT, where a protection developed and administered within an industrial-relations framework became judicialised. Through that history, I seek to bring into view the four failures alluded to at [3] above, and to explain how the law came to assume its confusing, unstable hybrid form, anecdotally evidenced by how the parties in this case—especially RPL—misunderstood the law governing s 14(2) claims.

Against that background, I will then attempt to reconstruct the law on s 14(2) claims in the light most favourable to its coherence and clarity. This will entail an examination of whether the juridical foundations of s 14(2) and its remedial framework can be regarded as preserved despite the practical alterations worked by the surrounding rules concerning proof, procedure, and relief, and whether those alterations can be understood as reflecting a considered balance. I will suggest that they can be, once ordinary techniques of legal analysis are applied to isolate the exact protection conferred by s 14(2) and, consequently, the sequence of inquiry which s 14(2) now requires the ECT to undertake when an employee falling within its scope initiates a claim.

The reconstruction that I offer will sit in some tension with both the history and the plain words of s 14(2), which collectively point towards the provision both (a) embodying a more robust protection, and (b) requiring a different sequence of inquiry from that which my reconstruction will suggest. Nonetheless, I am of the view that my reconstruction better accords with contemporary understandings of the employment relationship and security of employment, and tracks the direction in which Singapore’s labour and wider socioeconomic policies have developed since the ideas underlying the protection first began to find legal expression in the 1960s.

In substance, the tension arises because language bearing the imprint of an older conception of the protection has been carried forward into a materially different setting, in which the institution charged with its enforcement, prevailing sensibilities concerning labour and security of employment, and wider socioeconomic conditions and policies have all changed. Though that tension could probably have been avoided—or at least appreciably reduced—had Parliament more directly adapted the statutory language of s 14(2) to fit how it seemed to conceive of the modern protection, that tension is not so pronounced as to compel the conclusion that the operation now intended is unworkably mismatched with the statutory form in which it has been expressed.

Even on that reconstruction, however, a number of lesser troubles remain, and I will examine them in five groups thereafter.

(a) The first concerns institutional drift—that is, the drifting of dismissal claims out of their proper forum.

(b) The second concerns institutional fragmentation—that is, the splitting of the statutory and contractual claims arising from the same dismissal between different adjudicative routes.

(c) The third concerns statutory overlap—that is, how ss 14(2) and 84(2) of the EA relate to ss 35(2), 35(3) and 82 of the Industrial Relations Act 1960 (“IRA”).

(d) The fourth concerns process before dismissal—that is, the legal significance, if any, of the process followed by an employer before dismissing an employee.

(e) The fifth concerns the meaning of “dismiss”—that is, the reach of the definition in s 2(1) of the EA, including its application beyond s 14(2) and its relationship with forced resignation and constructive dismissal.

Because my analysis is both wide-ranging and comprises many strands, its detail and length may obscure the overall position. At the end, therefore, I will weave them back together in a comprehensive restatement of the law governing dismissal at large. That restatement will treat the contractual and statutory protections as separate but connected. For each protection, it will identify the institution in which such protection may be invoked, its personal and material scope, the procedure and proof required to engage it, and the relief which may follow. It will also explain the available routes where more than one protection is invoked.

Following that—and specifically because the ECT is a high-volume forum in which parties appear without counsel—I will reduce my restatement into a simplified account of the matters to which employees and employers must attend when bringing or answering a dismissal claim. After all, if the law is difficult, it is of little use to state it only in abstract propositions. Some guidance should therefore be articulated as to how employees and employers should think through the claims they bring, resist, or seek to resolve.

Finally, I return to the four dispositive issues in this case: who bore the burden of proof; whether Mr C established discrimination; whether he was nevertheless dismissed without just cause or excuse; and how that dismissal was to be remedied. A brief conclusion will follow and, thereafter, I add a postscript which draws my analysis and the resulting legal position into a more manageable summary.

I am acutely mindful that an exposition of this breadth may be thought atypical of a subordinate judicial body, especially one unaided by arguments from counsel, and all the more so given how much I have already said in JGP. I have therefore hesitated over whether it is proper to embark upon it. Nevertheless, three points led me to decide to do so:

(a) First, as I will show, the present law admits of several reasonable views, each producing a materially different application of s 14(2), so that how the provision is applied, and sometimes whether a claim succeeds, may depend—to an uncomfortable degree—not only on the views of the parties, but on the view taken by the individual tribunal. For a statutory protection administered principally among litigants in person, the need to reduce that uncertainty is sufficiently strong to warrant the breadth and atypicality of what follows.

(b) Second, I am wholly cognisant that, ordinarily, a tribunal should decide no more than the case requires and should be slow to develop questions which the parties have not fully explored. However, the statutory setting of the ECT creates an unusual difficulty. Parties are not permitted to be represented by counsel (see s 19(1)(b) of the ECA), so the absence of developed legal argument is a feature inherent in the forum. If a tribunal which has perceived a difficulty through actual cases and accumulated experience were nevertheless to say nothing because litigants in person had not fully articulated it, the result would be an unbreakable loop. The parties would seldom be equipped to expose the difficulty without counsel, while their lack of counsel would then supply the reason for leaving it unexposed.

(c) Third, the possibility of appeal does not solve that problem. Appeals from the ECT lie only on grounds involving a question of law or jurisdiction (see s 23(1) of the ECA). A legal difficulty left submerged in the tribunal’s reasons may never be recognised, still less formulated, as such a ground, and may therefore recur from case to case without reaching an appellate court in a form capable of correction, clarification, or development.

For those reasons, I consider that the difficulties which I have perceived in the law must be ventilated—and ventilated fully—despite the reluctance expressed at [3], and although much of what follows will necessarily be obiter. As to the breadth of that ventilation, it might appear from the roadmap set out above that—even on my own terms—I will be venturing far beyond that which is necessary. That is probably true but, as I also said at [3], once the inquiry is opened, it proves very difficult to contain. The reason is that the difficulties are not discrete. Each rests upon, and in turn exposes, anterior questions about the juridical source and content of the protection, its relationship with cognate provisions, and the institutions through which it is administered. To stop at any earlier point would therefore be to offer an account of the immediate difficulty while leaving unexamined the premises from which it arose and upon which any satisfactory answer must depend.

Eliciting the major troubles with the law

I said at [33] above that the four dispositive issues appeared simple because I had compressed within them several “major troubles”: difficulties that go to the very root of the protection conferred by s 14(2). One might fairly wonder if this exaggerates the depth at which the troubles exist—after all, how can an area of law such as this be so fundamentally troubled? In the course of this section, I will of course demonstrate why I have characterised the troubles as I have. However, before that, it is also useful to make two prefatory points: (a) as I suggested at [3] above, the major troubles only really arose because of the 2018 Amendment Act—so, they are not longstanding; and (b) for how fundamental they are, the major troubles are surprisingly well-hidden.

Thus far, the discussion at [34]⁠–⁠[43] has exposed the immediate consequence of the major troubles—namely, the parties’ unprincipled movement between contractual termination, common law “wrongful dismissal”, the Guidelines, and s 14(2). However, it has not revealed their full extent. That requires us to uncompress the four dispositive issues, which leads us to three prior and fundamental questions: first, what are the juridical foundations of the protection conferred by s 14(2) (“First Fundamental Question”); second, who must prove the facts needed to engage or resist the engagement of the protection (“Second Fundamental Question”); and third, if the protection has been engaged, what is the remedy which follows (“Third Fundamental Question”). Though these may seem to be questions of an elementary character, one will find that it is not easy to furnish clear, satisfactory, and principled answers to them.

For clarity, I should also note early that I have used the word “major” to contrast these with the “lesser” troubles mentioned at [48] above that are not quite so foundational. The lesser troubles assume some account of the protection and concern its operation in particular settings. The major troubles are logically prior. Unless they are resolved, one cannot state with confidence what s 14(2) even protects against. That is not to say that the lesser troubles are not difficult and potentially decisive in individual cases. However, they can generally be addressed by ordinary legal-analytic methods once the relevant premises are settled. The major troubles place those premises themselves in doubt and therefore call for the inquiry into statutory language, history, structure, and institutional setting I will now undertake.

As to juridical foundations

Before seeking to elicit the major trouble underlying the First Fundamental Question, I must be clear about what is meant by “juridical foundations”. By “juridical foundations”, I mean the legal basis or bases from which the protection derives its binding force and content. Though these are related, they are not identical. Binding force concerns why the protection has the authority of law and the sources from which it draws such authority. Content concerns what the protection requires or prohibits, though, naturally, such content depends in large measure on the sources from which the protection draws its binding force.

To be clear, I do not draw this distinction just for technical precision. A source may be legally relevant without every proposition contained in it being binding; conversely, the existence of a binding legal rule does not mean that its content can be ascertained without reference to other materials which may not, by themselves, have the force of law. Unless those possibilities are kept separate, an obligation to consider a source may imperceptibly become an obligation to apply it, and an accessible explanation of the law may come to be treated as the law itself. As will become apparent, that is substantially what has occurred in relation to the Tripartite Guidelines.

Binding force

I begin with the sources from which the protection conferred by s 14(2) of the EA acquires its force of law. The obvious starting point is s 14(2) itself since the protection only exists by virtue of that provision. Whatever meaning is ultimately attributed to the expression “without just cause or excuse”, the expression has legal authority because it forms part of the enacted text. The ECA supplies the adjudicative machinery through which the protection is now enforced and imposes several associated obligations upon the ECT. Nothing about that surrounding scheme, however, relocates the source of the protection itself. Thus, its binding force remains statutory.

There are, however, two other possible sources which require consideration. The first is the common law. That possibility arises principally because both the statutory protection and the contractual action are commonly described as concerning “wrongful dismissal”. From that shared label, it might be thought that the statutory protection derives some part of its binding force from the common law rules governing termination of the contract of employment, or that the statutory expression is somehow confined by those rules.

That suggestion is weak. The law is no stranger to polysemy, and the use of the same expression in two legal settings does not, by itself, establish that the expression performs the same juridical function in both. On the contrary, one should be cautious about making legal consequences depend upon verbal classifications rather than the rules and reasons which those classifications are intended to express (see, eg, Glanville Williams, “Language and the Law—I” (1945) 61(1) Law Quarterly Review 71 at 74). Indeed, words and legal labels commonly bear several related senses, and the existence of a shared label cannot alone resolve which sense is engaged in a particular legal setting (see, eg, Daniel J Hemel, “Polysemy and the Law” (2023) 76(4) Vanderbilt Law Review 1067 (“Polysemy and the Law”) at 1067–1071 and 1103–1104). Therefore, the mere fact of the common label “wrongful dismissal” may, at most, alert one to a possible historical or even conceptual relationship between the two regimes. However, it cannot bear the much greater conclusion that the protection obtains its legal force—or even part of such force—from the common law, or that the contractual and statutory inquiries are juridically related, much less identical. Those conclusions have to be proved, not assumed.

I would go so far as to say that any failure to appreciate this point is, in my view, a significant one. In systems derived from the English tradition, we are taught early—particularly in our schooling in equity—to understand the law’s deeply conceptual nature, and to cope with the fact that distinct legal concepts can travel under the same label; for example, the different senses of “assignment”, “rescission”, “estoppel”, “set off”, and so on. It would, as such, be a fundamental failure of legal learning for one to allow the nomenclature tail—without more—to wag the conceptual dog.

Next, I turn to the second possible source from which the protection might derive its binding force: the Tripartite Guidelines. The position in relation to the Guidelines is more difficult. Unlike the common law label, the Guidelines have been assigned an express role by ss 20(7)(a) and 35(2) of the ECA. The question is, as such, not whether they possess any legal significance. They plainly do. The question is what kind of significance Parliament has assigned to them and, in particular, whether their contents possess binding force independently of the words enacted in s 14(2), or at least, have the legal capacity to affect how the words in s 14(2) ought to be understood.

The starting point is that the Guidelines plainly do not have the force of primary legislation. This much should be clear from Arts 38 and 58(1) of the Constitution of the Republic of Singapore (2020 Rev Ed). Article 38 vests Singapore’s legislative power in the Legislature, consisting of the President and Parliament, while Art 58(1) provides that the Legislature exercises its law-making power through Bills passed by Parliament and assented to by the President. Calnan usefully explains the significance of that formality in the following terms (Richard Calnan, Principles of Statutory Interpretation (Oxford University Press, 2023) (“Calnan”)):

Of course, I do not mean that every legally binding rule must appear in primary legislation. It is possible for the Guidelines to have binding force independently of s 14(2) by other means. For example, the Legislature may confer powers to make subsidiary legislation, and the common law has binding force of its own. Obviously, the Guidelines have nothing to do with the common law. As to whether the Guidelines might have the force of subsidiary legislation, the answer is also clearly ‘no’. The Guidelines only exist by virtue of the Minister’s exercise of power under s 35 of the ECA, and this provision very plainly does not state that their contents are to have binding effect, much less that they are to be conferred the status of subsidiary legislation.

That, then, only leaves the question as to whether the Guidelines at least have the legal capacity to affect how the words in s 14(2) ought to be understood. To answer this, we need to begin with s 35(1) of the ECA, which empowers the Minister to issue guidelines on “what constitutes wrongful dismissal” in the form of tripartite guidelines. Section 35(2) provides that, upon their publication in the Gazette, “regard may be had” to them for the purposes of ss 20(7) and 25(4). Section 20(7)(a) then provides that, when deciding a claim involving a “wrongful dismissal dispute” (as defined in s 2(1) of the ECA), the ECT “is to have regard” to the Guidelines. Section 25(4)(a) materially repeats that direction in relation to appeals. It bears noting that the difference in phrasing between s 35(2) (“may”) on the one hand and ss 20(7)(a) and 25(4)(a) (“is to”) is not immaterial and, at the very least, produces some doubt as to whether it was intended by the Legislature that the ECT and High Court’s “regard” to the Guidelines be mandatory or discretionary.

Nevertheless, taking the Guidelines’ effect at their highest, it can be said that the ECT and High Court must “have regard” to them by virtue of ss 20(7)(a) and 25(4)(a), putting aside the inconsistent language in s 35(2). On that basis, and assuming that “regard” bears its natural meaning—“attention or heed paid to something as having an effect or influence” (Oxford English Dictionary, “regard (n), sense I.1.b”)—the direction can be understood as requiring genuine consideration of the Guidelines as material capable of influencing the decision. However, because “regard to” the Guidelines plainly does not entail the application of every proposition contained in them, there is doubt about the extent to which the Guidelines can be used as a legitimate aid to interpreting s 14(2), especially where the account in the Guidelines cannot be reconciled with the actual words used in s 14(2). Put simply, at the very highest, it is only mandatory for the ECT to consider the Guidelines. Given the limited direction (“have regard to”) enacted in s 20(7)(a), it is not mandatory for the ECT to apply the Guidelines. That, in turn, leaves open how exactly the ECT’s consideration of the Guidelines is supposed to fit in with principles which must be applied in the ordinary exercise of statutory interpretation.

To answer that, we must look at the Interpretation Act 1965 (Act 10 of 1965) (“Interpretation Act”). Section 9A of the Interpretation Act requires an interpretation which promotes the purpose or object underlying the written law, and regulates the consideration of extrinsic materials capable of assisting that inquiry. Whether the Guidelines are extrinsic materials capable of assisting that inquiry depends on s 9A(3). In this regard, because of the direction given in ss 20(7)(a) and 35(2), it can at least be argued that the Guidelines fall within s 9A(3)(f) which reads, “the material that may be considered … in the interpretation of a provision of a written law includes — (f) any document that is declared by the written law to be a relevant document for the purposes of this section”.

However, even if we accept that the Guidelines have a legal basis—whether through ss 20(7)(a) and 35(2) of the ECA or by virtue of s 9A(3)(f) of the Interpretation Act—from which they can derive the capacity to affect how s 14(2) is interpreted, it is unclear exactly how such capacity was meant to sit alongside the ordinary principles of statutory interpretation and the extent to which such capacity to bear on the meaning of s 14(2) may actually translate into bearing on the meaning of s 14(2). After all, ss 20(7)(a) and 35(2) simply use the open-ended “have regard to”, and s 9A(3)(f)—unsurprisingly—tells us little more than that since it merely makes clear that materials falling within its ambit “may be considered” in purposive interpretation.

At this time, it is important that I make clear that I am not suggesting that the ECT is not able to carry out the ordinary exercise of statutory interpretation without clearer directions as to how the Guidelines feature in that exercise. A judicial tribunal is amply equipped to interpret a statutory provision with administrative guidance in mind should such consideration be required by the applicable legislative framework. The problem, as I alluded to at [38] above and which I will develop more fully from [90]⁠–⁠[118] below, is that the Guidelines contain propositions and statements which are inconsistent with the plain text of s 14(2), and which seem to embody a dismissal-related protection that is distinct in conception from that which the text of s 14(2) (and its history: see [250]⁠–⁠[351] below) suggests s 14(2) actually embodies.

Yet, in tension with that, when one reads the debates during which Parliament considered the Guidelines, it can be inferred with some confidence that it was intended that they would bear quite substantially on how the ECT understands s 14(2) and adjudicates claims filed thereunder. Granted, the Parliamentary debates do not disclose a single—and certainly not a technically precise—account of the Guidelines’ juridical status vis-à-vis s 14(2), but they do at least disclose a sufficiently consistent expectation about the practical work they were to perform as an interpretive aid.

When moving the Bill, the Minister explained that the Ministry of Manpower (“MOM”) had—over the years—accumulated a considerable body of “wrongful dismissal” cases which put into practice the “broad principles” used by MOM and reflected the consensus reached between employers and unions. Because those cases had not been published, the transfer of adjudication to the ECT would be accompanied by the publication of guidelines containing illustrations of what did and did not constitute “wrongful dismissal”. The ECT, in turn, would have to consider the principles and parameters contained in them. Thus, on the Minister’s account, the Guidelines were to make explicit and carry into the ECT an approach which had previously been applied by MOM. They would make that approach publicly accessible and furnish the new adjudicatory body with the standard framework for deciding claims (Singapore Parliamentary Debates, Official Report (20 November 2018) vol 94 (“2018 Debates”) at 3.56pm (Mrs Josephine Teo, Minister for Manpower)).

Similar expectations may be discerned from some Members’ speeches:

(a) Mr Patrick Tay, who declared that he had been a member of the tripartite workgroup involved in the amendments, thanked the tripartite partners for agreeing to prepare guidelines and FAQs “to provide clarity on what amounts to wrongful dismissal”, and said that the resulting transparency would protect employees while ensuring that employers were not saddled with frivolous allegations (2018 Debates at 4.18pm (Mr Patrick Tay Teck Guan)).

(b) Mr Louis Ng approached the matter from the same direction. Having observed that the absence of a clear definition might render the protection ineffective, he asked whether “clear definitions” of dismissal without just cause or excuse could be supplied through subsidiary legislation or the intended guidelines, so as to clarify employees’ rights and assist the ECT in adjudicating claims. Though Mr Ng did not suggest that subsidiary legislation and guidelines possessed the same juridical status, that he offered them as alternative means of supplying the requested clarification is nevertheless revealing because it shows that he expected the guidelines to have the effect of clarifying the substantive reach of the protection (2018 Debates at 6.09pm (Mr Louis Ng Kok Kwang)).

(c) Assoc Prof Daniel Goh identified the underlying uncertainty of the Guidelines more directly. He asked what legal status the Guidelines would have, whether the ECT would be obliged to remain within their limits, how the ECT should otherwise take guidance from them, and whether it could depart from them in particular cases. The question on legal status did not receive a direct answer from the Minister (see [77] below). Unfortunately, that leaves the juridical character of the Guidelines and the degree of their influence unresolved. Nevertheless, for present purposes, one can see that each of the possibilities contemplated by Assoc Prof Goh seemed to assume that the Guidelines would influence the adjudication of disputes in the ECT in a substantial way (2018 Debates at 5.40pm (Assoc Prof Daniel Goh Pei Siong)).

The Minister’s response to these comments and questions confirmed the weighty practical expectation being placed on the Guidelines, though without resolving its juridical basis. She first answered the question of what constituted dismissal without just cause or excuse by identifying discrimination, denial of statutory benefits, and an intention to punish employees for exercising their rights as grounds which rendered a dismissal “wrongful”, while stating that dismissal for poor performance, misconduct or redundancy would not be considered “wrongful”. She then explained that it was impossible to define every scenario and that the “more feasible approach” was to use illustrations and set out principles and parameters which the ECT “must take into account when adjudicating cases” [emphasis added]. Since that explanation was given directly in answer to questions concerning both the meaning of the words used in s 14(2) and the interaction between the Guidelines and ECT adjudication, it presented the Guidelines as a means by which greater specificity would be supplied to the general words used in s 14(2). The Minister did not, however, say whether the ECT could depart from the Guidelines and, if not, on what legal basis an Executive-issued document could achieve that binding effect (2018 Debates at 7.03pm (Mrs Josephine Teo, Minister for Manpower)).

It is clear that, even taken together, these materials do not support the attribution to Parliament of a settled or juridically complete account of the Guidelines’ legal status. The Guidelines had not yet been issued and were not before the House, and the question of whether the ECT could depart from them received no direct answer. They do, however, support the narrower inference that the Minister and the Members who addressed the Guidelines were operating on the assumption that the Guidelines would exert substantial influence upon the interpretation of s 14(2) and the adjudication of disputes arising therefrom: that they would make public and systematise MOM’s earlier principles, furnish non-exhaustive illustrations and parameters, clarify what constituted “wrongful dismissal”, and shape the ECT’s determination of claims.

All that invites the question: can the Guidelines legitimately achieve such effect where they—as I suggested at [73] above—are inconsistent with the text of s 14(2) which, I should emphasise, Parliament decided to leave substantively unaltered when it enacted the 2018 Amendment Act. The answer, in my view, is quite plainly ‘no’, and I have three reasons.

First, given what I said at [67] above, it should be clear that—to the extent the Guidelines are inconsistent with the text of s 14(2)—they cannot be allowed to have the effect of indirectly amending primary legislation. Indeed, Parliament had the opportunity to amend s 14(2) in 2018 if it wished to recast the protection conferred by that provision in terms similar to that subsequently published in the Guidelines. However, it did not do so and, seeing as how it did not, it should be taken that such absence of an amendment affecting the substantive content of s 14(2) reflects its intention to maintain the content of such protection specifically in the terms used in s 14(2) (“without just cause or excuse”). After all, it may be taken that intended changes to the law—especially to matters as important as the substantive content of an employment right—would be “deliberately and carefully worked out”. We should therefore assume that Parliament would not have intended to alter the substantive content of s 14(2) “by a mere sidewind” (Diggory Bailey & Luke Norbury, Bennion, Bailey and Norbury on Statutory Interpretation (LexisNexis, 9th Ed, 2026) (“Bennion”) at para 26.7; also see David Lowe & Charlie Potter, Understanding Legislation: A Practical Guide to Statutory Interpretation (Hart Publishing, 2018) (“Lowe & Potter”) at para 5.6).

Second, the Guidelines cannot sensibly prevail where there exists an inconsistency between them and the text of s 14(2). As I said at [43] above, it is a basic precept of our constitutional order that primary legislation stands above administrative guidance. Indeed, the extent to which such guidance may exert influence in respect of the interpretation and application of primary legislation itself depends on the force given to said guidance by the primary legislation. Here, that is ss 20(7)(a) and 35(2), which—at their highest (cf [70] above)—only require the ECT to “have regard” to the Guidelines when adjudicating a s 14(2) claim. The limited framing of “have regard” is not enough to cause the Guidelines to prevail over s 14(2) where inconsistencies exist.

That brings me to my third and final reason. I have placed emphasis on the limited framing of ss 20(7)(a) and 35(2) for two purposes.

(a) One, there exists the sound principle of statutory interpretation articulated by the Court of Appeal in AnAn Group (Singapore) Pte Ltd v VTB Bank (Public Joint Stock Company) [2022] 1 SLR 771 (“AnAn”) at [76]: that it would be incorrect, absent good reason, to use the contents of subsidiary legislation to inform the meaning of primary legislation. Similarly, applying AnAn, the High Court in Alka v Public Prosecutor [2024] 5 SLR 1071 (“Alka”) at [18] held that subsidiary legislation cannot modify or delineate a definition used in primary legislation in the absence of a specific empowering provision. If this principle applies to subsidiary legislation, it would apply with even greater force to the Guidelines. The principle is supported both by authority and by the constitutional structure. Unless primary legislation expressly provides otherwise, an Executive body empowered to make subsidiary legislation—or, still less, to promulgate guidelines—cannot use that instrument to effectively recast primary legislation which only Parliament may amend (see, again, [67] above).

(b) Two, unless the primary legislation conferring interpretive force on the Guidelines made abundantly clear that the ECT’s duty to “have regard to” the Guidelines takes priority over the premise upon which the exercise of statutory interpretation is generally understood to proceed, it is highly doubtful that a judicial body such as the ECT may disregard its duty to interpret statutory provisions in accordance with the established principles. Those principles do not permit the ECT to simply have regard to the Guidelines alone in interpreting what s 14(2) means. On the contrary, statutory interpretation draws upon a settled body of techniques: the enacted words are read as a whole, in context and in light of their purpose, while recognised rules, presumptions, canons, and admissible external aids are brought to bear (Bennion at paras 10.1–10.2 and 10.4–10.6; Lowe & Potter at paras 3.2–3.10). Each of those techniques serves the same inquiry into the objective intention of the Legislature as expressed in the words it enacted; none can turn the Executive’s understanding of the law into an independent source of statutory meaning (see Craies on Legislation (Daniel Greenberg ed) (Sweet & Maxwell, 13th Ed, 2025) (“Craies”) at paras 16-003, 16-005, and 17-001; Calnan at paras 1.11–1.15 and 2.11–2.16). This point is, in fact, made directly in Craies: that, though government departments may offer guidance on the meaning and application of legislation, that guidance is “of no authority whatsoever”, because it is Parliament’s intention in passing the legislation, rather than the Government’s intention in promoting it, that stands to be discerned (at para 16-011). Therefore, unless ss 20(7)(a) and 35(2) had conferred on the Guidelines greater force—specifically, enough to override the basic premise of statutory interpretation—using the Guidelines as though they reflect the settled meaning of s 14(2) would invert that exercise, effectively making the Executive’s account the source of the protection’s content and Parliament’s words in s 14(2) something to be made to fit it.

Therefore, absent provisions in the primary legislation making very clear that the Guidelines are to have such potent interpretive significance and, therefore, bring to bear the effect which the Minister and Members seemed to think they would have (see [75]⁠–⁠[78] above)—which I do not think ss 20(7)(a) and 35(2) achieve—the Guidelines cannot be accorded greater weight than their legal status permits. In fact, that distinction appears within ss 20(7) and 25(4) themselves. In each provision, para (a) directs the adjudicator to “have regard to” the Guidelines, whereas para (b) directs that any compensation claimed be calculated “in accordance with” regulations made under s 34(1) (ie, the Employment Claims Regulations 2017 (“EC Regs”)). The Legislature thus used the language of consideration for the Guidelines and the language of compliance for subsidiary legislation. That contrast is difficult to explain if “have regard to” was intended to require the Guidelines to be accepted, still less to permit them to control the meaning of s 14(2). Simply put, the Guidelines cannot lead the interpretation of s 14(2). Their consideration must instead take place within the ordinary exercise of statutory interpretation.

The structure of that exercise is settled. As the apex court explained in Tan Cheng Bock v Attorney-General [2017] 2 SLR 850 (“Tan Cheng Bock”), primacy must be accorded to the text of the provision read in its statutory context, because the enacted text is what Parliament chose to embody and give effect to its purposes (at [43]). Where the ordinary meaning thereby ascertained is clear, extrinsic material may confirm but cannot alter that meaning. Where the provision is ambiguous or obscure, or its ordinary meaning produces a manifestly absurd or unreasonable result, such material may assist in ascertaining its meaning; even then, however, purposive interpretation cannot be used to rewrite the provision or give it a meaning which its words cannot bear (at [47]–[50] and [54]). The same principles must govern whatever interpretive assistance the Guidelines are capable of rendering.

The statutory direction to “have regard” to the Guidelines therefore answers one question, but not another. It answers whether the ECT and High Court may simply leave the Guidelines out of account: they may not. However, it does not answer whether a particular proposition in the Guidelines is correct, what weight it should receive, or whether it can properly influence the construction of s 14(2). Those questions remain governed by the ordinary interpretive exercise. This distinction also accords with the position stated in the authorities and commentaries concerning statutory guidance. Such guidance must be genuinely considered where the governing scheme so requires, and the weight to be given to it depends upon that scheme, the circumstances, and the cogency of its reasoning. However, guidelines do not acquire the force of legislation, remain incapable of altering the true legal meaning of the statute, and carry no presumption that their account of the law is correct (see, eg, Bennion at para 24.16; Lowe & Potter at paras 7.42–7.43.4; Craies at paras 27-031, 27-038, and 27-042; also see Wathen-Fayed v Secretary of State for Housing, Communities and Local Government [2025] 1 WLR 3693 at [59]–[61]).

Within those boundaries, the Guidelines may nevertheless perform some useful and substantial work. Most obviously, they may assist the ECT in understanding the industrial setting in which s 14(2) operates, including recurring grounds for dismissal and the prevailing practices which MOM, the National Trades Union Congress (“NTUC”), and the Singapore National Employers Federation (“SNEF”) considered relevant. Their illustrations may identify recurring situations in which just cause or excuse will ordinarily exist or be absent, while any reasoning which they contain may persuade the ECT that one construction available from the statutory language is preferable to another. They may also assist in applying a properly ascertained statutory standard to particular facts. In these ways, the Guidelines may materially influence adjudication without themselves supplying the governing rule.

The exact influence the Guidelines may have in respect of the meaning of s 14(1), however, must be stated carefully. First, where a proposition in the Guidelines accords with the meaning of s 14(2) independently ascertained, the tribunal may adopt it and use the accompanying illustrations to assist the application of that provision. Second, where the Guidelines do not address a situation, their silence cannot be treated as excluding it from the statutory protection, particularly where their illustrations are expressed to be non-exhaustive. Third, where a proposition in the Guidelines contradicts, narrows, enlarges, or otherwise displaces the meaning which s 14(2) bears upon its proper construction, the ECT must consider that proposition but decline to follow it. The inconsistency is itself a decisive reason for doing so. Having regard to the Guidelines does not require the ECT to commit an error of law, and the ECT’s judicial duty is discharged—not avoided—when it identifies the inconsistency and explains why the terms of the enacted provision must prevail.

This approach gives effect both to the statutory direction in s 20(7)(a), and to the practical expectation disclosed by the speeches of the Minister and the Members during the Parliamentary debates. The Guidelines may clarify recurring industrial situations, promote consistency, and exert influence where they accurately particularise the standard which Parliament enacted. What they cannot do is substitute a different framework for that standard. The expectation that the Guidelines would perform significant practical work must be realised through the juridical means which Parliament supplied; it cannot enlarge those means beyond the language Parliament enacted.

Drawing the threads together, the protection’s binding force remains exclusively statutory in the terms of s 14(2). The common law label of “wrongful dismissal” contributes no independent authority or content to that protection, and the Guidelines derive legal significance from the ECA’s direction that they be regarded but possess no independent force capable of controlling s 14(2). Their contents may assist where they illuminate or faithfully apply the enacted standard; where they depart from it, they must yield. The practical importance of that conclusion depends upon whether the Guidelines do, in fact, reproduce the content of s 14(2). I turn next to that question.

Content

That question must, in my view, be answered in the order required by the ordinary exercise of statutory interpretation. I must first ascertain the meaning which s 14(2) bears by reference to its text, context, and purpose, and do so without assuming that the Guidelines have stated that meaning correctly. It is only after that meaning has been ascertained that the account supplied by the Guidelines can sensibly be compared against the statutory protection. If one were instead to begin with the Guidelines and ask whether the words in s 14(2) can be made to accommodate them, one would repeat, at the level of content, the very inversion identified at [82(b)] above. For present purposes, I am concerned only with the matters which must be established before a dismissal attracts liability under s 14(2). The distinct questions of how those matters are to be proved, and what relief follows once they are proved, will be addressed later from [119] and [172] respectively.

It is useful, therefore, to begin by setting out ss 14(1), (2), and (3):

The natural starting point is the expression which Parliament chose to enact: “dismissed without just cause or excuse”. It bears recalling that the label “wrongful dismissal” does not appear in s 14(2). That label appears only in s 2(1) of the ECA, which describes a dispute arising under s 14(2) (as well as s 84(2) of the EA) as a “wrongful dismissal dispute”. Accordingly, the label cannot supply the meaning of the expression used in s 14(2). That meaning must be derived from the words enacted in the EA, read in their statutory setting and as part of the Act as a whole.

In this regard, the word “without” frames the statutory condition negatively. The question posed is therefore whether the requisite justification was absent, rather than whether the employee can attach some independently recognised label of “wrongfulness” to the employer’s reason (cf para 9 of the Guidelines reproduced at [27] above). The words “cause or excuse” appear naturally to direct attention to the basis upon which the dismissal occurred. A “cause” ordinarily connotes something which brought about, or bore upon, the decision, while an “excuse” connotes a reason advanced to account for or justify it. The word “just” then seems to call for an evaluative judgment about that basis. Thus, a fact may have been true and yet have played no material part in the decision; an operative reason may equally have been too slight, irrelevant, or otherwise insufficient in the circumstances to justify dismissal. A reason might also have been capable of justifying dismissal had it been known and acted upon, but formed no part of the basis upon which the employer in fact dismissed the employee. On this reading, the expression appears to impose two related requirements: the asserted basis must possess a sufficient connection with the dismissal, and it must be sufficient, in the circumstances, to justify it. Mere proof that some adverse fact concerning the employee was true does not, without more, establish either requirement.

The need to distinguish the fact of dismissal from the basis capable of justifying it is also reflected in the meaning which the EA assigns to “dismiss”. Under s 2(1), dismissal includes a termination at the employer’s initiative “with or without notice and for cause or otherwise”, as well as a resignation which the employee proves was not voluntary and was “forced” by the employer’s conduct, omission, or course of conduct. Two points follow. First, the question whether there was a dismissal is anterior to the question whether it was without just cause or excuse. A termination does not cease to be a dismissal merely because the employer gave notice, complied with the contract, or acted for an asserted cause (cf Noor Mohamed bin Mumtaz Shah v Apollo Enterprises Ltd (trading as Apollo Hotel Singapore) [2000] 1 SLR(R) 670 which was decided at a time where there may be said to have been some ambiguity about what “dismiss” in s 14 meant). Second, the question of justification may therefore arise even where the employer had, and validly exercised, a contractual power to bring the employment relationship to an end. Any other conclusion would leave the words “with … notice” and “for cause or otherwise” with little work to do when the definition is applied to s 14(2).

That distinction also explains why the requirements governing termination under ss 10, 11, and 14(1) must be kept analytically separate from the inquiry under s 14(2) (also see Chandran (Lexis) at para 6.152). Compliance with, or breach of, those requirements may, of course, bear upon that inquiry. Neither, however, determines whether the dismissal was supported by just cause or excuse. Read in this way, the surrounding provisions retain their respective contractual and statutory consequences, while s 14(2) subjects the employer’s decision to terminate to a distinct inquiry into the basis and merits of the dismissal.

The requirement of connection directs attention, in turn, to the basis which materially brought about, or bore upon, the dismissal. Evidence obtained after the dismissal may prove that such a basis existed and operated at the material time. It does not follow, however, that a fact first discovered after the dismissal, or a rationale assembled only then, can retrospectively become part of the basis upon which the dismissal occurred. Whether such a matter may possibly bear upon relief is a distinct question; but it should not alter the inquiry into liability under s 14(2) since that provision invites a search for the “cause or excuse” which justified the dismissal, and that naturally requires a link between the cause or excuse and the decision to dismiss.

It is against that prima facie meaning that the Guidelines must now be considered. The differences between the account of liability supplied by the Guidelines and the one which emerges from s 14(2) may, for ease, be arranged into six groups.

(a) First, there is a difference in both the statutory gateway and the governing formula adopted by the Guidelines.

(b) Second, contractual notice becomes, in their account, an organising principle of substantive liability.

(c) Third, where no reason was given, the absence of just cause or excuse is replaced by proof of a positive “wrongful reason”.

(d) Fourth, the truth or falsity of a communicated reason is made to attract categorical consequences.

(e) Fifth, particular grounds are classified as legitimate or wrongful without maintaining the statutory inquiries into existence, causal operation, and justificatory sufficiency.

(f) Sixth, misconduct and due inquiry are treated as determinative of whether a dismissal without notice is wrongful.

As to the first, para 1 of the Guidelines omits the words “for cause or otherwise” from the first limb of the statutory definition and reduces the second limb to “involuntary resignation”. That reduction compresses the statutory limb into two words and, in particular, omits both the distinct requirement of compulsion and the requirement that such compulsion be attributable to the employer’s conduct or omission. Paragraph 2 then states that, “in accordance with” the EA, a dismissal is wrongful if it is without “just or sufficient cause”, although s 14(2) uses the expression “without just cause or excuse”. By omitting “excuse”, inserting “sufficient”, and joining the resulting adjectives disjunctively, the Guidelines change the words which define the protection before any effort has been made to interpret them. In my view, that goes further than simply expressing the enacted rule shortly.

The same hybrid formula appeared in the 2018 Debates. In replying to Members, the Minister likewise described a dismissal as being without “just or sufficient cause”, before identifying discrimination, denial of statutory benefits, and punishment for exercising rights as wrongful grounds, and poor performance, misconduct, and redundancy as grounds which were not wrongful. No explanation was given for the disappearance of “excuse” or the introduction of “sufficient” (2018 Debates at 7.03pm (Mrs Josephine Teo, Minister for Manpower)). Assoc Prof Ravi Chandran, in an article commenting on the 2018 legislative changes, considered the different expression “sufficient cause” in s 84(2), but concluded that there should be no difference between that standard and s 14(2) because he could not discern why Parliament would have intended different standards (Ravi Chandran, “Employment Act Changes—Implications and Uncertainties” (2019) 31 Singapore Academy of Law Journal 316 (“Chandran (SAcLJ)”) at para 24). However, that reasoning seems to me to proceed in the wrong direction. The absence of an immediately apparent policy explanation cannot permit different enacted words to be treated as identical, and provides still less reason to combine the two expressions into a third which appears in neither provision.

I turn, then, to the second difference, which concerns contractual termination and notice. Paragraph 3 begins with the accurate proposition that employer and employee may each possess a contractual right to terminate the relationship. Paragraph 7 then says that, because both possess that right, a dismissal with notice is “presumed not to be wrongful”. Illustration 3 carries the proposition further by treating as non-wrongful a dismissal where the employer gave no reason and the employee could identify no facts suggesting some intention beyond contractual termination. The difficulty is that this conspicuously overlooks how s 14(2) itself frames the restraint upon the employer’s dismissal of an employee, namely, by requiring “just cause or excuse”, while ss 2(1) and 14(3) make clear that contractual notice does not take the dismissal outside that restraint. Granted, notice may be evidentially relevant, as I will explain from [124]–[125] and [166]–[169] below, and may also determine separate rights and liabilities under contract or ss 10 and 11. Standing alone, however, it cannot sensibly constitute a “just cause or excuse for dismissal”, and the Legislature enacted no presumption that it should be treated as one.

There is, moreover, a difficulty within the Guidelines’ own notice-based structure. Paragraph 1 includes an immediate termination accompanied by salary in lieu and an “involuntary resignation” within its account of dismissal. The provisions which follow then divide cases between dismissal “with notice” and dismissal “without notice”, but do not explain where either situation falls. If salary in lieu is treated as notice, paras 7 and 9 apply; if the immediacy of the termination makes it a dismissal without notice, paras 5 and 6 apply. Those routes may point in opposite directions. An apparent resignation forced by the employer also sits uneasily within that division. Section 14(2), by contrast, applies expressly with or without notice and does not organise substantive liability around the distinction.

Some of the discussion during the 2018 Debates reveals, I think, the same inclination towards reasoning in contractual terms. Mr Dennis Tan recounted a case in which an employer had terminated an employee under a contractual provision, paid the required sums, and was nevertheless asked by MOM to make an ex gratia payment. He objected that this failed to give full recognition to the employer’s contractual position. The case was not recounted with enough facts to determine whether payment had been sought because the dismissal lacked just cause or excuse. It cannot, therefore, establish that Mr Tan regarded contractual compliance as invariably decisive. It does, however, show how readily the statutory inquiry could be regarded as exhausted by the contract. The Minister’s reply moved in the same direction when she said that, in cases of contractual termination where the employer had met all its obligations, the Tripartite Alliance for Dispute Management (“TADM”) would first require the employee to substantiate a case before contacting the employer to attend mediation. That may have been intended merely to describe an administrative screening practice for the purposes of mediation. However, even then, one can easily see how such perspectives led to the contractual presumption stated in para 7 of the Guidelines (2018 Debates at 4.33pm (Mr Dennis Tan Lip Fong) and 7.03pm (Mrs Josephine Teo, Minister for Manpower)).

The difficulty which can arise from this view of notice’s relationship with the s 14(2) inquiry may be seen in a case decided in 2024. The employment agreement there stated that the employer could terminate it “without just cause” upon two weeks’ notice. The employee was dismissed without a stated reason and alleged that the employer intended to deprive her of severance and leave benefits. Applying paras 7, 9, and 11 of the Guidelines, the tribunal confined its inquiry to whether those entitlements and that purpose had been proved. More importantly, it held that the agreement permitted the employer to terminate without just cause upon notice, and supported that conclusion by reference to the supposedly reciprocal contractual right stated in the Guidelines. The alleged entitlements were not established, and the claim therefore failed. Those findings may well have defeated the particular benefit-deprivation case which the employee advanced. I am not sure, however, that they could make the contractual term an answer to s 14(2), for the term purported to permit precisely the state of affairs which engages the statutory protection, notwithstanding the authority in s 14(3) to order relief.

I turn next to the third difference, which follows from, but goes further than, the contractual presumption. Paragraph 9 provides that, where no reason is given for a dismissal with notice, an employee “must substantiate a wrongful reason”, including discrimination, deprivation of benefits, and punishment for exercising an employment right. Granted, the word “include” means that the listed reasons are not exhaustive. The more fundamental difficulty is that para 9 changes the object of proof. Section 14(2) asks whether just cause or excuse was absent, whereas para 9 requires the employee to establish the presence of a positively wrongful motive. An employee may, for example, be able to show from a consistently satisfactory record, the absence of any operational change, and the employer’s inability to identify a contemporaneous concern that no legitimate basis probably existed. Yet, she may remain unable to discover and prove which hidden motive did operate. Paragraph 9 thereby turns that evidential difficulty into a substantive limitation upon the protection.

Indeed, the same confusion appeared before the Guidelines were promulgated. When introducing the 2018 amendment to the definition of “dismiss”, the Minister said that an involuntary resignation could be wrongful where an employee had been forced to resign for “wrongful reasons”. The definition in s 2(1), however, asks only whether the resignation was voluntary and whether it was forced by the employer’s conduct or omission. Whether the resulting dismissal was without just cause or excuse is a distinct question under s 14(2). The Minister’s formulation merged those questions and, in doing so, treated a positive wrongful reason as part of both. Paragraph 9 makes the same conceptual move by requiring an employee who was dismissed with notice and given no reason to prove a positive wrongful reason as her route to liability (2018 Debates at 3.56pm (Mrs Josephine Teo, Minister for Manpower)).

The consequence of applying those premises to a s 14(2) claim, where the dismissal was effected with notice and without stated reasons, is illustrated by a case from 2025. The employee was dismissed with three months’ salary in lieu, and the termination letter gave no reason. At the hearing, the employer gave evidence of competency, attendance, and workplace-relational concerns which, it said, had led to the dismissal. The ECT nevertheless held that this evidence was “not material” because the termination letter stated no reason and the employee bore the burden of proving a positive wrongful reason. It therefore examined only her allegation that she had been dismissed in retaliation for complaints made within the organisation, which it rejected because the later whistleblowing complaint post-dated the decision to dismiss, while the earlier complaints were temporally remote, had been acted upon, or lacked evidence of any nexus with the dismissal. I make no comment about the result, which may have been supportable if the contemporaneous concerns were real, operative, and sufficient. For present purposes, what matters is that the framework supplied by the Guidelines seemed to make it unnecessary for the tribunal to decide those questions, because evidence directed to the statutory inquiry was put aside when the employee could not establish the different matter required by para 9.

That brings me to the fourth difference, which concerns a reason stated by the employer. The Guidelines move between three ideas which, in my view, must be kept separate: what the employer communicated; whether the facts stated were true; and what actually caused and justified the dismissal. Paragraph 4 provides that, where misconduct or poor performance is cited, the dismissal is considered wrongful if the employer cannot prove the ground. Illustrations 1 and 2 then tend to treat proof of the cited ground, together with the process described, as sufficient to make the dismissal “not wrongful”. Paragraphs 9 and 13 proceed in the converse direction by providing that a stated reason which is proved false makes the dismissal “wrongful”. Truth is thereby treated as tending towards sufficiency and falsity as conclusive. The natural reading of s 14(2), however, requires the ECT to examine and distinguish between factual truth, causal operation, and justificatory sufficiency. A true fact may have played no part in the decision, or may be insufficient, while a false communication may be evidence of pretext even though another operative cause or excuse remains possible. The Guidelines also leave open whether an honest and reasonable mistake may amount to an excuse.

Chandran (SAcLJ) identified part of this distinction though, unfortunately, he did not hold to the interpretive method needed to resolve it. He correctly observed that an employer which proves a cited ground must probably go further and show that it presents just cause or excuse. He likewise recognised that a real ground might not justify dismissal where, for example, long and exemplary service made a lesser response appropriate. He then suggested, however, that the common law rules concerning other reasons raised after dismissal should continue to apply because the ECA’s burden provisions did not exclude them, and that the ECT would probably apply the English “range of reasonable responses” approach (cf Blackstone’s Employment Law Practice 2023 (Lydia Banerjee and Gavin Mansfield eds) (Oxford University Press, 11th Ed, 2023) (“Blackstone’s”) at paras 25.03, 25.23–25.27 and 25.30–25.33). I can find no basis for either proposition in the text, context, or purpose of s 14(2). The first assumes, without argument, that the common law treatment of later-raised reasons—usually called the principle in Boston Deep Sea Fishing and Ice Co v Ansell (1888) 39 Ch D 339 (“Boston Deep Sea Fishing”)—carries into the statutory protection. The second risks replacing the ECT’s statutory judgment as to whether the cause or excuse was just with a deferential review of whether some reasonable employer might have dismissed, an approach which has itself been criticised on that basis (see, eg, Aaron Baker, “The ‘Range of Reasonable Responses’ Test: A Poor Substitution for the Statutory Language” (2021) 50(2) Industrial Law Journal 226).

One side of the Guidelines’ rule may be seen in a case from 2023. The employee there was forced to resign after departing from the employer’s travel-planning protocol during a business trip and changing scheduled customer meetings. The tribunal accepted that his intentions appeared well-meaning and were directed towards meeting customers whom he regarded as more valuable, but found that he had nevertheless departed from the protocol and caused unhappiness within the overseas team. It then applied the Guidelines’ contractual presumption and false-reason rule. Because the complaint stated to the employee was “not proven to be false”, and because a purpose to deprive him of benefits was not established, the dismissal was held not to be wrongful. The truth of the protocol departure, however, established only the factual ground. There would, I think, still remain the further question of whether, having regard to its setting, gravity, and any industrial latitude permitted by s 14(2), that departure supplied just cause or excuse.

The converse difficulty appeared in a case from 2025. The dismissal letter there gave two reasons: inadequate improvement and commitment in the employee’s performance, and a serious violation of honesty and integrity arising from information given to a colleague about a day off. The tribunal found the performance ground proved by written warnings and contemporaneous feedback. It nevertheless found the honesty allegation false because the employee had misunderstood the position, sought clarification, and promptly corrected it. From that falsity, the tribunal held that the dismissal was wrongful despite the separately proved performance ground. Thus, the finding that one stated ground was false determined liability; however, that was without any decision as to whether the proved performance concern had actually operated and was sufficient under s 14(2).

The fifth difference concerns the substantive categories adopted in the Guidelines. Paragraphs 5–8 treat misconduct, poor performance, and redundancy as legitimate grounds, differentiated principally by whether notice was given, while paras 9–12 identify discrimination, deprivation of benefits, and punishment for exercising an employment right as wrongful grounds. These may, of course, furnish sensible illustrations. Serious misconduct, sustained poor performance, and genuine redundancy may plainly provide just cause or excuse, just as discrimination, deliberate benefit deprivation, and retaliation may plainly demonstrate its absence. The labels cannot, however, answer the statutory inquiry by themselves. The ECT must still determine what constitutes the ground, whether it existed and actually bore upon the dismissal, and whether it justified dismissal in the circumstances. The Guidelines do not identify the causal threshold for discrimination or retaliation, the treatment of mixed grounds, what benefits an employee “would otherwise have earned”, or the limits of an “employment right”. Proof of some poor performance, for example, does not necessarily establish its gravity or causal role; conversely, a dismissal may lack just cause or excuse although it falls within none of the wrongful categories stated in the Guidelines.

The 2018 Debates disclose, in my view, the same tendency to treat categories as supplying the content of the protection, rather than illustrating its application. The Minister stated categorically that dismissals because of poor performance, misconduct, and redundancy were not wrongful, without qualifying that statement by reference to truth, causation, gravity, or sufficiency. Assoc Prof Daniel Goh, in turn, asked whether the Guidelines would “cover” discrimination, mental-health grounds, sexual harassment, and false retrenchment. As I have said, the situations which they identified may well disclose dismissals without just cause or excuse. The difficulty lies in the assumed method, under which categories supplied by the Executive through the Guidelines would clarify s 14(2)’s content. For the reasons given at [79]⁠–⁠[83], that inverts the proper approach. Once the statutory hierarchy is restored, those categories may illuminate the application of s 14(2), but acquire no status as necessary or sufficient conditions for establishing or resisting a claim (2018 Debates at 7.03pm (Mrs Josephine Teo, Minister for Manpower) and 5.40pm (Assoc Prof Daniel Goh Pei Siong)).

I come, finally, to the sixth difference, which concerns dismissal without notice and the process which precedes it. Paragraph 5 of the Guidelines states that misconduct is the “only legitimate reason” for dismissal without notice, while para 6, read with Illustration 2, proceeds on the basis that poor performance cannot justify such a dismissal unless it amounts to misconduct and a proper inquiry has been conducted. Although those propositions appear to be premised upon s 14(1), they bring together three distinct issues: (a) the specific circumstances in which an employer may, after due inquiry, dismiss an employee without notice for misconduct under s 14(1); (b) the circumstances in which a contract of service may more generally be terminated without notice, and the consequences which follow if none applies; and (c) whether a dismissal, however effected, was without just cause or excuse under s 14(2).

The first two issues are not governed by s 14(1) alone. Section 11(1) permits either party to terminate without notice by paying salary in lieu, while s 11(2) separately permits termination without notice where the other party has wilfully breached a condition of the contract. Section 14(1), for its part, permits an employer, after due inquiry, to dismiss without notice for misconduct inconsistent with the fulfilment of an express or implied condition of service. If “without notice” in para 5 bears the meaning used in s 11(1), and therefore includes immediate termination accompanied by salary in lieu, misconduct is plainly unnecessary. If the expression is intended to mean dismissal without either notice or salary in lieu, s 11(2) remains and prevents s 14(1) from being treated as exhaustive. Wilful breach and misconduct may certainly overlap, and the relationship between the two provisions may itself present some difficulty, but they are formulated differently. As such, the statutory scheme cannot simply be reduced to the categorical proposition stated in para 5.

The third issue is governed by s 14(2), which operates “[d]espite subsection (1)” and, when read with the definition in s 2(1), applies to dismissals “with or without notice and for cause or otherwise”. Whether there was a sufficient basis for ending the employment relationship is therefore distinct from whether the employer was entitled to end it immediately without giving notice or paying salary in lieu. Poor performance may, for example, be sufficiently serious to supply just cause or excuse for dismissal while falling short of the misconduct or wilful breach required to dispense with notice. The employee may then be entitled to salary in lieu, but it does not follow that the dismissal was also without just cause or excuse. Illustration 2 obscures that possibility because the factual basis remains the same in both versions of the illustration: documented poor performance which did not improve. The Guidelines nevertheless treat the dismissal as “not wrongful” if notice is given and “wrongful” if it is not, because the performance was not clearly misconduct and no proper inquiry had been conducted.

A case from 2026 illustrates the resulting conflation. The employee there was summarily dismissed during her probation although her contract required one week’s notice. The tribunal accepted that she had been late for work and had forgotten to send certain videos to a colleague, but found that those performance-related failings were insufficiently serious to constitute misconduct justifying summary dismissal. Applying para 5 of the Guidelines, it held that the employer could dismiss only upon notice and awarded salary in lieu. The tribunal then carried that conclusion directly into s 14(2), without deciding whether the real and operative performance deficiencies, though insufficient to dispense with notice, were weighty enough to justify dismissal.

The common feature of the six differences can now be drawn together. Each takes a circumstance which may bear upon the statutory inquiry—contractual notice, the communication of a reason, its truth or falsity, the category into which it falls, or the process preceding dismissal—and gives that circumstance a conclusive effect which s 14(2) does not itself supply. Because those effects may operate in either direction, a dismissal may escape scrutiny without any just cause or excuse being identified, or attract liability without its absence having been established.

The apparent solution to this odd state of affairs would be for the ECT, when determining a s 14(2) claim, to focus upon the statutory inquiry actually posed and to approach the Guidelines with the caution described at [87] above. That solution, however, exposes a further difficulty. The common law presently imposes no obligation upon employers to have a reason for dismissal, much less to furnish one to the employee, and nothing in our legislation creates such a duty (on this, also see MOM’s published guidance on termination with notice: “Is my employer required to give me a reason for terminating my employment?” (last updated 27 April 2026). Yet, once the Guidelines are put aside, the inquiry required by s 14(2) is a negative one directed at whether the dismissal was unsupported by just cause or excuse. If the employer need not disclose any reason, while the employee’s right to redress depends upon proving the absence of just cause or excuse, one is left with a practical question which cannot be avoided: how is the employee to prove that negative?

As to breach and evidential rules

I turn next to the Second Fundamental Question: who must prove the facts needed to engage or resist the engagement of the protection conferred by s 14(2). At first glance, this may appear to concern only the mechanics by which a breach is established. However, because the statutory condition is expressed through the absence of just cause or excuse, the allocation of proof bears directly upon the practical content of the right: it may determine both whether the protection can be enforced and the inquiry which the ECT must undertake. For the reasons already given from [66]–[88] above, that allocation must be found in the governing provisions and in the ordinary principles applicable to any gaps which they leave. It cannot be supplied by the presumptions or categorical consequences stated in the Guidelines where those propositions are not borne out by the statutory scheme.

The obvious starting point is s 27(2) of the ECA, reproduced at [31] above. It makes two allocations which are material to the present discussion. Under para (a), where an employee is dismissed without notice by an employer “under section 14(1)” of the EA and brings a claim under s 14(2), the employer bears the burden of proving the allegation that the employee was dismissed with just cause or excuse. Under para (b), where an employee is dismissed with notice and that notice is or purports to be given on the ground of poor performance or misconduct, the employer bears the burden of proving “that ground for giving the notice of dismissal”. The precision of those conditions matters for two related reasons. First, s 27(2) states no general rule that an employer which dismisses must prove justification. Second, the two paragraphs express the burden which they do place upon the employer in materially different terms.

That does not, of course, dispose of the cases which fall outside those two paragraphs. Sections 103 and 104 of the Evidence Act 1893 (“EVDA”) ordinarily place the legal burden upon the party whose asserted facts must exist before judgment can be given in her favour, and upon the party which would fail if no evidence were adduced on either side. Section 105 separately places the burden concerning a particular fact upon the party which wishes the court to believe in its existence. Applied in the ordinary way, those provisions would leave the employee with the ultimate legal burden of proving that her dismissal was without just cause or excuse, although an employer may bear an issue-specific legal burden concerning a positive fact which it asks the ECT to find. There is, however, an anterior complication. Section 21(1) of the ECA provides that the ECT is not bound by the rules of evidence and may inform itself in any manner which it thinks fit. It is therefore uncertain whether the provisions of the EVDA apply directly and without qualification, or instead furnish principles by which the ECT may properly inform itself.

Section 27(2) plainly governs wherever its stated conditions are satisfied. As to the residual cases, while it is not wholly clear, I think it safe—without presently resolving the anterior complication just mentioned—to proceed on the basis that ss 103–105 apply. After all, they represent the general rule. As stated in Phipson on Evidence (Hodge M Malek gen ed) (Sweet & Maxwell, 21st Ed, 2025) (“Phipson”) at para 6-06 (adopted in Cooperatieve Centrale Raiffeisen-Boerenleenbank BA v Motorola Electronics Pte Ltd [2011] 2 SLR 63 (“Cooperatieve”) at [31]–[32]) as follows:

Even if one proceeds on that basis, oddities and gaps remain in the proof of a s 14(2) claim. They are best exposed, I think, through four cases:

(a) The first is where an employer satisfies its contractual notice obligation and states no reason for dismissal. I will refer to these as “notice, no-reasons” or “NNR” cases.

(b) The second is where an employer satisfies that obligation but states a reason without relying upon it as entitling the employer to withhold notice. I will refer to these as “notice, with-reasons” or “NWR” cases.

(c) The third is where an employer withholds contractual notice and states a reason which, it says, entitled it to do so. I will refer to these as “no-notice, with-reasons” or “NNWR” cases.

(d) The last is where an employer withholds contractual notice but states no reason which, it says, entitled it to do so. I will refer to these as “no-notice, no-reasons” or “NNNR” cases.

Two qualifications should be made about those labels. First, the references to notice concern satisfaction of the contractual notice obligation, whether by working notice or salary in lieu. I recognise, as mentioned at [100] above, that s 11(1) describes an immediate termination with salary in lieu as termination “without notice”, whereas s 27(2)(b) refers to an employee “dismissed with notice”. That wording creates some doubt as to whether para (b) applies where salary in lieu is paid. I leave that point aside for now and use the labels in the general contractual sense just described. Second, a case may move from “no-reasons” to “with-reasons” if the employer identifies during the proceedings a reason which it says actually operated when the dismissal occurred. A reason first disclosed later must, however, be kept distinct from a reason first conceived, acquired, or discovered later. Only the former is capable of being evidence of the historical basis upon which the dismissal was effected.

Subject to those qualifications, the two axes perform different evidential work. The satisfaction or withholding of notice appears to determine—though imperfectly, given the difficulty just mentioned—which of the specific provisions in s 27(2) may be engaged. The identification of a reason determines whether the negative inquiry under s 14(2) has a known object. Where a reason is stated, the employee can direct her evidence towards its existence, its connection with the decision, and its justificatory sufficiency. Where none is stated, she may instead be required to negate an unknown range of possible explanations. NWR and NNWR cases most closely resemble the situations addressed by s 27(2)(b) and (a) respectively. As will become apparent, however, even those cases are not completely covered.

I begin with an NWR case involving poor performance or misconduct. Section 27(2)(b) applies where the notice is or purports to be given upon either of those grounds, and the employer must then prove “that ground”. In that respect, the provision fits the negatively framed inquiry under s 14(2): the employee need not first prove that the alleged poor performance or misconduct did not occur before the employer has adduced anything to show that it did. Further, the words “that ground for giving the notice” [emphasis added] suggest that the employer must prove both the underlying events and that the ground genuinely operated upon the decision to dismiss. A fact which existed, but bore no part in the dismissal, would not ordinarily be the ground “for giving” notice.

There is, however, an oddity in the drafting. Unlike para (a), para (b) does not reproduce the expression “just cause or excuse”. Even so, I do not think that the burden which it creates can be discharged merely by proving the existence of some adverse fact. In JGP, I concluded that the employer must prove that the stated poor-performance or misconduct ground was true, that it was the ground upon which notice was given, and that it supplied just cause or excuse for the dismissal (see JGP at [72]–[74], [79(i)]–[79(j)], and [91]). I remain of that view. Otherwise, para (b) would place the burden upon the employer as to factual truth, while leaving the employee to disprove justificatory sufficiency, even though the latter is the question which s 14(2) ultimately poses. The difference in wording between paras (a) and (b) is inelegant. However, it does not, in my view, justify dividing the same stated ground into separate burdens concerning its existence, its operation, and its legal sufficiency.

The gaps start to become clearer where an NWR employer states some other reason. Redundancy, reorganisation, incapacity, incompatibility, the loss of a necessary qualification, and a non-misconduct breach of contract cannot seriously—at least not without disingenuity—be characterised as “poor performance” or “misconduct”. Paragraph (b) therefore appears to leave the ultimate legal burden under s 14(2) upon the employee, although an employer which asks the ECT to find a positive justificatory fact may bear an issue-specific legal burden concerning that fact (cf s 105 of the EVDA).

A separate uncertainty also arises where the asserted reason is disclosed for the first time in the proceedings. Late disclosure does not, by itself, mean that the reason was acquired late: the employer may be adducing later evidence of a consideration which actually operated at the time. Conversely, a reason first conceived, or a fact first discovered, only after dismissal cannot retrospectively become the ground upon which notice was given. The words “is … given on the ground” may extend to an historical ground which actually operated despite its earlier non-communication, whereas “purports to be given” seems more naturally directed to a ground communicated when notice was given. That lack of clarity is real, but is less serious for the employee than the gap which arises where no reason is identified at all. Once the employer commits itself to an asserted historical reason, the employee can seek to disprove its truth, operative role, or sufficiency and, in that practical sense, the reason bounds the negative which she must prove.

Next, in an NNWR case, s 27(2)(a) supplies the closest match to the full statutory inquiry. Where the employee was dismissed without notice “under section 14(1)”, the employer bears the burden of proving that the dismissal was with just cause or excuse. Unlike para (b), para (a) uses the expression found in s 14(2) itself. It therefore appears to require proof of the relevant misconduct, its connection with the dismissal, and its justificatory sufficiency. In such a case, the negatively framed right and the legal burden face in the same direction.

The difficulty lies in the words “under section 14(1)”. That provision concerns dismissal without notice, after due inquiry, for misconduct inconsistent with the fulfilment of an express or implied condition of service. Section 27(2)(a) does not extend in terms to a dismissal “under or purportedly under” s 14(1), although para (b) expressly includes notice which “purports” to be given upon one of its specified grounds. It is therefore uncertain whether para (a) applies where, for example, an employer dismisses for alleged misconduct without conducting due inquiry and makes no express invocation of s 14(1). A narrow reading risks making the allocation of proof depend upon the very merits which the allocation is needed to determine. A wider reading, however, must account for Parliament’s use of different language in the two paragraphs. Neither course is entirely satisfactory.

Quite apart from that difficulty, an employer which withholds notice may rely upon s 11(2), an express contractual term, an accepted repudiatory breach, or some other basis which does not depend upon s 14(1). Paragraph (a) does not clearly govern such a case (see Phosagro Asia Pte Ltd v Piattchanine, Iouri [2016] 5 SLR 1052 (“Phosagro Asia”) at [41]–[47]).

Taking stock, NWR and NNWR cases correspond only partly with s 27(2). Paragraph (b) is confined to two specified grounds, while para (a) is confined to a dismissal under s 14(1), the boundaries of which are themselves uncertain. The Guidelines may explain the practical effect of an allocation which s 27(2) actually makes. They cannot extend that allocation to cases which Parliament omitted, alter the matter which the statute requires a party to prove, or give legal force to the presumption stated in para 7 where the statutory scheme provides none. For the reasons already given, any proposition in the Guidelines which would produce one of those effects must yield.

The gaps become still more conspicuous in an NNNR case. As a matter of practical inference, the withholding of notice may suggest that the employer considered some cause to entitle it to dismiss summarily. The manner of termination may therefore call for an explanation more readily than a dismissal effected with notice. However, neither proposition establishes that the employer acted under s 14(1). It may have relied upon s 11(2), an express term, an accepted repudiatory breach, or no legally sufficient basis at all. An inference that the employer probably had some reason cannot identify the juridical source of that reason or, without more, bring the case within the particular language of s 27(2)(a).

Nor can para 7 of the Guidelines supply the missing allocation. As explained above, any presumption which it states must first be assessed against the ECA and cannot be treated as a legal presumption merely because it appears in the Guidelines. Indeed, it would be especially unsafe to derive a legal presumption not even from the Guidelines expressly, but by inverting a proposition stated in the Guidelines. Paragraph 7, after all, only addresses cases where dismissal is effected with notice. It says nothing about cases where a dismissal is effected without notice.

While the reverse may seem intuitive, it must be remembered that presumptions can be justified on various grounds—epistemic relevance would be the obvious justificatory candidate (see, eg, Gerhard Schurz, “Relevance as Difference-Making: A Generalized Theory of Relevance and Its Applications” (2024) 181 Philosophical Studies 2279 at 2280–2281); however, broader grounds of policy such as risk allocation are certainly also available (see, eg, Edna Ullmann-Margalit, “On Presumption” (1983) 80 The Journal of Philosophy 143 (“Ullmann-Margalit”) at 147–150). The Guidelines themselves justify the presumption at para 7 on this basis: “[a]s both employee and employer have a right to contractually terminate employment with notice, …” It is unclear what connection exists between the contractual premise and the evidential conclusion. Section 14(2) is a reasons-based protection rather than a contractual one, and the Guidelines do not explain why satisfaction of the contractual notice obligation should make the existence of just cause or excuse more probable, or justify allocating the risk of non-proof to the employee. Without a clear understanding of why a presumption is said to operate in one direction, it would be unsafe to infer from the fact of such presumption alone that another necessarily operates in the opposite direction.

That is not to say that an NNNR case can never fall within the provisions of the ECA if the Guidelines cannot aid us. Section 27(2)(a) does not expressly require the employer’s reason to have been communicated. An employer may plead during the proceedings that it acted under s 14(1), and adduce evidence that this was the historical basis upon which the dismissal was effected. A later pleading does not itself create that historical fact; the ECT must still determine whether the asserted basis actually operated at the time. If it did, para (a) may be engaged despite the employer’s earlier silence. However, if the employer elects to remain silent, that leaves both para (a) and para (b) without clear application. The spirit of para (a) may favour an explanation whenever notice is withheld, but its text states no rule of that breadth.

I come, then, to the most difficult of the four cases: NNR. Where the employer later identifies a particular consideration as the ground which, it says, actually operated, the employee will at least know the case which she must meet, though the separate uncertainty remains whether a ground disclosed only then can bring the notice within s 27(2)(b). My present concern is with the narrower case in which the employer remains silent and the objective record discloses no historical ground for the dismissal.

Neither limb of s 27(2) then clearly applies and, on the residual position described at [121] above, the employee bears the ultimate legal burden of proving that her dismissal was without just cause or excuse. Paragraph 9 of the Guidelines offers one means of making that burden manageable: it requires the employee to prove a positive “wrongful reason”, such as discrimination (see the text of para 9 reproduced at [27] above). However, as explained at [103]⁠–⁠[105], to make that the mandatory form of every claim would alter the content of the right conferred by s 14(2). For the reasons given at [79]⁠–⁠[87], an employee remains entitled to rely directly upon the statutory words. In a truly silent NNR case, she may then be left to prove what I will call an “unbounded negative”, because no one has identified the possible cause or excuse which her evidence must answer.

By an “unbounded negative”, I mean something more precise than a proposition expressed in negative form. As Phipson makes clear at [121] above, an allegation which forms an essential part of a party’s case remains for that party to prove whether it is affirmative or negative (also see Roderick Munday, Cross & Tapper on Evidence (Oxford University Press, 13th Ed, 2018) (“Cross & Tapper”) at pp 133–134). A negative which has a known object may be capable of ready proof. An employee alleged to have been absent on a specified day may, for example, produce attendance records or call persons who saw her at work. The difficulty in a true NNR case is different: what has not been supplied is the proposition whose negation the employee is expected to establish. Thus, the problem is evidential, rather than merely grammatical.

Of course, the difficulty will not arise in every case where the employer gives no reason. The surrounding circumstances may themselves disclose the apparent explanation and thereby confine the field. An immediately preceding investigation, an announced reorganisation, the loss of a necessary licence, or a precipitating exchange may each do so. My present concern is narrower. It is with the silent employer and an objective record which reveals no apparent reason, leaving the employee without a known proposition towards which her evidence can be directed.

I accept, too, that the employee need not exclude every fanciful or metaphysically possible explanation. Section 14(2) is proved upon the ordinary civil standard, rather than beyond all doubt. Even after impossible and fanciful explanations are put aside, however, a considerable range of realistic causes or excuses may remain. A favourable performance history may weigh strongly against poor performance or misconduct, while saying little about a genuine redundancy, a regulatory impediment, or the loss of a necessary qualification. The absence of warnings may answer some explanations, but have little bearing upon an operational reorganisation or an irreparable breakdown arising from events unknown to the employee. If she attempts to answer every realistic possibility, she may have to range across the employer’s business, the position of other employees, and internal deliberations to which she has never had access. This is the difficulty classically illustrated by Russell’s amusingly exaggerated celestial teapot (see Bertrand Russell, “Is There a God?” (1952) in Last Philosophical Testament, 1943–68 (John G Slater ed) (Routledge, 1997) at pp 547–548): a party cannot sensibly be required to refute one unadvanced hypothesis after another.

The seriousness of that difficulty has been recognised by the courts, though in settings quite removed from the present one. In Ang Jian Sheng Jonathan and another v Lyu Yan [2021] 1 SLR 1091, the Court of Appeal described the burden of establishing that a person did not exist as the “unenviable (if not nearly impossible) task” of proving a negative (at [10(b)]). Lord Kerr likewise observed in Crawford Adjusters (Cayman) Ltd v Sagicor General Insurance (Cayman) Ltd [2014] AC 366 that proving an absence of reasonable and probable cause involved a negative proposition, “normally among the most difficult of evidential requirements” (at [109]). More recently, and albeit in an entirely different context, the Court of Appeal in Jumaat bin Mohamed Sayed v Attorney-General [2025] 1 SLR 1287 recognised both sides of the point. It said that “[t]he rules of evidence are not validated by their ease of being fulfilled”; but, at the same time, that the burden of proving a negative “should not be so onerous that it becomes virtually impossible to discharge” (at [88]–[89] and [134]–[135]).

In some settings, the difficulty may be avoided by attending more closely to the substance of what the opposing party has asserted. Thus, in SCT Technologies Pte Ltd v Western Copper Co Ltd [2016] 1 SLR 1471 (“SCT Technologies”), once the creditor proved the debt, the debtor which pleaded payment bore the burden of proving that positive fact; the creditor was not required conclusively to prove non-payment (at [22]–[27]). In Bintai Kindenko Pte Ltd v Samsung C&T Corp [2019] 2 SLR 295 (“Bintai Kindenko”), the party relying upon a contractual exclusion had to prove it, rather than requiring the other party to prove that it was not disentitled (at [49]–[53]). Cross & Tapper similarly observes that a party which bears the burden upon an issue need not necessarily anticipate every possible defeasing defence, particularly one which could readily have been stated (at p 132).

It might be tempting to solve the NNR difficulty by treating every plausible cause or excuse as a positive defence which the employer must plead and prove. I do not think that a categorical solution of that kind is open to the ECT. The absence of just cause or excuse forms part of the condition which s 14(2) places upon the employee’s right, and s 27(2) expressly selects the circumstances in which the employer must prove the converse. To recast every uncommunicated justification as a defence would extend to all cases an allocation which Parliament confined to specified ones. To be clear, I am not suggesting that an employer can never bear an issue-specific burden concerning a positive fact which it asks the ECT to find. Sections 105 and 108 may produce that result where their respective conditions are satisfied. However, I do not think the composite condition in s 14(2) cannot be converted into a general defence merely because its negative form is difficult for the employee to prove.

The foregoing does not quite exhaust the rules concerning the burden of proof. Sections 103–105 of the EVDA are accompanied by s 108, which provides that, “[w]hen any fact is especially within the knowledge of any person, the burden of proving that fact is upon him”. As Pinsler explains, where s 108 applies, it allocates the legal burden concerning that particular fact and therefore goes beyond the movement of a merely tactical or evidential burden (Jeffrey Pinsler, Evidence and the Litigation Process (LexisNexis, 8th Ed, 2024) (“Pinsler”) at paras 12.030–12.037A). Section 108 warrants separate attention here because a consideration which actually operated upon an employer’s decision to dismiss may, in some cases, be known only to those who made that decision.

Section 108 is, however, a narrow exception, and three restrictions are material here. First, as Phosagro Asia makes clear, it does not operate simply by asking which party has better access to the evidence (at [67]–[74]). It is engaged only in exceptional circumstances where proof would be impossible or disproportionately difficult for one party, while the relevant fact can be proved without comparable difficulty by the other. Second, a bare assertion is insufficient. The party invoking s 108 must first establish a prima facie case by proving facts from which, in the absence of an explanation, liability could properly be inferred (Phosagro Asia at [68]–[72]). Third, Yap Son On v Ding Pei Zhen [2017] 1 SLR 219 (“Yap Son On”) adds that the fact must not also be capable of being known by persons other than the party said to possess the special knowledge, and that regard must be had to the balance of convenience and any disproportion of labour involved in proof (at [79]–[80]). Those restrictions do considerable work in an NNR case. If the employer’s silence were, by itself, enough to engage s 108, an employee could reverse the residual burden simply by alleging that no justification existed. That would render the specific allocations in s 27(2) substantially otiose, in much the same way as the course considered at [144] above.

Against those restrictions, how might s 108 operate in an NNR case? In my view, it can operate only at the level of an identified historical fact. Where the evidence discloses a specific consideration capable of having operated upon the decision-maker; the employee has first proved facts from which an absence of justification may properly be inferred; and whether that consideration actually operated is especially within the employer’s knowledge and cannot realistically be proved from elsewhere, s 108 may place upon the employer the legal burden concerning that particular fact. Depending upon how the issue arises, that may require the employer to prove that the identified consideration actually operated upon the dismissal. However, it is not enough to describe “the actual reason” at large as a fact especially within the employer’s knowledge. That phrase identifies no particular fact. It describes the subject matter of the whole inquiry and, if accepted as sufficient, would impose through s 108 the general burden of justification which s 27(2) does not impose.

At this point, it is especially important to keep three questions separate. The first is whether the event or condition relied upon existed. The second is whether it was known to, and actually operated upon, the decision-maker when the dismissal occurred. The third is whether that operative consideration was sufficient in law to amount to just cause or excuse. The first two are questions of historical fact and may, depending upon the evidence, concern facts especially within the employer’s knowledge. The third is an evaluative conclusion which the ECT must draw and cannot be shifted to the employer through s 108. This distinction also clarifies the temporal point. A consideration need not have been communicated, or reduced to a precise verbal formula, at the time of dismissal. It may have operated contemporaneously and be articulated or proved only later. However, a consideration which first arose after the dismissal—or a pre-existing fact which was discovered only later and was therefore unknown to the decision-maker—could not have operated upon the decision when it was made. Section 108 concerns proof of the historical reason; and I do not think it can be made to convert a subsequently acquired reason into one.

Section 108 therefore supplies only a limited qualification to the residual position described at [120]⁠–⁠[121]. It may place upon an employer the legal burden concerning a specially known fact even though the employer does not itself ask the ECT positively to find that fact. In that respect, it goes further than the ordinary movement of an evidential burden. Its operation nevertheless remains conditional upon the employee first establishing the necessary prima facie basis, and it does not impose a general burden of justification upon every silent employer. There also remains the anterior uncertainty identified at [120]⁠–⁠[121]: whether, and if so in what manner, the EVDA applies before an ECT which s 21(1) of the ECA frees from the rules of evidence. For present purposes, and consistently with the basis adopted above, I proceed on the footing that this freedom does not require the ECT to disregard the principled allocation of proof expressed in ss 103–108 and, at the least, permits those provisions to inform the ECT’s approach.

It remains to explain why this limited operation of s 108 does not render s 27(2) redundant. Paragraphs (a) and (b) impose their respective burdens from the outset whenever their stated conditions are satisfied, and each addresses the full matter which the employer is required to prove. Section 108 operates only after the necessary prima facie foundation has been laid, and only in relation to the particular fact which is especially within the employer’s knowledge. Thus, s 27(2) makes a defined allocation for specified classes of dismissal, while s 108 makes a fact-specific allocation in exceptional circumstances. Further, s 108 concerns the proof of a fact which existed. Where that fact is said to be the reason for dismissal, it must have actually operated at the material time, though it may be evidenced only later. Section 108 does not require the employer to create, communicate, or retrospectively formulate a reason where none existed.

The difficulty in reconciling these rules of proof may cause one to ask whether the employer should simply be required to state its reason for dismissal. Such a requirement would confine the inquiry from the outset. The employee could decide whether to bring a claim and, if she did, direct her evidence towards the truth of the stated ground, its actual operation, and its justificatory sufficiency. The ECT would likewise know which dispute it was being asked to decide. It is therefore easy to understand why a jurisdiction which confers a reasons-based protection against dismissal may also bring the employer’s reason expressly into the inquiry.

The two statutory settings also considered in JGP illustrate the point. In the United Kingdom, s 92 of the Employment Rights Act 1996 (c 18) (UK) (“UK ERA”) confers upon a qualifying employee a right, generally upon request, to a written statement of the reasons for dismissal, while creating special automatic entitlements in specified cases. In Malaysia, once dismissal is admitted or established in a representation under s 20 of the Industrial Relations Act 1967 (No 177 of 1967) (M’sia) (“M’sian IRA”), the employer generally bears the burden of proving that the dismissal was supported by just cause or excuse (see, eg, Dr A Dutt v Assunta Hospital [1981] 1 MLJ 304; Ireka Construction Bhd v Chantiravathan a/l Subramaniam James [1995] 2 ILR 11; and Weltex Knitwear Industries Sdn Bhd v Law Kar Toy & Anor [1998] 7 MLJ 359). The juridical means differ, but the employee is not, under either arrangement, left to disprove reasons which no one has advanced.

Singapore law has adopted neither arrangement. As stated at [117] above, there is presently no general duty to give reasons at the point of dismissal. To derive such a duty from s 14(2), merely because the right may otherwise be difficult to vindicate, would be a substantial step. It would involve inferring a distinct positive obligation of communication from the practical consequences of a negatively framed protection, even though the provision says nothing about what an employer must communicate (cf JIF v JIG [2026] SGECT 3 at [47]–[55], where I made the same broad point in relation to the negatively framed s 45 of the EA).

There remains a narrower possibility: requiring the employer to identify a reason only after an NNR claim has begun. Sections 20 and 21 of the ECA give the ECT substantial powers to identify the issues, question witnesses, obtain documents, and seek further evidence where the parties have not adequately developed the case. Those powers plainly permit the ECT to ask whether an employer relies upon a cause or excuse, require any justification which it chooses to advance to be particularised, and obtain existing material which bears upon that case. Unlike a general duty operating at the point of dismissal, a direction at the litigation stage would arise only where the information is needed to determine an actual claim.

I accept that a litigation-stage direction of this kind has considerable attraction. It would bound the negative, reduce the possibility of surprise, and permit both parties to direct their evidence towards a defined factual contest. Where the employer chooses to advance a positive justification, requiring that justification to be stated clearly is an ordinary matter of case management. The difficulty arises where the employer advances no positive case and elects instead to put the employee to proof. A direction made solely because the employee may otherwise find it difficult to prove the statutory negative goes beyond organising the case which the parties have brought. In substance, it would require the ECT to procure from one party the very case which the other requires to discharge her legal burden.

That was the course which I described in JGP as “a slippery path” (at [A.91]–[A.95]). To be clear, nothing in that description prevents the ECT from receiving evidence first furnished in the proceedings of a consideration which actually operated when the dismissal occurred. Nor does it insulate the employer from the ECT’s powers of questioning and production. The concern is with a routine direction requiring a silent employer to formulate a reason which it has not chosen to advance. Once such a reason is required, a judge-led hearing risks turning into whether the reason can be substantiated. If the employer’s failure to substantiate that compelled account becomes decisive before the employee’s own evidence has called for an answer, procedure will, in practical effect, have shifted the legal burden beyond the cases identified in s 27(2).

That result would sit particularly uneasily with the granularity of s 27(2). An NNR employer which says nothing could be required to identify and prove a reason, while an NWR employer which candidly states redundancy or reorganisation appears to fall outside para (b). Similarly, an NNNR employer could be required to establish a general entitlement to dismiss summarily even though para (a) is confined to a dismissal under s 14(1). Silence would thereby attract a wider burden than candour, and the limitations which Parliament placed upon both paragraphs would disappear through the use of general procedural powers. I do not think that ss 20 and 21 can safely be used to produce that categorical result.

I therefore think it important to keep two matters separate: the ECT’s power to obtain and manage evidence, and the question of which party fails if the evidence remains insufficient. The ECT may require an employer to particularise a justification which it advances. It may ask the decision-maker factual questions, direct the production of an identified class of documents, summon a person whose evidence appears material, or seek further evidence on its own initiative. In an appropriate case, it may also require the employer to answer whether it relies upon an identified cause or excuse where that question is necessary for the fair disposal of the particular proceedings. However, the breadth of those powers does not support a standing rule that every NNR employer must state a reason for dismissal where the law otherwise imposes no obligation of disclosure, especially if that rule would lead down the slippery path just described.

That conclusion does not mean that an NNR employee is left unable to prove the statutory negative, even where it is truly unbounded (cf [140] above). Facts may be proved circumstantially, and the absence of just cause or excuse is no exception. The employee may rely upon the objective record as a whole and ask the ECT to infer that, more probably than not, no cause or excuse sufficient to justify the dismissal actually operated. She is not required to negate every imaginable reason separately and directly. The question is whether the evidence, considered cumulatively and in its setting, makes the statutory conclusion more probable than its converse.

To see how that inquiry operates, it is necessary to distinguish the legal burden from the evidential burden. The distinction is explained in Britestone Pte Ltd v Smith & Associates Far East, Ltd [2007] 4 SLR(R) 855 (“Britestone”) at [58]–[60] (also see SCT Technologies at [16]–[19]; Pinsler at paras 12.006–12.007). The legal burden is the enduring obligation to persuade the ECT of the fact necessary for relief. The evidential burden is the tactical onus to adduce evidence, or to contradict, weaken, or explain evidence already led. In a residual NNR case, both burdens will ordinarily rest upon the employee at the outset. Once she adduces evidence capable, if accepted and left unanswered, of supporting an inference that the dismissal was without just cause or excuse, the evidential burden may move to the employer. If no answer is given, the ECT may conclude that she has discharged her legal burden, though it is not bound to do so. Subject to any burden which ss 105 or 108 place upon a particular fact, the legal burden of establishing the composite statutory conclusion remains with the employee throughout.

A real evidential basis is, however, required before that tactical burden moves. As a general starting point, the employer’s silence and the employee’s assertion that she did nothing wrong will not, by themselves, suffice. What will suffice must necessarily depend upon the circumstances. Favourable appraisals may not, for example, negate every credible operational, regulatory, or external concern. Consistently strong appraisals, a recent promotion, and the absence of any prior warning may, however, provide powerful answers to suggested performance or conduct concerns. There can be no exhaustive list of considerations which will move the evidential burden. The question is whether the matters proved are capable of supporting the general inference which the employee asks the ECT to draw.

Once that point is reached, the employer may answer in several ways. It may expose weaknesses in cross-examination, rely upon documents already before the ECT, show that another inference is at least equally probable, or adduce evidence of the consideration which in fact operated upon the dismissal. Such evidence may be furnished for the first time in the proceedings, and the consideration may likewise be articulated there for the first time. The lateness of the evidence or its articulation does not make the historical reason a later-acquired reason; what matters is whether the consideration truly operated when the decision to dismiss was made. The employer may, of course, elect to give no answer and maintain that the employee’s evidence remains insufficient. However, if the ECT considers that her evidence called for an answer, the employer runs the risk that the unrebutted inference will be accepted.

At this point, I should distinguish the movement of the evidential burden just described from the conventional adverse inference arising from withheld evidence. Section 116 illustration (g) of the EVDA permits an inference that evidence which a party could have produced, but withheld, would have been unfavourable to that party. As Sudha Natrajan v Bank of East Asia Ltd [2017] 1 SLR 141 makes clear, the inference is discretionary, may be drawn only where the existing evidence gives rise to a case to answer, and may be displaced or weakened by a sufficient explanation for the absence of the witness or document (at [19]–[23]; also see Pinsler at para 12.068). Its scope must also remain confined to what the missing evidence could sensibly have shown. Thus, where an employee has established a case to answer, an employer’s unexplained failure to call the person who decided upon the dismissal, or to produce an identified contemporaneous record after a proper direction, may strengthen that case. However, the employer’s mere refusal to formulate or advance a positive reason in the proceedings is not, without more, the withholding of “evidence”. A reason is a proposition to be proved; the testimony, documents, or other material capable of establishing whether that reason operated are the evidence. Illustration (g) may support an existing case by permitting an inference about missing evidence. It cannot supply the case which must first exist before the inference may be drawn.

The same reasoning applies to an NNNR case in so far as s 27(2)(a) is not engaged (cf [133]⁠–⁠[135] above). The withholding of notice remains relevant because it may suggest that the employer considered itself entitled to dismiss immediately. If the record also discloses no investigation, warning, precipitating event, or contractual basis for immediate termination, those matters may support an inference that no sound cause for dismissal actually operated at the time. Two distinctions must, however, be maintained. First, a failure to establish a lawful basis for withholding notice does not, without more, prove that the dismissal itself lacked just cause or excuse (cf [112]⁠–⁠[115] above). Entitlement to notice raises a distinct contractual inquiry. Second, later evidence of a reason which actually operated at the time of dismissal is different from a justification acquired only after the event. Under the Boston Deep Sea Fishing line of cases, an employer may rely, in answer to a damages claim, upon pre-termination repudiatory misconduct discovered only after dismissal, although Phosagro Asia left open whether such misconduct can defeat an accrued contractual debt such as salary in lieu of notice (at [42]–[45] and [77]–[87]). Whatever its contractual effect may be, later-discovered misconduct which did not operate upon the decision to dismiss cannot supply the historical cause required by s 14(2).

The positions I have reached may therefore be drawn together as follows. First, in an NWR case involving poor performance or misconduct, s 27(2)(b) places upon the employer the legal burden of proving the stated ground in the full sense described at [125]⁠–⁠[126]: its factual truth, its actual operation upon the decision, and its sufficiency as just cause or excuse. In an NWR case involving another stated reason, the employee appears to retain the ultimate legal burden. Once identified as the employer’s asserted historical case, the reason will ordinarily bound the negative which the employee must prove, though she remains entitled to show that it was pretextual and that some other consideration actually operated. The employer may also bear an issue-specific legal burden under s 105 concerning a positive fact which it asks the ECT to find or, where its strict conditions are satisfied, under s 108 concerning a fact especially within its knowledge. Neither allocation, by itself, transfers to the employer the ultimate burden of disproving the composite statutory conclusion.

Second, in an NNWR case which falls within s 14(1), s 27(2)(a) places upon the employer the burden of proving that the dismissal was with just cause or excuse. Where the employer relies instead upon s 11(2), an express contractual term, an accepted repudiatory breach, or some other basis for withholding notice, para (a) does not clearly govern. The asserted historical reason will nevertheless ordinarily bound the inquiry. If the employer asks the ECT to find a positive fact in support of that reason, s 105 may place the burden of proving that particular fact upon it, while s 108 may operate in relation to a fact especially within its knowledge. Separately, the substantive law governing the entitlement to withhold notice may itself place upon the employer the burden of proving a positive fact necessary to that contractual justification. Unless some applicable rule reallocates the composite statutory burden, however, the employee appears to retain the residual burden under s 14(2). The withholding of notice may support her case, but does not, without more, shift that burden or establish the statutory conclusion.

Third, in an NNR case to which s 27(2)(b) does not apply, and in any residual NNNR case outside s 27(2)(a), the employee bears the ultimate legal burden of proving that her dismissal was without just cause or excuse. That remains subject to any issue-specific burden arising under ss 105 or 108. She may discharge her burden by circumstantial evidence, and an evidential burden may move to the employer once her evidence is sufficient to call for an answer. The ECT may use its powers under ss 20 and 21 to identify the issues, obtain existing material, and test any justification which the employer advances, but those powers do not alter the statutory allocation of proof. Therefore, the inquiry should not begin with the presumption stated in the Guidelines, derive an opposite presumption from the withholding of notice, or require a silent employer to identify and prove a reason before the employee has adduced sufficient evidence to call for an answer.

I fully recognise that this is not an especially tidy result. Uncertainty remains as to whether, and if so how, ss 103–108 and 116 of the EVDA apply in proceedings before an ECT; what counts as a dismissal “under section 14(1)”; and how an immediate termination with salary in lieu fits within provisions framed around dismissal with or without notice. More fundamentally, Parliament has enacted a right expressed in negative terms but prescribed specific rules of proof for only some of the ordinary forms of dismissal to which it applies. Those gaps cannot be removed by enlarging s 27(2) through general procedural powers or by treating the Guidelines as though they supplied rules of proof which Parliament did not enact.

The remaining cases can nevertheless—at least on my analysis of how the various rules of proof may be woven together somewhat coherently—be decided by recognising that the absence of just cause or excuse may be proved circumstantially; by applying ss 105 and 108 only to the particular facts which properly fall within them; by using the ECT’s powers to obtain relevant existing material where the case requires it; and by recognising that an evidential burden may move to the employer once the employee has adduced sufficient evidence to call for an answer. Even so, despite it being possible to apply the law in this rather technical way, I think the very fact that so much needs to be drawn from outside s 27(2) of the ECA is, itself, the major problem for it highlights the oddities and gaps left by the ECA. In any case, supposing the two prior troubles are overcome, there still remains the last question as to what remedial consequences follow once the scope for liability is identified and proven.

As to remedial consequences

The final difficulty concerns what is to follow once a dismissal without just cause or excuse has been established. Section 14 identifies two remedies, namely reinstatement and compensation, while the EC Regs divide compensation into “loss of income” and “harm”. Each expression is familiar enough when read on its own. The difficulty appears when one asks what legal object or interest each remedy is intended to restore, replace, or recognise; how the three relate to one another; and which facts should govern their assessment. I begin with reinstatement because s 14 does.

Reinstatement

The structure of s 14 appears, at least formally, to give reinstatement priority. Section 14(2) permits an employee to lodge a claim for “either” reinstatement in her former employment or compensation, and lists reinstatement first. Section 14(3) adopts the same order. It first addresses a claim for reinstatement and empowers the ECT, notwithstanding any rule of law or agreement to the contrary, to direct the employer both to restore the employee to her former employment and to pay the wages which she would have earned had she not been dismissed. Only thereafter does it address a claim for compensation and empower the ECT to direct the payment of an amount of wages determined by it. The statutory sequence therefore moves, as a matter of form, from restoration to monetary substitution.

I do not think that the order of these provisions creates a substantive preference for reinstatement, since s 14(2) frames the remedies disjunctively and the word “may” in s 14(3) preserves a discretion even after liability has been established. Nevertheless, the sequence does show that Parliament placed reinstatement at the front of both the provision creating the remedial claim and the provision conferring the remedial power. The history examined at [283]⁠–⁠[347] below explains that arrangement and the greater importance which reinstatement once possessed.

Practice, however, appears to point in the other direction. So far as I am aware, the ECT has not ordered reinstatement under s 14(2) since jurisdiction was transferred to it on 1 April 2019 (also see JGP at [A.152]). I may, of course, be mistaken. However, even allowing for that possibility, the remedy appears to remain exceedingly rare, and two ECT cases usefully illustrate the broader reluctance to order it after liability has been established.

The first is the case discussed in Annex 2 to JGP at [A.25]–[A.40]. Although the employee established that the aggravated allegation used to dismiss her had not been proved, the tribunal found that her underlying conduct still amounted to misconduct. She sought reinstatement alone, contending that it would most effectively vindicate her record and protect her reputation and future employment prospects. The tribunal declined to order it because it regarded reinstatement as a discretionary and extraordinary remedy analogous to specific performance and, more particularly, considered that restoration would not answer the wrong which had been established, given that the employer could have dismissed her upon a properly framed account of the underlying misconduct. There were, in addition, practical considerations. Her former role and its headcount no longer existed, restoration would cause operational and financial hardship, and her allegations against management demonstrated that mutual trust had been irretrievably damaged. Since she had not claimed compensation in the alternative, her success on liability produced no remedy (see JGP at [A.25]–[A.36]).

The second is a case heard in 2026. A teacher was summarily dismissed after his employer adopted the findings of an investigation conducted by his former school into safeguarding concerns. The tribunal found that the established concerns amounted to serious misconduct, but treated the dismissal as wrongful because the employer had not conducted a due inquiry of its own. For the reasons explained at [112]⁠–⁠[115] above, I do not think that the absence of due inquiry, without more, necessarily answers the distinct question posed by s 14(2). Even upon the tribunal’s premise, however, there remained a separate question as to the remedy which should follow. Although the employee’s formal claim form sought both remedies, he maintained that reinstatement was the only remedy capable of addressing the stigma of having been regarded as a safeguarding risk, and the tribunal accordingly treated it as his chosen remedy. It declined to order reinstatement because the wrong which it had found was the failure to conduct due inquiry, and the directly responsive order would therefore have been one requiring such an inquiry, which the ECT had no power to make. The tribunal also found that the employer continued to regard the employee as a safeguarding risk and that the trust necessary for the relationship had broken down. Liability, once again, produced no remedy.

These two decisions form a very thin body of remedial practice, and I would not draw from them any complete account of when reinstatement may be ordered. They nevertheless disclose the questions which have so far governed the exercise of discretion after liability is established: whether restoration corresponds with the particular feature which rendered the dismissal “wrongful”; whether the former employment remains available in any practical sense; what hardship restoration would impose upon the employer; and whether sufficient trust remains for the employment relationship to resume. Although those considerations are unsurprising, the difficulty lies in identifying the juridical basis upon which they are to control a statutory remedy which Parliament placed first and made available “despite any rule of law or agreement to the contrary”.

That last expression assumes particular importance because both decisions approached reinstatement through the language of specific performance. The common law’s resistance to compelling the performance of employment contracts is well-known (see De Francesco v Barnum (1890) 45 Ch D 430; Chappell v Times Newspapers Ltd [1975] 1 WLR 482; Arokiasamy Joseph Clement Louis v Singapore Airlines Ltd [2002] 2 SLR(R) 924 (“Arokiasamy Joseph Clement Louis”) at [50]). Section 14(3) plainly removes any categorical bar which might otherwise follow from that rule. Matters such as feasibility, hardship, and the state of the parties’ relationship do not thereby become irrelevant, for they may bear directly upon whether restoration can sensibly be achieved. Their relevance and weight must, however, arise from the purpose and operation of the statutory remedy, rather than be imported without examination from the common law starting point which the words of s 14(3) were enacted to overcome.

There is also a practical consequence which should not be overlooked. In both cases, the employees selected reinstatement and received no compensation when that remedy was refused. It remains unclear whether the word “either” in s 14(2) prevents compensation from being claimed in the alternative. The risk is nevertheless evident: an employee may establish liability and yet receive no relief because she sought reinstatement. In a jurisdiction intended to provide simple and expeditious redress, that risk gives employees a substantial reason to prefer compensation.

The rarity of reinstatement may, in this way, become self-reinforcing. It may cause the remedy to be approached with caution; that caution may discourage applications; and the scarcity of applications may leave the ECT with few occasions to develop principles governing its exercise. Employees may be especially reluctant to seek the remedy where the available decisions show that its refusal can leave a successful claim without any relief.

A case which I heard after the present matter gives some indication of how employees may now regard the remedy. The employee remained unable to find work several months after his dismissal. He had lost a well-paying position from which he firmly maintained there had been no just cause or excuse to remove him and, for the most part, had preserved professional relationships with his former employer and its personnel. Yet, he did not seek restoration, explaining that “I am not seeking reinstatement because, in my opinion, that is not a feasible option available”. That explanation suggests that, despite the position which reinstatement occupies in s 14, employees may now regard it as practically unavailable.

As the history at [283]⁠–⁠[387] below will explain, reinstatement once possessed a public and industrial importance which later conditions no longer sustained. Thus, although it remains formally first in s 14, practice has plainly deprioritised it.

I do not regard the rarity of reinstatement, by itself, as a major trouble of the same kind as those discussed above. A discretionary remedy may properly be uncommon because the circumstances in which it can work are themselves uncommon. The difficulty, rather, is that s 14 retains an inherited remedy which is only partly aligned with present labour policy and supplies no criteria for its exercise. Whereas compensation is at least addressed in the EC Regs, reinstatement is left to broad principles derived largely from the common law’s aversion to specific performance and a body of ECT practice which—as far as I know—contains no example of its exercise. Unless the position is clarified by legislation, the remedy may therefore disappear in practice without any indication that Parliament intended that result.

Compensation for loss of income

I turn next to compensation for loss of income. Section 20(7)(b) of the ECA requires the ECT to calculate compensation for a wrongful dismissal dispute in accordance with the EC Regs, and reg 17A directs attention to the Second Schedule. Paragraph 1 of that Schedule provides that compensation “is to consist of” two components: the claimant’s loss of income determined under para 2, and the harm caused to the claimant by the respondent as a result of the wrongful dismissal determined under paras 3–7. Paragraph 2, however, says only that compensation for loss of income must not exceed three months of the claimant’s gross rate of pay on the date of dismissal. It prescribes a ceiling, but leaves unstated the measure by which the award is to be calculated.

That distinction matters because the expression “loss of income” does not, when read on its own, answer the questions upon which an award depends. It does not identify the counterfactual against which loss is to be measured, the period over which the loss is to be assessed, the effect of salary in lieu of notice, or the place of mitigation. Nor does it say whether the employee must prove that the income was lost because of the fact of dismissal, because the dismissal was without just cause or excuse, or because of the manner in which it was effected. Finally, para 2 leaves unclear whether the three-month ceiling merely limits loss which has otherwise been proved, or supplies a conventional range within which the ECT may make an award upon proof of the statutory wrong. As the cases show, each possibility has some support.

In approaching these questions, notice must be kept in the juridical place identified at [93]⁠–⁠[117] and [118]⁠–⁠[170] above. Although notice may affect an allocation of proof, or generate a separate contractual or statutory entitlement, it does not supply the content of just cause or excuse. Compensation for the statutory wrong may therefore extend beyond the notice period, subject to whichever measure the EC Regs require and, of course, the avoidance of double recovery.

Any apparent tension with that conclusion arises from para 7 of the Guidelines, which presumes a dismissal with notice not to be wrongful. For the reasons given at [99]⁠–⁠[102] and [132]⁠–⁠[165], that statement is most defensibly understood as an imprecise account of the employee’s residual burden in a notice-without-reason case. Once liability under s 14(2) has been established, it supplies no reason to confine compensation to notice.

There is, moreover, a more concrete reason why loss of income cannot invariably mean salary in lieu of notice, for such salary may be claimed independently under s 11(1) of the EA or the contract, without any claim under s 14(2). A claim heard in 2024 provides an example. The employee claimed three months’ salary in lieu of notice, abandoned the part exceeding the ordinary jurisdictional limit, and recovered $20,000 without bringing a claim for dismissal without just cause or excuse. The distinction was subsequently recognised in Goh Hui En Rebecca v IG Asia Pte Ltd [2025] 4 SLR 1477 (“Goh Hui En Rebecca”), where the court explained that a claim for salary in lieu remains fundamentally contractual, whereas a claim under s 14(2) goes further and compensates the losses and harm caused by the distinct statutory wrong (at [68]–[73]).

Section 12(7) of the ECA reinforces that distinction by placing salary-related disputes and “wrongful dismissal disputes” within separate aggregate claim limits. Of course, those claims cannot be used to obtain double recovery. Where salary in lieu has compensated a period for which income-loss compensation would otherwise be awarded, the same wages cannot be recovered twice. The rule against double recovery does not, however, make the juridical functions of the claims identical. If the loss-of-income component were always and wholly exhausted by salary in lieu, it would perform no work where notice had already been paid and add nothing where notice pay was separately claimed. The more natural inference is that the two claims may overlap as a matter of fact while continuing to perform distinct juridical functions.

Once notice is restored to its proper place, one must ask which counterfactual para 2 requires. One possible answer is simply factual: where an employee has made reasonable efforts to obtain comparable work, remains unemployed, and would have continued earning but for the dismissal, the wages thereby lost constitute loss of income. The fact of dismissal would then supply the relevant causal event. A second answer asks more narrowly what income was lost because the dismissal was without just cause or excuse. On that approach, the ECT must determine what would probably have occurred had the employer observed the statutory protection. The employment might have continued; a proper process might have been undertaken and produced dismissal later; or another lawful ground might have brought the relationship to an end. A third possibility arises where the manner of dismissal itself impairs the employee’s ability to secure replacement work. A false accusation or deliberate reputational act may then cause financial loss even if the employment would otherwise have ended.

The contractual minimum-obligation rule cannot select among those possibilities because, as explained at [186]⁠–⁠[189] above, s 14(2) imposes a statutory restraint additional to the notice power. The relevant counterfactual must therefore concern when and how the employment would probably have ended had the statutory wrong not occurred. The EC Regs do not identify that counterfactual, and the decisions have supplied materially different answers.

The contractual minimum-obligation approach was applied directly in a case from 2022. Although the employee’s fixed term remained unexpired, the employer could terminate the agreement upon one month’s notice. Relying on common law authorities concerning damages for wrongful dismissal, the tribunal treated one month’s salary as the relevant loss. Since the same $7,000 had been awarded as salary in lieu of notice, it awarded nothing under para 2 so as to avoid overcompensation. The statutory income component was, on that reasoning, exhausted by the employer’s minimum contractual obligation.

Three other decisions reached the same practical result by treating the two heads as overlapping and allowing recovery under only one. In a case from 2020, the tribunal awarded two months’ salary in lieu and set loss of income at nil because the two overlapped. An ECT decision in 2021 likewise treated the salary-in-lieu award as accounting for the relevant income loss. Another decision from 2021 approached the overlap from the other direction: it awarded loss of income and dismissed the overlapping claim for salary in lieu. Those outcomes may have been entirely correct if the only proved income loss occupied the same period as the notice entitlement. It is not entirely clear, however, whether later loss was excluded for want of proof, remoteness, or failure to mitigate, or whether the overlap was treated as exhausting the relevant compensation as a matter of law.

A different approach appeared in yet another case heard in 2021. Salary in lieu was awarded for the unserved portion of a notice period between 1 and 21 August 2020, while loss-of-income compensation was awarded separately for the period between the wrongful dismissal on 15 July 2020 and the beginning of that notice period. Since the periods did not overlap, the tribunal awarded both. Notice therefore operated as a credit against factual duplication, rather than as the juridical measure of the statutory loss.

The latter understanding is the one which I adopted in JGP. The employee there had received full notice, but remained unemployed despite sustained efforts to obtain comparable work. I therefore awarded three months’ income for the period following the expiry of notice, treating the continuing shortfall as loss caused by the fact and manner of dismissal (see JGP at [103]–[106] and [A.153]–[A.156]). Unlike the decision from 2022 (see [192] above), contractual notice did not exhaust the statutory head.

A further difference may be seen in two cases heard in 2024. In the former, only one day’s salary in lieu had been payable, the employment had lasted about 11 weeks, and the tribunal found that the employee had not been serious in seeking substitute work. It nevertheless awarded one month’s loss of income, without identifying the counterfactual or explaining why those facts yielded that period. In the latter, the employee remained unemployed, had looked for similarly flexible work in a small industry, and there was no evidence that her employment would otherwise have ended. The tribunal expressly found that, but for the dismissal, she would have remained employed and awarded the maximum three months. The measure in the first decision was left unstated; the second adopted continued employment as its starting point. Unfortunately, the EC Regs say nothing about either approach.

A more particular counterfactual was supplied recently in JIW v JIX [2026] SGECT 7 (“JIW”). The employee had made more than 30 applications, attended about five interviews, and remained without an offer. The tribunal found that, absent the wrongful dismissal, her probation would have been extended for at least three months. It therefore assessed her income loss at three months’ salary, or $34,500, although the prescribed claim limit reduced the recoverable sum to $30,000. That period was derived neither from notice nor from the actual time taken to obtain replacement employment. It arose from a judicial reconstruction of what the employer would probably have done had the statutory wrong not occurred.

The cases therefore disclose four possible methods: the employer’s minimum contractual obligation; notice treated as a credit against duplication; factual continuation of employment; and a more specific lawful counterfactual, such as extended probation. Since the EC Regs neither select among them nor explain the role of mitigation, their three-month ceiling limits the award only after a measure has been chosen.

A final complication arises if the manner of dismissal may affect the assessment of loss of income (cf [190] above). Where a false allegation or deliberate act impairs the employee’s reputation and therefore her employment prospects, para 2 of the Second Schedule would appear to take into account consequences which para 4(a) treats expressly as relevant to the separate award in respect of “harm”. It then becomes necessary to identify what the second monetary component is intended to compensate, and how it is to be kept distinct from loss of income.

Compensation for harm

Paragraph 1(b) of the Second Schedule describes the second component as compensation for “harm caused to the claimant by the respondent as a result of the wrongful dismissal”. Paragraph 3 permits the ECT to consider any relevant aggravating or mitigating factor, while para 4 supplies non-exhaustive examples. The aggravating factors include a humiliating or degrading manner of dismissal, physical harm, the use of a false accusation as the ground of dismissal, and a deliberate act adversely affecting the employee’s later employment. The mitigating factors include misconduct or poor performance, and insubordination which was taken into account by the employer but was insufficient to amount to just cause or excuse. Paragraph 5 then provides that the harm component is to comprise a “base amount” not exceeding two months of the employee’s gross rate of pay, which may be increased or reduced by up to 50% under paras 6 and 7.

One might expect these provisions to identify the kind of harm with which the component is concerned. I do not think that they do. A degrading dismissal and physical injury point towards personal and non-pecuniary consequences. A deliberate act affecting later employment points towards reputational and economic consequences. A false allegation may produce all three. Misconduct, poor performance, and insubordination, however, do not ordinarily reduce the injury which the employee has in fact suffered. They seem instead to bear upon the employee’s responsibility for the events, the gravity of the employer’s wrong, or the strength of the employee’s claim to vindication. Thus, the listed factors move among consequence, manner, culpability, and the employee’s responsibility without disclosing a single identified form of harm.

There is a related difficulty concerning the “base amount”. If it represents a conventional sum for the loss of a job protected against dismissal without just cause or excuse, para 5 does not explain why it prescribes only a maximum and no starting point. If the base is to be chosen by reference to the injury proved in the particular case, the Second Schedule does not identify which facts determine that base and which are reserved for the subsequent aggravating or mitigating adjustment. A third possibility is that the base operates as the monetary substitute for restoration and therefore recognises the employee’s loss of the protected employment relationship itself. If that is its function, its boundary with loss of income assumes particular importance, yet the Second Schedule does not state it. The expression “base amount” therefore assumes an object of valuation which the Schedule never identifies.

Some awkwardness also arises from the mechanics of adjustment. Paragraph 6 provides for an increase where there is any aggravating factor, or where aggravating factors outnumber mitigating factors. Paragraph 7 provides for a reduction in the converse terms. Where one aggravating and one mitigating factor are both present, the first limb of each paragraph appears to be engaged. It may be possible to apply both an increase and a reduction to the same base, but the Schedule says nothing about their order or the relative weight which either factor should receive. The cases have commonly met the problem by counting the factors, treating equal numbers as cancelling, or making a single net adjustment. Those methods may well yield sensible results, though they are not clearly prescribed by the text.

In an ECT decision from 2019, the tribunal treated loss of income as analogous to ascertainable financial loss and harm as closer to general or aggravated damages for non-pecuniary injury. The tribunal recognised that the Second Schedule supplied no method for selecting the base. It chose half a month’s salary, treated the employer’s overstatement of the accusation as aggravating, treated the employee’s contractual breaches as more strongly mitigating, and reduced the base by 30%, producing an award of $1,750. That account at least gives the component an identifiable function. It does not, however, explain why half a month was the appropriate starting point for that function.

The smaller bases selected in later decisions rest upon methods which are no more evident from the text. In an ECT case from 2021, the tribunal began at one-fifth of a month and reduced that sum by 50% because the employee had missed a meeting, had not properly worked from home, and had interfered with the employer’s equipment. The result was $740, or one-tenth of a month’s salary. In a case from 2021, the tribunal began with a sum representing 30% of one month and reduced it by 50% because of the employee’s non-cooperation and abrasive communications, producing an award equal to 15% of one month’s salary. The reasons explain the reductions, but do not identify the norm which made harm worth one-fifth of a month in the first case and three-tenths in the second.

Nor does the uncertainty disappear where the factors are neutral. A tribunal, in a case from 2021, found no aggravating or mitigating factor, selected a base of half a month, and made no adjustment. In a case from 2022, short service was used to set the base at one-quarter of a month. An abrupt and public dismissal was then treated as aggravating, while an argument and poor performance were treated as mitigating. Since there was one factor on each side, the tribunal made no adjustment. The first decision does not explain why a neutral case was worth half a month. The second treats length of service as relevant to the base, although the Second Schedule does not say whether that base values service, injury, or the loss of the protected position.

There is a more fundamental difference as to whether any harm must be proved. In a case from 2024, the tribunal awarded one month’s loss of income but no compensation for harm because the employee had adduced no evidence of it. In JIW, by contrast, the tribunal observed that it would have awarded two months for harm because the employment had ended suddenly and without justification and it was difficult to suppose that the employee had not experienced considerable distress (at [99]). The first approach treats harm as an actual consequence requiring evidence, whereas the second permits it to be inferred from the “wrongful dismissal” itself and the circumstances in which it occurred. The difference therefore arises at the anterior stage of deciding whether compensation for harm is available at all.

The uncertainty between the base and the adjustment may, in turn, be seen in another case from 2024. The employee was informed of her termination while hospitalised, pregnant, and close to giving birth. The tribunal selected the maximum base of two months, found the manner degrading and insensitive and the resulting anxiety substantial, and then applied the maximum uplift of 50%. It thereby awarded three months’ salary for harm. The severity of the circumstances readily explains why the award was high. What remains unclear, however, is whether the employee’s vulnerability and anxiety supplied the reason for choosing the maximum base, the reason for applying the maximum uplift, or both. Unless the base is given an identified function, the same facts may readily do work at both stages.

Nor is the difficulty confined to exceptional facts. In a case from 2025, the tribunal began with the maximum two-month base although it identified no aggravating feature, and reduced that sum by 25% for unsatisfactory performance. The reasons do not explain why the maximum was the proper starting point before the reduction. Conversely, a case from 2024 selected one month as the base and then stated that it would increase the sum “by another month”, producing an award of two months. If that description is read literally, the increase was 100% of the base, although para 6 permits no more than 50%. It remains unclear whether the description was merely imprecise or whether the method departed from para 6.

In JGP, I adopted a different approach. I treated two months as a conventional base recognising the affront to dignity, reputational sting, and disruption inherent in dismissal without just cause or excuse. I found no aggravating factor, but reduced the amount by one month because the employee’s negligent medical claims had materially contributed to the dismissal (JGP at [107]–[108] and [A.157]–[A.161]). On that approach, the base performs a normative function capable of application without proof of psychiatric or other personal injury. That function differs from the one adopted in the decision which required evidence of actual harm. It also leaves unexplained why the statutory maximum, rather than some lower point within the permitted range, was chosen.

Drawing those decisions together, the differences cannot be attributed merely to the breadth of the discretion or the variety of the underlying facts. The base has ranged from one-fifth, one-quarter, three-tenths, and one-half of a month, through one month, to the maximum of two months before adjustment. More importantly, tribunals have variously treated it as compensation for non-pecuniary injury, an amount affected by length of service, a response to proven distress, and a conventional recognition of the wrong itself. Some decisions require evidence of harm while others infer it; some facts determine the base, others determine the adjustment, and still others appear to determine both. Without an identified object of valuation, the discretion cannot be exercised by reference to a common measure.

That difficulty is compounded by the absence of a stable boundary between the two monetary components. A deliberate act which impairs future employment is an express aggravating factor under the harm component, while the resulting difficulty in obtaining work may also produce loss of income. A false accusation may cause reputational injury and distress and, at the same time, explain why the employee remains unemployed. If para 2 responds to the manner as well as the fact of dismissal, the overlap is wider still. The Second Schedule contains no rule assigning such consequences to one component or the other, and does not explain how the same fact may properly bear upon both without producing double recovery.

Comparative approaches

At this stage, it is useful to consider two other statutory regimes, though only by way of comparison because their language, history, and institutional settings differ from ours. Each nevertheless separates compensation for the statutory loss from contractual damages and, in doing so, identifies the work performed by its different remedial heads.

Under the United Kingdom’s unfair dismissal regime, the basic award and the compensatory award are governed separately (see Blackstone’s at paras 33.01–33.03, 33.49–33.51 and 33.67–33.68). The basic award is ordinarily calculated by reference to age, length of service, and a capped week’s pay (UK ERA, s 119), and conventionally recognises the loss of job security which follows from unfair dismissal. That function had appeared before the introduction of the modern basic award in Norton Tool Co Ltd v Tewson [1972] ICR 501, which recognised compensation for the loss of redundancy protection. The compensatory award, for its part, is such amount as the tribunal considers just and equitable having regard to the loss sustained by the employee in consequence of the dismissal, in so far as that loss is attributable to the employer’s action (s 123(1)). Thus, the first head assigns a conventional statutory value to the loss of the job, while the second addresses the financial consequences which the dismissal has in fact caused.

That separation allows the compensatory inquiry to proceed upon an identified counterfactual. The tribunal asks what earnings and benefits the employee would probably have received if the unfair dismissal had not occurred, subject to mitigation, contribution, and the possibility that a fair dismissal would have occurred at a later time. The last possibility is reflected in Polkey v A E Dayton Services Ltd [1988] AC 344, under which an award may be reduced to reflect the chance that a fair process would have produced dismissal in any event. The inquiry may be difficult, and its result may depend upon an estimate. Its object is nevertheless clear: the employee’s probable financial position absent the statutory wrong, rather than the employer’s minimum contractual obligation to give notice (also see Blackstone’s at paras 33.71, 33.80–33.84, 33.145–33.168 and 33.172–33.183).

The same separation can be seen in the treatment of overlapping and non-pecuniary losses. Contractual notice damages may be awarded for the notice period and the statutory compensatory award assessed thereafter, or the statutory award may be assessed across the whole period with the contractual sum deducted. Either method avoids double recovery while preserving the distinction between the causes of action (see, eg, Shifferaw v Hudson Music Co Ltd (UKEAT/0294/15, 30 August 2016); Harvey, ch 9 at paras [515]–[517]). Further, injury to feelings does not form part of the ordinary compensatory award for unfair dismissal (see Dunnachie v Kingston upon Hull City Council [2005] 1 AC 226), although financial loss caused by an illness attributable to the dismissal may do so (see Dignity Funerals Ltd v Bruce [2005] IRLR 189). The line is therefore drawn by reference to the kind of loss which the compensatory head addresses, rather than by asking only whether the employee has suffered harm in some broad sense (also see Blackstone’s at paras 33.94, 33.140–33.142 and 33.186–33.191).

Malaysian law adopts a different structure, though it preserves the same need for separation. The statutory jurisdiction begins from reinstatement and, where that remedy is not ordered, the Industrial Court may award compensation in lieu together with back wages. Compensation in lieu is ordinarily assessed by reference to length of service and serves as the monetary substitute for restoration. Back wages address the earnings lost following dismissal and are usually calculated to the conclusion of the hearing, subject to the statutory limits and deductions prescribed by the Second Schedule to the M’sian IRA. Both heads extend beyond the notice which the employer might have given under the contract.

The cases illustrate the distinct functions performed by those two heads. In Unilever (M) Holdings Sdn Bhd v So Lai @ Soo Boon Lai [2015] 4 MLJ 326, the employee had passed the compulsory retirement age by the date of the award and could no longer lawfully be reinstated. Compensation in lieu was therefore unavailable because the primary remedy for which it substituted was itself unavailable. In Koperasi Serbaguna Sanya Bhd (Sabah) v Dr James Alfred (Sabah) [2000] 3 CLJ 758, post-dismissal earnings were relevant to back wages, while compensation in lieu continued to reflect the loss of the position as an interest distinct from interim earnings. The expansion of compensation to include future earnings in R Rama Chandran v Industrial Court of Malaysia [1997] 1 MLJ 145, and the later rejection of future earnings as a general head in Telekom Malaysia Bhd v Ramli Akim [2008] 1 MLJ 770, further show why the present Second Schedule separately regulates and caps back wages and excludes compensation for future earnings (see Donovan Cheah, Navigating Employment Disputes in Malaysia (LexisNexis, 2nd Ed, 2024) at paras 5-26–5-35; Sharifah Suhanah Syed Ahmad, Employment Law in Malaysia (LexisNexis, 2012) at pp 120–126).

The two regimes therefore achieve juridical order by different routes. Malaysian law separates the monetary substitute for restoration from lost wages, while United Kingdom law separates a conventional award for job security from proved financial loss and contractual notice damages. Each then regulates mitigation, contribution, overlap, and the possibility of a later lawful termination by reference to the interest protected and the counterfactual assumed.

The Second Schedule supplies no equivalent ordering. It does not identify compensation as the monetary substitute for reinstatement, state the counterfactual for loss of income, or explain what the base amount for harm represents. Its aggravating and mitigating factors move among consequence, culpability, manner, and the employee’s responsibility. Those omissions explain much of the methodological variation in the decisions considered at [192]⁠–⁠[212] above.

The result is that materially similar dismissals may be approached from different assumptions about notice, counterfactual periods, proof of harm, and the base amount, even before the facts peculiar to the employee are considered. That, in my view, is the major remedial trouble: the EC Regs prescribe ceilings and modifiers without adequately identifying either the injury being compensated or the normative position against which loss is to be measured. Those questions return the analysis to the juridical foundations of s 14(2).

The juridical foundations revisited

The content of a legal right

At [56] above, I defined the juridical foundations of s 14(2) by reference to the protection’s binding force and content. In so far as the former is concerned, the reasons given at [58]⁠–⁠[85] establish that the protection derives its authority from the words enacted by Parliament, and that neither the common law label “wrongful dismissal” nor the Guidelines supplies another source. What remains to be considered is whether the rules concerning proof and remedy have altered the content identified at [90]⁠–⁠[117].

To answer that question, a short account of what a legal right comprises will assist. In Hohfeld’s scheme, a right in the strict sense is a claim held by one person against another, correlated with a duty borne by that other person. A privilege, power, and immunity describe different relations, correlated respectively with a no-right, liability, and disability (see Wesley Newcomb Hohfeld, “Some Fundamental Legal Conceptions as Applied in Judicial Reasoning” (1913) 23(1) Yale Law Journal 16 (“Hohfeld (1913)”) at 30–32; Brian H Bix, Jurisprudence: Theory and Context (Sweet & Maxwell, 8th Ed, 2019) (“Bix”) at pp 137–139; MDA Freeman, Lloyd’s Introduction to Jurisprudence (Sweet & Maxwell, 9th Ed, 2014) at pp 337–339; Raymond Wacks, Understanding Jurisprudence (Oxford University Press, 6th Ed, 2020) at pp 297–300).

Accordingly, a claim-right may be stated by identifying its holder, the person against whom it is held, and the act, omission, or state of affairs to which it relates: A has a right against B for X (Bix at 139). The governing rule and facts determine when that relation arises and when it is infringed. Although the evidence by which those facts are proved and the consequence which follows infringement are connected with the right, neither ordinarily forms part of the conduct which the right required from the duty-bearer.

Lawyers also use “right” more broadly to describe an aggregate of incidents. A claim, liberty, power, and immunity may together form one legal institution, and that wider usage will often give a fuller account of the right-holder’s position (Bix at 138). Even then, however, the incidents must be kept distinct. The power to begin proceedings does not define the claim which those proceedings vindicate, and the order made at their conclusion does not necessarily define the conduct required before they began.

That distinction continues after breach, for Hohfeld separated the primary claim existing before violation from the secondary claim which may arise thereafter, just as modern accounts distinguish a primary duty from the later right of action which supplies redress (see Wesley Newcomb Hohfeld, “Fundamental Legal Conceptions as Applied in Judicial Reasoning” (1917) 26(8) Yale Law Journal 710 (“Hohfeld (1917)”) at 752–753; J E Penner & E Melissaris, McCoubrey & White’s Textbook on Jurisprudence (Oxford University Press, 5th Ed, 2012) at p 114). Procedure or remedy may weaken the protection which a right affords in practice, but that does not, by itself, change what the primary duty required.

I recognise, of course, the limit of that proposition. Parliament may authorise an interference upon payment of compensation, in which case the legislation creates a liability-rule structure instead of a prohibition followed by redress. Remedial provisions may also illuminate the substantive provision which they accompany. The relationship must therefore be found in the particular legislation and cannot be fixed by conceptual analysis alone (see Charles Lowell Barzun, “Legal Rights and the Limits of Conceptual Analysis: A Case Study” (2013) 26(2) Ratio Juris 215 at 225–232).

The form in which s 14(2) is expressed requires some care. Its words give an employee who considers that she has been dismissed without just cause or excuse the power to lodge a claim, while s 14(3) gives the ECT power to grant relief if satisfied that the condition is established; they do so without using the ordinary language that an employer “must not” dismiss in that way. If those words are read alone, they might describe only powers and liabilities, with dismissal without just cause or excuse serving as the event which permits intervention. Whether the scheme also presupposes a primary claim must therefore be ascertained from the provisions read as a whole.

The right conferred by section 14(2)

In my view, the scheme does presuppose such a claim. As explained at [92]⁠–⁠[95], the definition of dismissal and the surrounding provisions subject the employer’s effective termination to a separate statutory restraint. Section 14(3) requires the anterior finding that the employee “has been dismissed without just cause or excuse” before relief is considered; reinstatement is restorative in form; and the EC Regs speak of loss and harm resulting from the wrongful dismissal. Read together with the enacted standard, these features treat the unjustified dismissal as a wrong which precedes the remedial decision.

The primary relation may, accordingly, be stated in this way: the employee holds a claim against her employer that she not be dismissed without just cause or excuse, and the employer bears the correlative duty. The wider scheme gives the employee power to lodge a claim, subjects the employer to the ECT’s jurisdiction, and gives the ECT power to make an authorised order if it is satisfied that the duty was breached. Because each relation arises upon its own conditions, the discretion preserved by “may” in s 14(3) means that the antecedent claim does not carry an unconditional entitlement to either remedy (see Hohfeld (1913) at 54; Hohfeld (1917) at 760–764).

The same analysis also explains how an employer’s legal power to end the relationship can coexist with the statutory duty. A power is the capacity effectively to alter legal relations, and need not entail an absence of duty concerning its exercise. An employer may therefore terminate the relationship effectively while incurring contractual or statutory consequences for doing so. Accordingly, although the employee possesses no immunity from dismissal or unconditional entitlement to remain employed, she has a claim that she not be dismissed in the absence of just cause or excuse (see Hohfeld (1913) at 52–53).

On that analysis, the content of the claim is supplied by the three-part inquiry stated at [92]⁠–⁠[95]: whether there was a basis capable of amounting to cause or excuse; whether it possessed the necessary connection with the dismissal; and whether it was just in the factual and industrial setting. Notice, the communication and truth of a reason, the category into which it falls, and the process preceding dismissal may bear upon those matters. The remedial consequences which follow do not add further elements to the antecedent breach.

Section 27(2), for its part, allocates the risk of non-persuasion without changing that standard of conduct. Paragraph (a) places the composite question of just cause or excuse upon the employer where its conditions are met; para (b), on the construction adopted at [121]⁠–⁠[122], requires proof of the stated poor-performance or misconduct ground in its full sense; and, in the residual cases, the employee bears the legal burden of establishing the statutory negative. Those allocations may decide the proceedings where the evidence remains insufficient, but they do not alter the duty which governed the employer when the dismissal occurred.

It follows that a failure of proof resolves the issue against the party which bears the relevant risk, though it does not necessarily establish the opposite historical fact. In particular, an employee’s failure to discharge the residual burden means only that she has not established a violation in those proceedings; it confers no general privilege upon employers to dismiss without just cause or excuse. The same distinction must be kept in mind where a stated ground fails under para (b), for the result depends upon the scope of that paragraph and the case which was advanced (see [106]⁠–⁠[109] above).

The reconciliation proposed at [141]⁠–⁠[164] preserves that order. Section 108 concerns an identified historical fact especially within the employer’s knowledge; circumstantial proof, a moving evidential burden, and an adverse inference concern the rational establishment of facts; and ss 20 and 21 permit the ECT to obtain and test relevant material. None creates a general rule requiring every employer to formulate and prove a reason. I recognise that the remaining gaps may make an NNR claim difficult to establish. That practical difficulty in proof does not, however, change the norm which governed the dismissal.

So too, the remedial provisions preserve the same sequence, for s 14(3) begins with a finding that the employee has been dismissed without just cause or excuse, after which the ECT considers reinstatement or compensation. Those remedies may illuminate the interest protected by s 14, but the scheme does not use them as alternative definitions of the wrong.

In this way, reinstatement gives the protected relationship an object capable of restoration, though its rarity and the discretion in s 14(3) do not convert the antecedent claim into an absolute right to continued employment (see [172]⁠–⁠[183] above). Salary in lieu and compensation for loss of income may overlap factually while performing different juridical work, and the unresolved counterfactual under para 2 affects the assessment of relief rather than the elements of liability (see [184]⁠–⁠[199] above). The same is true of the uncertainty concerning what compensation for harm represents and how that component overlaps with income loss (see [200]⁠–⁠[212] above).

It follows, in my view, that the juridical foundations have not changed: the protection remains a statutory claim held by an employee against her employer that she not be dismissed without just cause or excuse. The ECA governs the invocation and adjudication of that claim and the specified burdens of proof, while s 14(3) and the EC Regs govern the powers available after breach and, where compensation is sought, the calculation of the award. Although those later incidents change the parties’ particular legal positions as the relevant events occur, they do not retrospectively alter the antecedent relation.

Seen in that way, the principal troubles become manageable in the manner already explained. The Guidelines must yield where they depart from the enacted standard (see [62]⁠–⁠[118] above); the statutory negative may be proved under the burdens and evidential principles described at [119]⁠–⁠[171]; and remedies must remain analytically distinct from notice and from the elements of breach (see [172]⁠–⁠[222] above). Nevertheless, a substantial residue remains because the Guidelines continue to occupy a prominent but uncertain place; the ECA leaves several common evidential configurations incompletely addressed; and the EC Regs do not identify the governing objects and counterfactuals with sufficient clarity.

Some troubles can therefore be managed by restoring each source, burden, and remedy to its proper juridical position. Others arise from omissions or tensions which a court may accommodate, but cannot remove without supplying rules which Parliament did not enact. The underlying protection remains simple in form, although its present administration has accumulated a considerable number of foundational difficulties. To understand how those difficulties came to be, it is necessary to examine the way in which these rules were placed together. I turn now to that history.

Deconstructing the major troubles

In the course of the preceding paragraphs, one will have gathered just how troubled the law is. It is of no use to anyone if these grounds merely pointed out those troubles without offering the best solution that I am able to devise within the remit of my judicial function. Before any solution can be furnished, however, we first need to understand how the law came to find itself in this state, and that requires us to turn to history. I am, of course, mindful that how legal officials and actors—including judges—think about history can vary greatly, and there are certainly a multitude of perspectives one can take. For ease of explanation as to how I think the history matters, I rely simply on the contrast between Maitland and Dicey.

Shortly after the publication of “Canon Law in England” (1896) 11(43) The English Historical Review 446, Maitland wrote to Dicey from Horsepools, his home in Gloucestershire near Stroud. Dicey’s own letter has not survived, so the exact reason Maitland replied as he did is not known. In any event, having thanked Dicey for his letter and a book that he was to be sent—likely Dicey’s Digest of the Law of England with Reference to the Conflict of Laws (Stevens & Sons, 1896)—Maitland then continued (see The Letters of Frederic William Maitland: Volume II (P N R Zutshi ed) (Selden Society, 1995) (“Maitland’s Letters”) at 104–105):

While Maitland expressed admiration for Dicey’s views on legal history, it must be said for those who are less familiar, the two scholars could not have differed more in how they saw the relation between law and history. Dicey, for his part, was juristic and analytical, seeking clarity, classification, and general propositions—an approach which served him well where the material was legal in character and capable of being arranged in that way. His Conflict of Laws was indeed exactly such a work—a skilled lawyer’s attempt to bring order to a difficult, relatively new and unsettled field by reducing it to clear principles, propositions and rules. That outlook owed much to Austinian jurisprudence and probably to the broader 19th-century aspiration to treat law as a coherent science (see Richard A Cosgrove, The Rule of Law: Albert Venn Dicey, Victorian Jurist (Macmillan, 1980) (“Cosgrove”) at 23–28 and 166–167; for a cognate expression of that broader aspiration in late 19th-century American legal thought, see, eg, Morton J Horwitz’s work in The Transformation of American Law, 1780–1860 (Harvard University Press, 1977) at 256–257 and The Transformation of American Law, 1870–1960: The Crisis of Legal Orthodoxy (Oxford University Press, 1992) at 6, 9–15, and 199).

Where history was concerned, however—with its need to reckon with so much of society’s plurality, complexity, and contingency—its unwieldiness and refusal to yield to coherent systemisation seemed to leave Dicey regularly dissatisfied. To him, the value of history, if any at all, seemed to lie in its ability to produce clear understanding for present use. If past events could be arranged so as to reveal general principles, improve judgment, or make legal development more intelligible, history had value. But if the history only produced ambiguity, complexity, qualification, and resistance to generalisation, it appeared to him to be of far less use. In the end, even if it would be slightly reductive to put it this way: Dicey’s attitude to history was very closely connected to his lawyerly search for order, principle, and certainty. He was drawn to history when it could be made to reinforce those things, but not when it unsettled them (see Cosgrove, Ch 8 generally or 171–179 specifically).

Maitland, by contrast, did not treat legal history as valuable because it supplied answers or even lessons for the present. Indeed, as the quoted passage makes clear, his view was the exact opposite (on this, also see T F T Plucknett, “Maitland’s View of Law and History” (1951) 67(2) Law Quarterly Review 179 at 188–189). If history was to do its “liberating work”, it had to be “as true to fact as it [could] possibly make itself”, and it would not be true to fact if it began by asking what lessons it can teach. Legal history then, on this view, matters because it reminds one that law is made by fallible human beings, in particular circumstances, and for reasons which may later be forgotten, misunderstood, or re-described. Its utility thus lies not in using the past to justify the decisions of—or as a model for—the present, but in freeing the present from the assumption that the legal ideas, categories, phrases, and institutions we have inherited are natural, inevitable, or incapable of being otherwise.

It is in that sense that Maitland’s observation of each generation having “free hands” is useful here. It is also why, in my view, the Diceyan perspective on when history has utility, especially in relation to law—one commonly adopted by those in the legal profession—is intellectually limiting. For present purposes, the value of history lies not in its ability to arrange the past so as to render it coherent with the present, and thereby make the structure and application of the modern law appear intelligible and defensible on its own terms. Instead, its value lies in its capacity to expose equivocation, ambivalence, and discontinuity where later legal forms may otherwise give the impression of simplicity and order.

One must however keep in mind that, liberating though Maitland’s perspective may be, it necessarily comes with a less comforting corollary; one which tends to underlie the plurality, complexity, and contingency with which history often has to cope. That is, if each generation has free hands to shape its own law, it cannot be taken for granted that those hands will necessarily be steady, used coherently, thoughtfully, or with unity over time. Legal ideas come, go, transmute, and transform. Legal officials and actors may borrow them for one purpose, retain them for another, narrow them by procedure, expand them by rhetoric, rename them by practice, re-shape them by habit, or harden them by adjudication. Often, the language of the law survives even as the assumptions which first gave such language meaning—and which animated its application—recede, and the same words may continue in use long after the world in which they first made sense has changed. There are, indeed, so many examples of this general phenomenon scattered throughout the law (for a recent and illuminating historical account of how inherited legal ideas can outlive, and later obscure, the assumptions that once made them intelligible, see Astron Benjamin Douglas, Property and Contract in the History of the Law of Breach of Confidence until 1825 (14 June 2025) (unpublished Doctor of Philosophy thesis, University of Cambridge, archived at Corpus Christi College, Cambridge)).

Regrettably, that is the difficulty which now attends s 14(2) of the EA. Basic legal training makes it easy and therefore tempting to treat s 14(2) as providing for the existence of a discrete statutory cause of action whose content may be ascertained simply by reading its present text, its supporting legislative framework in the ECA and—on the back of s 35 of the ECA—the Guidelines. That could have been enough had the free hands of the relevant legal officials and actors been used to develop the law differently. But, as I have taken some pains to show from [57]⁠–⁠[241] above, as things have turned out, that would be too thin an account of what the provision is, and of why it now gives rise to the many difficulties that it does.

The inherited settlement

That thinness supplies the point from which this history must begin. My view, put shortly, is that when Parliament enacted the 2018 Amendment Act and transferred the determination of claims under s 14(2) of the EA from the Minister to the ECT, the settlement which those words had inherited was largely left out of account. I do not mean that the age of the EA, or its place within Singapore’s wider programme of industrialisation, went wholly unnoticed, for both matters were mentioned (see, eg, 2018 Debates at 3.56pm (Mrs Josephine Teo, Minister for Manpower) and 4.18pm (Mr Patrick Tay Teck Guan)). What appears to have fallen from view was the more particular history which gave the protection its content: why an employee had been permitted to complain that she was “dismissed without just cause or excuse”; what those words had come to protect; and how that protection had been administered before it was recast as a modern judicial claim in the ECT.

Once that settlement had disappeared from view, contemporary administrative practice and the Guidelines could naturally appear to furnish the content which s 14(2) lacked. As I explained at [90]⁠–⁠[117] above, the divergence is most apparent where an employer gives notice without reasons: para 9 requires proof of a positively wrongful explanation, whereas the words Parliament enacted direct the inquiry towards the existence of just cause or excuse. It is therefore necessary to recover the older settlement if one is to understand why those inquiries differ in spirit.

I begin with the industrial and legal conditions which preceded the Industrial Relations Ordinance 1960 (“1960 Ordinance”), and with the collective machinery which that Ordinance established. From there, I turn to the dismissal awards, the recasting of the protection in 1968, and its extension through s 14 in 1973 to employees who were not unionised. Taken together, those matters disclose what Parliament inherited in 2018.

The industrial and legal setting before 1960

It would take this account much too far back to begin with the reception of the law of master and servant. For present purposes, it is enough to keep in mind that the ordinary law governing the individual employment relationship was received as part of English law and rested, in its classical form, upon assumptions of voluntariness, individualism, and equality of bargaining power. By the middle of the 20th century, however, the difficulty of applying those assumptions to industrial employment had become increasingly apparent (on this, see generally the writings of Sir Otto Kahn-Freund, eg: “Labour Law” in Law and Opinion in England in the 20th Century (Morris Ginsberg ed) (Stevens & Sons, 1959) at p 215, especially pp 223–225; Labour and the Law (Stevens & Sons, 1972) at pp 4–9, especially p 8).

Thus, in his contemporary survey of the law as at 1968, Tan Pheng Theng observed that the courts had developed no distinct body of private law around the peculiar problems of employment, with the result that the relationship continued to be governed by ordinary contract even as that law became less adequate to the conditions which industrial employment had produced (see Tan Pheng Theng, “A Conspectus of the Labour Laws of Singapore” (1968) 10(2) Malaya Law Review 202 (“TPT (1968)”) at pp 204–206). Legislation had, of course, long intervened in particular ways—principally through immigration, the regulation and protection of identified classes of labour, wages, hours, health, and enforcement of the individual contract. What had yet to emerge was a developed legal order for collective industrial relations.

The immediate foundations for such an order were laid by three enactments passed in 1940 and 1941. Prompted by a series of serious strikes and the authorities’ difficulty in finding responsible representatives with whom to negotiate, the Trade Unions Ordinance 1940 provided for the compulsory registration and regulation of unions of employers and employees, although, because of the War and the Japanese Occupation, it did not come into operation until May 1946. The Industrial Courts Ordinance 1940 (“Industrial Courts Ordinance”) then supplied peaceful machinery for resolving trade disputes through a standing Industrial Court, an appointed arbitrator, or a board of arbitration. The Trade Disputes Ordinance 1941 (“Trade Disputes Ordinance”), in turn, regulated the legality of strikes and lockouts, industrial intimidation and picketing, conspiracy, and certain breaches of contracts of service. In this way, the three Ordinances recognised organised labour, created a means by which industrial disputes might be adjudicated, and marked out the legal field in which collective pressure could be exercised (see TPT (1968) at pp 220–223; Oun Hean Loh, Industrial Relations in Singapore: Practice and Perspective (World Scientific, 2018) (“OHL (2018)”) at pp 13–16).

The limitation of the Industrial Courts Ordinance lay in the voluntary character of the machinery which it created. The Controller of Labour could refer a trade dispute to the standing Court with the parties’ consent, while the alternative routes to an arbitrator or board likewise depended upon an agreement to submit. Where an industry already possessed arrangements for settlement by conciliation or arbitration, the statutory machinery could be used only after those arrangements had failed and both parties had consented. Further, the Ordinance gave no award binding force in law and supplied no legal framework for collective agreements. Thus, although the State had made an adjudicative forum available, its use and efficacy remained dependent upon the parties’ agreement. Once negotiation and voluntary conciliation had failed, an unwilling party could resist adjudication and leave the dispute to be resolved by the relative economic power which each side could exercise (see TPT (1968) at pp 220–221).

That limitation assumed particular importance in the unsettled years which followed the War. The Japanese Occupation had destroyed the small pre-war trade-union movement, while the conditions which followed liberation—scarcity, inflation, unemployment, low wages, poor working conditions, and the re-emergence of political organisation—produced an exceptionally turbulent conjunction of industrial and political conflict. In 1946, there were 47 recorded work stoppages involving 50,325 workers and 845,637 man-days lost; in 1947, there were 45 stoppages and 492,708 man-days lost. Many unions were influenced by, or associated with, the Malayan Communist Party, and strikes could therefore serve political as well as industrial objects. The line between a contest over employment conditions and the wider struggle against colonial authority became, in consequence, difficult to maintain. I mention that conjunction because the insufficiency of voluntary machinery cannot sensibly be understood apart from the wider conflict within which it had to operate (see David B Johnson & Robert Quek Sio Chua, “The Role of the Industrial Arbitration Court in Singapore’s Economic Development” (“Johnson and Quek”) in The Law of Industrial Relations in Singapore (Kan Ting Chiu ed) (University of Singapore Law Society, 1970) (“LIRS (1970)”) at pp 113–114 and Table I; OHL (2018) at pp 16–18).

The Emergency declared in 1948 brought that first post-war wave to an end, as unions were deregistered or weakened, leaders detained, and work stoppages reduced to only 25 in the six years between 1949 and 1954. That relative quiet did not, however, mean that the underlying industrial problems had been resolved. When political activity and union organisation revived in the middle of the 1950s, industrial conflict returned with extraordinary force. In 1955 alone, there were 275 stoppages involving 57,433 workers and 946,354 man-days lost. The Hock Lee Bus dispute, which began as a contest over recognition and working conditions, culminated in violence on 12 May 1955 in which four persons were killed and 31 injured. It has since become perhaps the most familiar illustration of how quickly an industrial grievance could become a contest over political authority and public order (see Johnson and Quek in LIRS (1970) at p 113; OHL (2018) at pp 17–18).

By the time Singapore attained internal self-government in 1959, that gap had become part of a much larger economic and political problem. The new Government had come to office with the support of organised labour and with a programme directed towards improving wages and working conditions. At the same time, unemployment was serious, the entrepôt economy was vulnerable, and industrial expansion required capital, confidence, and continuity of production (see Woon Kin Chung & Loo Ya Lee, 50 Years of Singapore’s Productivity Drive (World Scientific, 2018) (“Woon and Loo”) at pp 41–44). The Government therefore expressed its object as “industrial peace with justice”: workers were to receive a fair share of the wealth which their labour helped to create, while conditions conducive to industrial expansion were established. That settlement was to be pursued through a legal process in which bargaining was encouraged, conciliation supplied, and an unresolved dispute finally determined, thereby constraining the unregulated exercise of economic power by employers as well as recurrent industrial action by labour (see Singapore Legislative Assembly Debates, Official Report (10 February 1960) vol 12 at cols 140–156 and 189–190 (Mr K M Byrne, Minister for Labour and Law, and Mr Lee Kuan Yew, Prime Minister); TPT (1968) at pp 224–228).

The settlement made by the 1960 Ordinance

Against that background, the central innovation of the 1960 Ordinance may be seen more clearly. The Legislative Assembly passed the Bill on 13 February 1960; the Ordinance came into operation on 15 September; and the Industrial Arbitration Court (“IAC”) was inaugurated on 24 October. What began was the unique “experiment”, as Kleinsorge described it, of combining free collective bargaining with compulsory arbitration. Drawing upon, while adapting, the system used in Western Australia, the Ordinance placed the course of collective industrial relations within a legal framework extending from the initiation of bargaining to the enforcement of an agreement or award. Freely negotiated agreement remained the preferred outcome, and the Ordinance created several opportunities for the parties to reach one. Its distinctive work began when those opportunities were exhausted, for the dispute no longer had to remain an indefinite contest of industrial strength: the sequence which began with bargaining and continued through conciliation could, as a final resort, end in a binding determination (see Paul L Kleinsorge, “Singapore’s Industrial Arbitration Court: Collective Bargaining with Compulsory Arbitration” (“Kleinsorge”) in LIRS (1970) at pp 55–72; TPT (1968) at pp 224–228).

It is useful, therefore, to keep in view the sequence contemplated by Part III of the 1960 Ordinance. Under s 16, a trade union of employees or an employer could serve a notice setting out proposals concerning industrial matters and inviting negotiation. If that invitation was not accepted within seven days, s 18 required the Commissioner for Labour to consult the refusing party with a view to persuading it to negotiate. A persistent refusal then required the Commissioner to notify the Minister and, unless otherwise directed, the Registrar that a trade dispute existed. Where negotiation was accepted but no complete agreement was reached within the prescribed period, a party could notify the Commissioner under s 19, who might assist the parties through conciliation. If he formed the view that negotiation was unlikely to produce agreement, or conciliation unlikely to assist, s 20 again required notification that a trade dispute existed. Even then, the possibility of agreement was not exhausted, since s 21 permitted the Minister to compel attendance at a private conference for conciliation or further conciliation, with non-attendance constituting an offence.

Where the parties reached agreement, it was reduced to writing, delivered to the Registrar, and brought before the President of the IAC for certification under s 23. Certification could be refused on public-interest grounds and had to be refused if the statutory requirements—including duration and, unless inappropriate, a procedure for referring disputes concerning the agreement’s operation to a referee—were not satisfied. Once certified, s 24 deemed the agreement to be an award, thereby supplying the legal force which collective agreements had lacked under the earlier system and making it enforceable against the parties and persons whom the Ordinance bound. Voluntary bargaining was, in that way, brought within the legal order, and its result became one of the principal means through which that order took effect.

If agreement remained impossible, Part IV supplied three routes by which the IAC could acquire cognisance of the dispute: a joint written request by all parties; a direction by the Minister published in the Gazette; or a proclamation by the Yang di-Pertuan Negara (later, President) that special circumstances made arbitration essential in the public interest (s 28). Only the first required agreement, while the remaining two supplied the compulsory endpoint which the earlier machinery had lacked. Once cognisance was acquired, s 29 required the IAC carefully and expeditiously to inquire into and investigate the dispute, together with every matter affecting it and its just settlement, and then to determine it by arbitration. The determination was industrial in object and range, for s 31 permitted regard to the interests of those immediately concerned, the community as a whole, and even the condition of Singapore’s economy.

The IAC which performed that function was deliberately unlike an ordinary court. It ordinarily sat as a tripartite body comprising its President and members drawn from employer and employee panels, while advocates and solicitors could not appear as of right. The Court controlled its own procedure, was not bound by the then-Evidence Ordinance, and was required by s 56(1)(c) to act according to “equity, good conscience and the substantial merits of the case without regard to technicalities and legal forms”. To those ends, it could take evidence on oath, compel witnesses and documents, proceed in a party’s absence, sit in private, seek expert assistance, inspect premises, join parties, and do what was necessary or expedient for an expeditious and just determination (ss 56–63). Its equally broad remedial authority appeared in the original s 38(a), which allowed an award to contain any matter or thing thought expedient to settle the dispute or prevent another, although employers continued to dispute whether those general words reached dismissal and authorised the recreation of the employment relationship.

Parliament removed that ambiguity through the Industrial Relations (Amendment) Act 1965, which was passed on 31 December 1965 and came into operation on 18 February 1966. The Act amended s 38(a) to refer expressly to matters relating to dismissal and reinstatement, and added to s 57(1) the power to order reinstatement or grant such other relief as the IAC considered desirable. When the Bill was tabled, the then-Minister for Labour explained that its purpose was “[t]o remove any ambiguity that may exist”, following work undertaken by the IAC President, NTUC, Singapore Employers’ Federation, and Ministry officials (Singapore Parliamentary Debates, Official Report (30 December 1965) vol 24 at cols 776–778 (Mr Jek Yeun Thong, Minister for Labour); also see Singapore Parliamentary Debates, Official Report (1 August 1968) vol 27 at cols 778–780 (Inche Abdul Aziz Karim) (“1 August 1968 Debates”)). The chronology is important because the express provisions were added after the IAC had already made reinstatement awards, and were intended to place beyond dispute the breadth of authority which the IAC understood the Ordinance to confer.

Taken together, these provisions produced a model in which collective bargaining supplied the primary means by which terms were settled, although bargaining now occurred alongside compulsory conciliation and the possibility of an adjudication capable of binding both sides without their agreement. An agreement emerging from that process could acquire the force of an award; where the IAC itself had to decide, its award was binding, final, and capable of regulating future conduct despite inconsistent rules of private law. The character of the adjudication was correspondingly different, because the IAC sought a settlement of the industrial relationship upon its substantial merits and could grant relief which an ordinary court determining a pleaded contractual dispute would not ordinarily supply. Thus, as Tan Pheng Theng observed in 1968, reinstatement overcame the equitable reluctance to decree specific performance of a personal contract and allowed the award, in that sense, to override the common law (TPT (1968) at pp 226–228). It was through this combination of bargaining, conciliation, adjudication, and remedy that the aspiration of industrial peace with justice was translated into legal form.

Those criticisms reveal that the 1960 settlement redistributed industrial power through a process which the State ultimately controlled, although that account should be set against the way in which the IAC actually worked. The President began a hearing by asking whether negotiation had been exhausted, and the IAC could adjourn at any stage if settlement remained possible. Its members questioned witnesses and assisted representatives, while the Registrar and his staff undertook conciliation of their own, with the result that a substantial number of disputes which reached the IAC ended in agreement. As I see it, compulsory arbitration exerted its influence in two connected ways: the IAC could finally determine a dispute, and the availability of that determination altered the circumstances in which the parties negotiated and conciliated before an award became necessary (Kleinsorge in LIRS (1970) at pp 60–66; Lam Pin Foo, “A Commentary on Some Aspects of the Work of the Industrial Arbitration Court” in LIRS (1970) at pp 94–110).

Before turning to the practical operation of the scheme, I should distinguish one further part of it. Section 77 proscribed dismissals for specified union or industrial reasons, made the conduct an offence, and supplied post-conviction remedies: the District Court could order payment of lost wages and reinstatement. Section 79 dealt with proof of the employer’s reason and required the Attorney-General’s consent for a prosecution; once the facts necessary to establish the offence apart from that reason were proved, the employer had to prove the reason. Brown observed that no prosecution was then known to have been brought, although several requests for consent had apparently been made (D J M Brown, “Initiation of Collective Bargaining under the Singapore Industrial Relations Ordinance” in LIRS (1970) at p 32). I draw attention to this narrower route because it addressed specifically prohibited reasons through a criminal process and co-existed with the IAC’s general jurisdiction to examine whether dismissal was justified upon the industrial merits. When s 34(2) of the 1960 Ordinance was enacted in 1968 (which I will turn to in a moment), it applied expressly outside s 79. Accordingly, when I consider the likely scope of s 34(2) below, care must be taken against collapsing its broader words into the narrower prohibited-motive inquiry in s 79—much as the Guidelines do in relation to s 14(2)—because the Ordinance had already dealt separately and more specifically with such prohibited reasons.

Dismissal within the industrial climate of 1960–1967

The enactment of the 1960 Ordinance did not immediately bring industrial peace. There were 45 work stoppages in 1960, 116 in 1961, 88 in 1962, and 47 in 1963. The figures for 1961 and 1962 must be understood against the division in the labour movement which accompanied the political split within the People’s Action Party (“PAP”). The Singapore Trades Union Congress divided into the Singapore Association of Trade Unions (“SATU”), associated with the Barisan Sosialis, and the NTUC, associated with the PAP Government. Rival unions competed for recognition and members, industrial action served political as well as workplace ends, and neither the IAC nor the legal regulation of bargaining could remove that wider contest. Operation Coldstore, the deregistration of unions, and the eventual elimination of SATU as an effective federation formed part of the explanation for what followed (see Edwin Lee, Singapore: The Unexpected Nation (ISEAS Publishing, 2008) (“Edwin Lee”) at pp 159–161, 185, 220–223 and 243–244). So too did the development of the NTUC’s Labour Research Unit, the professionalisation of bargaining and case preparation (also see Lee Kuan Yew, From Third World to First: The Singapore Story, 1965–2000 (HarperCollins Publishers, 2000) (“From Third World to First”) at pp 92–93), the maturing use of conciliation, and the growing acceptance of the IAC. The decline in stoppages after 1963 cannot, therefore, be attributed to any single institution or event (see C M Turnbull, A History of Singapore, 1819–1988 (Oxford University Press, 2nd Ed, 1989) (“Turnbull”) at pp 276–278; Gareth Curless, “The Triumph of the State: Singapore’s Dockworkers and the Limits of Global History, c 1920–1965” (2017) 60(4) The Historical Journal 1097 at 1119–1123; Kleinsorge in LIRS (1970) at pp 69–72).

Nevertheless, the overall change was pronounced, with recorded stoppages falling from 39 in 1964 to 30 in 1965, 14 in 1966, and ten in 1967. The IAC remained active throughout that decline: in those four years, 476 disputes were referred to it, 287 awards delivered, and 787 collective agreements registered and certified. As I have said, one cannot safely determine how much of the reduction followed from the IAC’s work. The figures show, at most, that reduced direct industrial action coincided with substantial use of bargaining, conciliation, and adjudication. Indeed, Johnson and Quek’s near-contemporary assessment was appropriately cautious: the IAC had promoted collective bargaining and helped to reduce stoppages through conciliation, although the available figures could not establish how many strikes it had prevented (Johnson and Quek in LIRS (1970) at p 115, Table III, and pp 118–119).

Within that changing industrial setting, dismissal occupied an especially difficult position. For an employee, it ordinarily meant the loss of a livelihood and, often enough, the income upon which a household depended. The employees in the awards considered below (see [287]–[304] below) commonly earned little, typically between $120 and $160 each month. Even in 1973—when conditions had no doubt improved from the 1960s—a Member of Parliament described dismissal as “analogous to a death sentence” (Singapore Parliamentary Debates, Official Report (25 July 1973) vol 32 at col 1163 (Mr N Govindasamy) (“1973 Debates”)). The expression was, of course, rhetorical, but it helps one now to appreciate the economic gravity which dismissal carried at the time and which the IAC itself recognised (see, eg, Singapore Manual & Mercantile Workers’ Union v Singapore Steam Laundry Ltd IAC Case No 327 of 1962 (28 September 1963) (“Singapore Steam Laundry”) at [68]). The apparent simplicity with which an employment contract could be ended therefore gives a misleading account of what dismissal meant to the person who lost the work, especially if one stands in the position of a worker in the 1960s.

For the union, the same dismissal could carry a wider significance. It might punish union activity, remove an effective branch officer, intimidate other employees, defeat a recognition claim, or demonstrate that the employer could disregard a negotiated disciplinary process. Even where no anti-union motive was alleged, a union unable to protect a member from what the workforce considered arbitrary treatment risked losing its credibility and the cooperation upon which stable bargaining depended. The employer, for its part, could regard the ability to discipline, select, reorganise, and end employment as indispensable to the proper management of the enterprise. Thus, although a dispute might concern only one employee, the disagreement beneath it could concern the distribution of authority across the workplace. That helps to explain why dismissal could so readily become a “trade dispute”, and why reinstatement could be pursued as an industrial remedy.

The history before 1960 supplies vivid illustrations of that capacity. In the Hock Lee Bus dispute, the employer’s dismissal of union branch officials and later dismissal of 229 members were followed by picketing and sympathetic action across other bus companies, while attempts at settlement included reinstatement and a Court of Inquiry before the confrontation ended in riot and, eventually, arbitration. I do not mean to make the flattened suggestion that dismissal alone caused the violence, since recognition, working conditions, political mobilisation, policing, and public order were all involved. The narrower point is sufficient: a controversy which began in part with dismissal and reinstatement did not remain confined to the contracts of those dismissed. Similarly, in September 1961, the dismissal of a Robinsons sales assistant prompted picketing by other employees and police intervention before management reversed its decision the following day (Turnbull at 255–256; Michael Fernandez & Loh Kah Seng, “The Left-Wing Trade Unions in Singapore, 1945–1970” in Paths Not Taken: Political Pluralism in Post-War Singapore (Michael D Barr & Carl A Trocki eds) (NUS Press, 2008) at p 217).

The Government’s own account of the 1968 amendments confirms that dismissal remained industrially combustible. Mr S Rajaratnam told Parliament that 106 of the 389 work stoppages—or about 27%—arose from dismissals, while a further 120 arose from assorted matters which he treated collectively as showing that the boundary between management and trade-union functions had not been clearly demarcated. That figure should be taken for what it was: a Government classification advanced in support of the amendments, rather than an independently reconstructed causal series. Even so, the total of 389 corresponds with the official annual figures, and the prominence assigned to dismissal was not seriously doubted. The labour Members disputed the inference to be drawn, emphasising that the turbulent years between 1961 and 1963 dominated the totals, that stoppages had since fallen sharply, and that dismissal and retrenchment disputes were frequently resolved by direct negotiation, the Labour Department, or the IAC without a stoppage (Singapore Parliamentary Debates, Official Report (31 July 1968) vol 27 at cols 733–738 (Mr S Rajaratnam) (“31 July 1968 Debates”); 1 August 1968 Debates at cols 769–780 (Mr Seah Mui Kok and Inche Abdul Aziz Karim)).

The work of the Labour Department also shows how much of the dismissal problem remained below the level of a strike or published award. Citing official material published by the Department, Inche Abdul Aziz said that 702 disputes had been referred to it for conciliation in 1966 and that 34.9% were dismissal cases. He added that the IAC had heard 13 dismissal cases in 1966 and four in 1967. Those figures formed part of his argument that the existing institutions had the capacity to cope with the problem (1 August 1968 Debates at cols 773–775 (Inche Abdul Aziz Karim); also see the figures in OHL (2018) at pp 367–369).

By the eve of 1968, the position was therefore more complicated than either the continuing number of dismissal disputes or the falling number of stoppages could reveal on its own. The legal machinery, together with the political and economic changes which surrounded it, had produced a considerably more orderly industrial climate than had existed in the 1950s and early 1960s. Dismissal nevertheless remained a point at which the employee’s need for security, the union’s representative authority, and the employer’s claim to manage could collide. Indeed, the increasing success with which the 1960 settlement moved those collisions from direct action into negotiation, conciliation, and adjudication helps to explain why the Government and the labour Members could read the same experience so differently. To the Government, the persistence of contests over dismissal demonstrated the uncertainty which managerial decision-making continued to face; to the labour Members, the decline in stoppages and the growing capacity to resolve those disputes through established machinery showed that the settlement had begun to work.

How dismissal disputes were resolved

Before examining the IAC’s awards, it is helpful to be precise about what the available routes of redress did and did not provide. At common law, an employer possessing a contractual right to terminate on notice did not have to justify that decision by reference to some general standard of fairness. If the employer dismissed in breach of contract, the action lay for the contractual loss caused by the failure to give the notice or observe the term which the contract required. Addis v Gramophone Co Ltd [1909] AC 488 (“Addis”) is the familiar authority: the employee could recover the salary and commission lost during the notice period, but no damages for the harsh manner of dismissal, injured feelings, or the difficulty which dismissal caused in obtaining later employment. The position could differ where the contract itself confined dismissal to specified grounds, as McClelland v Northern Ireland General Health Services Board [1957] 1 WLR 594 (“McClelland”) illustrates, although such contracts would probably have been rare in 1960s Singapore. In any event, the cause of action enforced the bargain which the parties had made; it did not confer upon every employee a general right to remain employed unless the employer proved just cause.

The remedy was correspondingly narrow because courts of law did not ordinarily compel the continuation of a personal employment relationship, and equity did not ordinarily decree specific performance of a contract of service. A wrongfully dismissed employee therefore received damages, usually measured by the remuneration which would have accrued during the contractual notice period or the unexpired fixed term, subject to mitigation (see Addis at 491–492 and 504–505; Brace v Calder [1895] 2 QB 253; Automatic Fire Sprinklers Pty Ltd v Watson (1946) 72 CLR 435; cf McClelland). There was, of course, Vine v National Dock Labour Board [1957] AC 488. Its exceptional statutory setting—where an invalid removal from a statutory register left the underlying relationship intact—explains the unusual result, for without such a structure the court would neither have examined industrial fairness nor restored the employee to work.

Before 1960, any wider protection arose, if at all, through collective industrial relations. A union could take up the dismissal, negotiate for withdrawal or compensation, seek the Labour Department’s assistance in conciliation, invoke any grievance machinery agreed with the employer, persuade the employer to submit the dispute to the standing Industrial Court, an arbitrator, or a board under the Industrial Courts Ordinance, or seek an investigation and report from a Court of Inquiry. None of those institutions could, however, impose an arbitral settlement without the parties’ consent. If the employer refused negotiation or arbitration and conciliation failed, the union’s remaining leverage lay in direct industrial action, subject to the Trade Disputes Ordinance and the restrictions applicable to essential services. That weakness made an individual dismissal capable of producing a collective stoppage, since the merits of the dismissal and the industrial peace could ultimately depend upon which party was better able to withstand the other’s economic pressure (TPT (1968) at pp 220–223; OHL (2018) at pp 13–16 and 367–369).

The 1960 Ordinance retained the earlier stages of negotiation and conciliation, although it placed them within a process capable of reaching a binding conclusion. A union could include dismissal or reinstatement among its proposals; the parties could settle and have the agreement certified as an award; and a dispute arising under an existing agreement could be referred to its referee. Where bargaining failed to produce agreement, the Commissioner and the compulsory Ministerial conference offered successive opportunities for conciliation, after which the matter could reach the IAC by joint request, Ministerial direction, or, in special circumstances, presidential proclamation. Industrial action remained available at the earlier stages, subject again to the applicable law. Once the IAC acquired cognisance, however, the dispute could no longer be pursued by strike or lockout and would ultimately be disposed of through an award. The availability of that endpoint necessarily altered the setting in which each earlier attempt at settlement occurred (Kleinsorge in LIRS (1970) at pp 58–61).

An individual dismissal could move through that machinery because the statutory definition of “trade dispute” extended to disputes between employers and employees connected with employment or non-employment, including a threatened, impending, or probable dispute. The breadth of that definition was directly challenged in Singapore Steam Laundry. The employer argued that the employee’s dismissal arose from his individual contract of service and, since that relationship had ended, could neither constitute a trade dispute nor be recreated by the IAC. Having obtained the opinion of the State Advocate-General, the IAC rejected both propositions. The Ministerial direction brought the dispute within its cognisance, and the IAC’s broad arbitral power allowed it, according to the industrial merits, either to uphold the dismissal or restore the relationship, with reinstatement serving in that way as one means by which the trade dispute itself could be settled.

The character of the ensuing inquiry followed from that jurisdiction. The IAC had to examine every matter affecting the dispute and its just settlement; it could inform itself beyond the strict rules of evidence, compel material, question the parties, and grant relief which they had not formulated. Its tripartite composition and the ordinary absence of lawyers further distinguished that inquiry from a common law action. Although notice could matter for the reasons considered below, the IAC could examine the employer’s asserted reason, whether the underlying facts had occurred, whether that was the true reason, whether the employee had been heard, and whether dismissal was an appropriate industrial response.

I should make clear that the awards do not show the IAC approaching every case in precisely the same way. In Singapore Steam Laundry, for example, it appeared to require the employer to justify the dismissal; in a case where a union alleged victimisation, the union might instead be required to prove that positive allegation. Its remedial approach was likewise not uniform, and reinstatement could be refused even where the employer’s conduct attracted criticism. Nevertheless, the current which appears to run through the dismissal awards was an inquiry into whether dismissal was justified. Coupled with the IAC’s broad procedural and evidential powers (see [261] above), that focus meant that contractual formality seldom detained it, while the substance of the employer’s justification remained the central concern.

The dismissal awards

I should preface this discussion by noting that the IAC issued 15 dismissal-related awards between 1960 and 1968, before the amendment which removed dismissal disputes from its jurisdiction (see [288]–[304] below), of which fourteen were determined between March 1963 and April 1967. The final award, concerning Singapore Manual & Mercantile Workers’ Union v Singapore Glass Manufacturers Co Ltd (IAC Case No 41 of 1966) (“Singapore Glass”), was handed down on 14 August 1968, after the Industrial Relations (Amendment) Bill 1968 had been debated and enacted. Given their small number and the relatively short period during which the IAC exercised this jurisdiction, the cases can, in my view, disclose no more than an emerging conception of industrial justice in dismissal. Some concerned substantive cause; others, the legal effectiveness of termination or retrenchment; and one records only a mediated settlement. Across those differences, however, dismissal appears as an act whose cause, proof, severity, process, and consequences could be examined independently of the employer’s contractual power to give notice.

The first award, Singapore Textiles and General Merchants Employees Union and Beng Hin Co Ltd IAC Case No 289 of 1962 (20 March 1963) (“Beng Hin”), concerned Quek Gim Par, an odd-job worker who earned $120 each month and was verbally dismissed after almost three years’ service. The employer initially gave neither him nor the union any reason, although it later alleged generally unsatisfactory service and oral warnings. The IAC made no clear finding that the late explanation had been proved and concluded, from the parties’ conduct and their deteriorated relationship, that reinstatement would serve no useful interest for either side. Even so, it described the method of dismissal as “undesirable and unsatisfactory” and ordered $380, comprising $60 in earned wages, $120 in lieu of notice, and $200 ex gratia. The award thus occupied the space between reinstatement and contractual notice: it declined to restore what the IAC regarded as an unworkable relationship, while treating the manner and industrial consequences of dismissal as calling for further relief, although the basis upon which that further sum was quantified was left unclear.

Singapore Steam Laundry, decided six months later, placed the substantive question in much sharper relief. The employee, Kemat, had worked for more than three years without complaint. After he quarrelled with employees of a customer who belonged to a rival union, the employer transferred him to another van without loss of status or wages. Acting upon his union’s advice, he refused the transfer and was dismissed with one month’s salary in lieu. The IAC found that the instruction was lawful, that the union had wrongly advised disobedience, and that Kemat had committed an offence against discipline.

The IAC nevertheless examined the circumstances which had produced the instruction, Kemat’s lack of fault in the original quarrel, his unblemished record, the inadequate explanation given to him, and the haste with which dismissal followed, before concluding that dismissal was disproportionate. In words which disclose the conception of industrial justice then emerging, the IAC said that dismissal was “final and conclusive”, could have “the most serious economic repercussions” for an employee, and “should therefore be used sparingly and for well-grounded reasons”. Kemat was reinstated from the date of the award; his past service and leave were preserved; and he remained entitled to the appropriate notice pay and bonuses, although the period out of work was treated as leave without salary. Thus, neither a proved act of insubordination nor compliance with notice prevented the IAC from asking whether dismissal was the justified sanction.

Goodwood Park Hotel Ltd and Singapore Catering Services Staff and Workers’ Trade Union IAC Case No 132 of 1963 (8 June 1964) (“Goodwood Park”) added a procedural dimension to the proportionality seen in Kemat. Tan Eng Hoon, a porter, had quarrelled with the Chef, seized an iron bar, and, as the IAC found, intended to threaten or cause bodily harm, although the more serious allegation that he threatened to kill the Chef was rejected. Tan also had a recent disciplinary history, and the IAC regarded his temper as unsuitable for hotel work, leading it to refuse reinstatement. The manner in which the decision had been reached nevertheless remained material: the Manager had not interviewed the Chef, had stopped Tan’s attempt to explain, and had summarily ordered him out. The IAC accordingly held that Tan had been dismissed in an unsatisfactory manner and awarded his accrued wages, one month’s wages in lieu of notice, and compensation for the balance of December 1963. The cause for dismissal, the process followed, Tan’s suitability for future work, and the relief appropriate to the case were therefore treated as related, although distinct, questions.

Two retrenchment matters decided in June 1964 illustrate the corresponding restraint which the IAC exercised where the operational reason was genuine. In Singapore Bank Employees’ Union and The Hongkong and Shanghai Banking Corp IAC Case No 48 of 1964 (25 June 1964) (“Bank property dispute”), three employees in the Bank’s Property Department were given three months’ notice following a reorganisation. The IAC examined the Department’s workload, the work expected in future, the employees’ comparative experience, and the selection of the senior draughtsman who was retained, before finding that the redundancy was genuine and upholding the terminations. Because the union had received only one day’s oral notice and had thereby been deprived of a proper opportunity to investigate or propose alternatives, however, the employees were allowed to retain two months’ salary already paid, received additional ex gratia sums ranging from $1,140 to $1,640, and were to be preferred for suitable future vacancies. Three months’ contractual notice did not, therefore, foreclose scrutiny of the operational reason or the selection process. Once those matters were justified, the defect in consultation was answered through a different remedy.

A different anterior question arose in Singapore Manual and Mercantile Workers’ Union and Grant Advertising, Inc IAC Case No 14 of 1964 (3 October 1964) (“Grant Advertising”), where the employer sent registered letters terminating four clerks without notice, although it withheld their final wages and notice pay while awaiting income-tax clearance. A majority held that the mandatory payment requirements in the Clerks’ Ordinance had not been observed, with the result that the terminations were ineffective. The four clerks were deemed continuously employed and received salary from 1 November 1963 to the award on 3 October 1964, less the sums eventually paid; a fifth employee outside the Ordinance received damages only until he found other work. Since the record disclosed no reason for the terminations, the IAC did not have to decide whether the employer possessed substantive cause, and the case turned upon the legal effectiveness of termination. Even so, the consequence is striking when viewed from the modern premise that notice pay ordinarily exhausts a defective termination: almost a year of salary and continuity followed from the conclusion that the relationship had never legally ended.

The majority in The Hongkong and Shanghai Banking Corp and Singapore Bank Employees’ Union IAC Case No 351 of 1962 (24 May 1965) (“Rodrigues”) was, at the same time, prepared to uphold a dismissal where the cause proved was sufficiently serious. Terence Rodrigues was accused of a course of indiscipline and insubordination which culminated in an assault upon his supervisor outside the Bank. The Court examined each incident, his explanations, whether the earlier conduct had been waived, the connection between the assault and the workplace, and the matters advanced in mitigation. It rejected the union’s case of victimisation and held that the cumulative misconduct and assault supplied “sufficient justification” for dismissal. The employee-panel member dissented, viewing the earlier matters as condoned and reinstatement with a three-year loss of increments as the proportionate response. Although the majority and dissent differed sharply in outcome, both examined the sufficiency and proportionality of dismissal as an industrial sanction; the employer’s contractual capacity to terminate did not resolve their disagreement.

The first award in 1966, Singapore Manual and Mercantile Workers’ Union and Transport and Storage Ltd (IAC Case No 47 of 1965) (“Transport and Storage”), returned to the anterior problem considered in Grant Advertising. The employer conceded that it had failed to satisfy the payment requirement applicable to P Palsamy as a clerk. Palsamy was accordingly deemed still employed, his service remained continuous, and he received salary from 5 August 1965 to 27 January 1966, less what had eventually been paid and his outside earnings. Since the threshold defect was accepted, no inquiry into substantive sufficiency was required, although the order again attached continuity and intervening wages to a termination which had never taken effect.

The machinery did not, however, require every reinstatement dispute to end in an adjudicated result. In Singapore Electrical and Wireless Employees Union and Meng Hin Electric Co (1966) IR Supp 569, the parties reported at the opening of the hearing that the Assistant Registrar’s mediation had produced a settlement. The union abandoned reinstatement, the employer agreed to pay each of three employees 45 days’ wages, and those terms were embodied in a consent award. As the published file records no reason for dismissal, it offers no assistance on substantive justification. It does, however, show that a dispute which had reached compulsory arbitration could still be mediated within the IAC and converted into an enforceable compromise.

The most developed analysis of proportionality appeared in Singapore Manual and Mercantile Workers’ Union and Scott and English (SEA) Ltd (IAC Case No 114 of 1964) (“Scott & English”). P T Murthee, an accounts clerk, was dismissed for persistent lateness after three warnings, the last issued 17 months earlier. The union alleged that the true reason lay in personal hostility, union activity, and Murthee’s knowledge of corruption. Although the majority held that neither victimisation nor corruption had been proved even on a balance of probabilities, it found personal animosity, selective scrutiny, no consistent rule requiring latecomers to explain, satisfactory work, and occasions on which Murthee worked late. It also confined the employer to the reason given at the time and declined to entertain its later reliance upon general insubordination. The Court accepted that habitual lateness could supply a valid ground, although “each case” had to be judged on its merits and dismissal was too severe on these facts. Murthee was consequently reinstated from the date of dismissal with continuous service and full basic salary for the period specified in the award, less outside earnings. At the same time, the IAC withheld allowances, bonus, increments, and other benefits, and subjected his attendance to a six-month condition. The order thus preserved financial and disciplinary consequences for the misconduct found, even as the dismissal itself was reversed.

Singapore Manual & Mercantile Workers’ Union and Messrs Kripalson (IAC Case No 11 of 1966) (“Kripalson”) returned once more to statutory effectiveness. Rose May Samy was dismissed for alleged unfaithful service and an uncooperative attitude, with the employer offering two months’ salary in lieu which she declined and which was never tendered or paid. Applying the payment requirement in the Shop Assistants’ Ordinance and its earlier decisions, the IAC held that an offer did not amount to payment and that the termination had not taken effect. It therefore ordered Rose’s immediate reinstatement without loss of service and her full salary from 1 December 1965 until the award on 4 April 1966. Once the termination failed at that threshold, the accusations did not require determination; reinstatement, continuity, and all salary accruing during the invalid termination followed as the industrial consequence.

Salvage Engineers Ltd and Singapore Manual and Mercantile Workers’ Union IAC Case No 48 of 1965 (6 May 1966) (“Salvage Engineers”) brought the technical and substantive inquiries together within the same dispute. Of 45 dismissed employees, the parties reduced the contested list to 19. Fifteen had not been terminated in accordance with the applicable statutory payment requirements and were deemed continuously employed; having regard to mitigation and the company’s precarious finances, they received one or two months’ salary and some were permitted to return. The remaining four had been dismissed for removing and attempting to sell cargo recovered from a salvaged vessel. In determining that allegation, the IAC checked ownership and disposal with the underwriters, considered the inconsistent accounts, and drew an adverse inference from the employees’ decision not to testify. It found sufficiently serious misconduct and upheld those four dismissals despite the manner in which their services had been terminated. The statutory defect thus preserved continuity for 15 employees, whereas proved serious misconduct justified dismissal for the other four.

Singapore Manual and Mercantile Workers’ Union and Lion Tourist Service Ltd IAC Case No 15 of 1966 (8 November 1966) (“Lion Tourist”) made the relationship between notice and cause still more explicit. Four of the 14 employees raised preliminary questions about the notice received, and the IAC dealt separately with acceptance of notice pay, alternative work, and statutory non-compliance before making tailored orders. The other ten were said to have been dismissed for misconduct. Although two had received the contractually adequate 14 days’ notice, the interim award required the employer to prove the gross misconduct invoked, which it failed to do as regards all ten. The employer ultimately conceded that misconduct had been “only an excuse” for some and advanced redundancy, poor work, and financial difficulty as the true grounds. The Court rejected that shifting account after observing that the vacancies had been filled and inferring that higher-paid employees had been replaced with cheaper recruits. Four employees were reinstated without a break in service and received between one and one-and-a-quarter months’ salary; the others received between one and two-and-a-half months’ compensation, reflecting their wishes, alternative work, and the practicality of reinstatement. Notice had therefore ended two contracts in the stipulated manner, although the cause which the employer chose to invoke remained to be proved and justified.

The award in the final case, Singapore Glass, was delivered on 14 August 1968 after the Bill amending the 1960 Ordinance. The union said that it did not fully know why Kamarudin bin Hassan had been dismissed and asked the employer to begin, which the IAC allowed. After extensive evidence, it found that Kamarudin had been absent without authority on three days, although the records did not conclusively establish the fuller accusation that he had fraudulently made attendance entries to obtain wages. The IAC criticised the employer’s failure to catch him making the entries despite surveillance and the union’s failure squarely to meet the factual case. It refused reinstatement because a return would impede harmonious relations and, acting “in equity”, awarded $3,000 in addition to $524.64 in salary, leave pay, and bonus. The IAC therefore awarded compensation after establishing misconduct but finding restoration of the relationship inappropriate.

Read together, the awards dispel any suggestion that the IAC invariably preferred the employee. Genuine retrenchment was upheld in the Bank property dispute, North British, and, by majority, Kie Hock; serious misconduct defeated relief in Rodrigues and for the four employees in Salvage Engineers; and Beng Hin, Goodwood Park, and Singapore Glass produced no reinstatement. The strength of the jurisdiction lay in making the employer’s justification, and the industrial consequences which should follow from it, independently examinable, while leaving the IAC free to uphold the dismissal where the reason and response were sufficiently established.

That point also explains the place of notice in the awards, subject to an important anterior qualification. Grant Advertising, Transport and Storage, Kripalson, part of Salvage Engineers, parts of Lion Tourist, and Singapore Glass concerned mandatory statutory requirements governing notice or prompt payment of final wages. Non-compliance could mean that no termination had legally occurred and notice could, in that setting, be decisive, although the preliminary objection in Singapore Glass failed because the company had done what it could to make payment and the employee had not collected it. Its significance differed once substantive justification arose: Kemat had received a month’s salary; the Bank’s redundant employees, three months’ notice; Kie Hock relied upon compliance with the Labour Ordinance; and two employees in Lion Tourist had received the agreed 14 days. The IAC nevertheless examined the asserted cause in each case. Notice and accrued entitlements could be brought into account when relief was framed, but did not answer whether the dismissal was justified.

The process by which that examination occurred is equally significant, although the awards contain no uniform statement of the formal burden of proof. Lion Tourist expressly required the employer to prove gross misconduct; Scott & English required the union to prove its positive allegation of victimisation; and the majority in Kie Hock considered that the union had not proved the absence of redundancy. Salvage Engineers, for its part, drew an inference from the employees’ decision not to give evidence. Those differences arose, however, within an institution which could bring the employer’s reason and its supporting material into the open. Singapore Glass required the employer to begin because the union did not know the reason; Kie Hock called for certified accounts; Salvage Engineers checked the employer’s explanation with a third party; and the IAC in Singapore Steam Laundry obtained the State Advocate-General’s opinion before deciding the jurisdictional objections. As I see it, the historically important point is that, whatever the formal allocation of proof in a particular dispute, the process enabled the asserted reason to be identified, investigated, and tested. It did not ordinarily leave an employee who had received no reason attempting to prove a prohibited motive while the employer remained silent.

The remedial position completes the picture and would, I think, appear almost unrecognisable to an employment lawyer today. Depending upon what industrial justice required, the IAC could restore the relationship prospectively or from the date of dismissal, preserve continuity and accrued benefits, and award all or part of the wages lost during the separation, subject to mitigation. It could also impose conditions upon reinstatement, substitute compensation where a return was impracticable, require retrenchment benefits or preference for future vacancies, and give binding force to a mediated compromise. Thus, Grant Advertising treated four employees as continuously employed for almost a year; Scott & English restored the relationship from the date of dismissal and awarded full basic salary for a substantial part of the intervening period; and Kripalson and Transport and Storage ordered immediate continuity together with the salary which had accrued. Viewed from the common law premise that wrongful dismissal ordinarily produces notice damages and no reinstatement, those orders were extraordinary. Within the industrial jurisdiction of the 1960s, however, they were applications of the remedial authority which the Ordinance had conferred for settlement of the dispute.

That reading accords with the understanding of contemporary legal observers. Chia Quee Khee wrote that, although the Ordinance had stated no general standard of dismissal, the IAC and referees had developed doctrines and standards by which disciplinary action could be measured. He regarded the union’s capacity to check the employer’s wide disciplinary discretion as one of the labour movement’s strongest bases of success. His examples extended beyond the awards gathered here: a referee in the Borneo Motors case upheld dismissal only after repeated negligence and ignored warnings, but still recommended compensation for long service; Ford Motor recognised the employer’s inherent right to dismiss while treating a show-cause process as part of its exercise; and Goodwood Park treated procedural unfairness as rendering dismissal unjustifiable for remedial purposes. Chia concluded that the IAC sought compromise solutions rarely seen in ordinary courts because its function was to resolve the dispute, balance the parties’ interests, heal the relationship where possible, and encourage better industrial relations (Chia in LIRS (1970) at 82–85). That near-contemporary description is especially helpful because it enables us to see more accurately how the IAC was perceived at the time.

The settlement recast in 1968

The legislative programme of 1968 accordingly altered the individual terms of employment as well as the collective distribution of industrial authority. The Employment Act 1968 (Act 17 of 1968) consolidated the Labour, Clerks’ Employment, and Shop Assistants’ Employment Ordinances, standardised minimum conditions, and reduced or controlled a number of benefits which the Government considered liable to impede industrialisation. The Industrial Relations (Amendment) Act 1968 (Act 22 of 1968), passed on 1 August and brought into operation on 15 August 1968, lengthened the duration of agreements, restricted more favourable terms in new undertakings, and, most importantly here, withdrew specified matters from the statutory field of collective bargaining and arbitration. Writing as the changes were made, Tan described them as retrogressive from labour’s perspective but prompted by grave economic necessity. The 1960 Ordinance, he said, had given workers a share in management matters formerly regarded as the employer’s absolute prerogative, and the 1968 amendments deliberately reversed part of that industrial democracy (see TPT (1968) at pp 228–229; Tan in LIRS (1970) at pp 134–137; Edwin Lee at p 267; From Third World to First at pp 87–88; also see Chandran (Lexis) at paras 4.3–4.9).

The individual contractual side of the arrangement should also be noticed. Sections 10(1) and 11(1) of the new EA stated, respectively, that either party could at any time give notice of an intention to terminate the contract and could bring it to an immediate end by paying the salary which would have accrued during the notice period. Those provisions gave clear statutory expression to the ordinary contractual freedom to terminate by notice or salary in lieu. Enacted at the same time within the industrial-relations legislation, the new s 34(2) supplied a countervailing restraint. Read together, the provisions allowed termination of the contract in the prescribed manner while preserving the distinct inquiry into whether the employer’s decision to dismiss could be justified. Had payment of notice answered both questions, the protective provision would have had very little work to perform. OHL (2018) has since expressed the same understanding by treating the protection as extending to termination with notice or salary in lieu (OHL (2018) at p 143). That later commentary cannot establish Parliament’s meaning in 1968, although it accords with the distinction disclosed by the contemporaneous statutory structure.

The Government expressed the change somewhat differently. Mr Rajaratnam accepted that the 1960 Ordinance had been enacted to secure “industrial peace with justice” and had worked satisfactorily in many respects, while maintaining that it had left “fundamental areas of latent dispute” over decisions which properly belonged to management. Promotion, transfer, recruitment, reorganisation, retrenchment, assignment of work, and dismissal had become subjects of negotiation and industrial action. So long as management could not know whether those decisions would be reopened through collective pressure or the IAC, investment and efficient organisation remained exposed to uncertainty. The Bill therefore sought to restore those functions to management, subject to identified restraints against abuse (31 July 1968 Debates at cols 733–740 (Mr S Rajaratnam)).

That policy found expression in provisions which must be read together. The new s 17(2) prohibited a trade union from including, in a notice of proposals for a collective agreement, any proposal concerning promotion, a non-detrimental transfer, appointment to a vacancy, termination because of redundancy or reorganisation (including selection criteria), dismissal and reinstatement in circumstances to which the new s 34(2) applied, or assignment of compatible duties. The union could no longer invoke the statutory collective-bargaining process over an ordinary dismissal or reinstatement. A proviso added to s 34(1) then prohibited the IAC from considering a dismissal dispute or ordering reinstatement, save where dismissal arose from a contravention of s 79—the existing protection against specified anti-union and industrial victimisation. Clause 10 completed the symmetry by deleting from s 40(a) the express words concerning dismissal and reinstatement which had been inserted in 1965. The old route by which dismissal could be negotiated, conciliated as a trade dispute, and ultimately determined by the IAC was thus closed at every relevant stage of the statutory sequence (Industrial Relations (Amendment) Bill 1968, cls 5, 8 and 10, and Explanatory Statement).

The labour Members directed their principal objection towards that redistribution of authority. Mr Seah Mui Kok accepted the need for industrialisation, although he feared that unscrupulous employers would treat the amendments as a licence to exploit workers. He emphasised the marked decline in stoppages and asked the Government to respond firmly to abuse. Abdul Aziz traced the 1965 amendment which had placed the IAC’s reinstatement power beyond dispute, explained that it followed an NTUC resolution, and argued that workers had acquired a legitimate share in the procedures through which management functions affected their lives. He was, however, neutral as to the identity of the decision-maker: what mattered was that general termination cases remained capable of effective resolution, whether through the IAC or the Ministry. Other labour Members likewise pressed for safeguards, fair inquiry, and continued institutional protection, while accepting that some reallocation might be required (1 August 1968 Debates at cols 765–780).

There was also a genuine disagreement over the evidence upon which the Government relied. Mr Rajaratnam treated the 106 dismissal-related stoppages, together with the 120 stoppages concerning other asserted management functions, as demonstrating that the boundary remained unstable. The labour Members answered that most occurred during the exceptional political conflict of the early 1960s, whereas by 1966 and 1967 employers, unions, the Labour Department, and the IAC had learned to resolve dismissal and retrenchment disputes with decreasing resort to industrial action. Whether the 1968 restriction completed the earlier settlement or abandoned a system which had begun to mature is a historical and policy question upon which reasonable views may differ. For present purposes, the shared premise matters more: dismissal was already subject to a substantive industrial protection, and the principal question was which institution should administer it after collective contest and IAC adjudication were removed.

The protection enacted as section 34(2)

The protective part of the 1968 arrangement appeared in the same clause which withdrew the IAC’s jurisdiction. The new s 34(2) permitted an employee who considered that she had been dismissed without just cause or excuse, otherwise than in contravention of s 79, to make written representations through her trade union to the Minister within one month, seeking reinstatement in her former employment. Under s 34(3), the Minister could request the Commissioner to inquire into the dismissal and report whether, in his opinion, it was without just cause or excuse. If the Minister dealt with the representations himself, s 34(4) required him to give the employer an opportunity to state in writing why the dismissal was considered justified.

The remedial provisions are equally important. If the Minister was satisfied, after considering the representations of the union and employer and any report by the Commissioner, that the employee had been dismissed without just cause or excuse, s 34(5) empowered him—“notwithstanding any rule of law or agreement to the contrary”—to direct reinstatement in the former employment together with the wages which the employee would have earned, or to order such amount of wages as he determined by way of compensation. The employer was bound to comply; the decision was final and conclusive and could not be challenged in any court or in the IAC (s 34(6)). A direction barred an action for damages in respect of the wrongful dismissal (s 34(7)), while non-compliance was punishable by imprisonment for up to 12 months, a fine of up to $5,000, or both (s 34(8)). After conviction, any unpaid amount could be recovered as though it were a fine and paid to the employee (s 34(9)).

The changes in process and institution should not be understated. The employee had to act through her trade union; the matter no longer took the form of a trade dispute adjudicated publicly by a tripartite industrial court; and the Minister’s determination became an executive decision insulated from challenge. The statute prescribed no procedure equivalent to the IAC’s open hearing, broad evidential powers, or duty to act according to equity, good conscience, and substantial merits. Those changes explain why several Members would have preferred the IAC to remain seised, and they eventually contributed to the fading of the earlier industrial conception.

The question, then, is what became of the substantive protection when its institutional setting changed so markedly. In my view, the better reading is that s 34(2) transferred and reconfigured the merits-based protection which the IAC had administered, leaving the decision with the Minister and removing the collective routes through which dismissal had previously been contested. Four connected features of the settlement support that view: the place at which s 34(2) was inserted; the language of justification which runs through the provision; the remedies and coercive enforcement which it preserved; and the understanding disclosed by the Parliamentary debates. There is also a practical feature of some importance, namely that the employer’s reason remained part of the statutory inquiry. I address those matters in the order just stated.

The natural starting point lies in the structure, for ss 34(2)–(9) were enacted at the very point when Parliament removed dismissal from collective bargaining and the ordinary jurisdiction of the IAC. The common law and the statutory provisions governing notice would have remained available without any further enactment. Parliament nevertheless created a public route through which an employee could contend that her dismissal lacked just cause or excuse, seek reinstatement, and obtain a direction enforceable by criminal sanction. The location and elaboration of that substitute are difficult to reconcile with a protection confined to notice pay or some contractual entitlement already enforceable elsewhere. They point, rather, to preservation within different machinery of the inquiry which withdrawal of the IAC would otherwise have removed.

The language of the provision leads in the same direction, since the employee represented that she had been “dismissed without just cause or excuse”, the Commissioner could inquire into the dismissal and report upon that same question, the employer was given an opportunity to explain why dismissal was considered “justified”, and the Minister had to be satisfied that just cause or excuse was absent. Through those successive stages, the employer’s asserted justification was placed before an institution charged with determining its sufficiency. As I see it, the resemblance to the IAC’s mode of analysis is strong: the reason had to be identified; what occurred and why could be examined; and the decision-maker then determined whether dismissal was an industrially justified response.

The remedies confirm the point because reinstatement with all wages lost appeared first, followed by compensation measured in wages, and either could be directed “notwithstanding any rule of law or agreement to the contrary”. Those words prevented an otherwise applicable contractual power, an agreement which purported to foreclose reinstatement, or the common law refusal to compel a personal employment relationship from determining the statutory result. Section 34(7) then made a Ministerial direction a bar to an ordinary damages action, thereby presupposing that the statutory proceeding and contractual action supplied different forms of redress. In content and effect, the remedial structure of the industrial jurisdiction had been placed in different hands.

The legislative explanation was framed in substantially the same way. Mr Rajaratnam said that the employer’s right to dismiss was “subject to certain restraints” and described the new representations as the means by which an employee dismissed without just cause or excuse could seek reinstatement. Directing those cases to the Minister would provide “more speedy machinery” for individual dismissals and prevent the IAC from being “bogged down by prolonged arguments over isolated cases” (31 July 1968 Debates at cols 738–740). His criticism was directed towards the time consumed by, and the institutional allocation of, the former process. It gives no indication that the substantive question which that process had examined—the justification for dismissal—was intended to disappear with the forum.

Mr Low Guan Onn, the Member for River Valley, makes the connection still clearer. He observed that, before 1966, dismissal of a single employee could produce prolonged negotiation, an unsettled dispute, and industrial action, although he preferred the Ministerial procedure because it promised a quicker determination. Mr Seah, concerned that the protection originally proposed might be confined too narrowly to trade-union victimisation, said that the new s 34 procedure “takes care” of dismissals arising in other circumstances. Abdul Aziz raised other objections to the Bill, but expressly remained neutral as between adjudication of general termination cases by the IAC and their resolution through the Ministry. That neutrality provides affirmative evidence that he understood the protection to continue notwithstanding the institutional change. The labour Members’ wider concerns addressed collective participation, the loss of the specialist tribunal, the risk of employer abuse, and the adequacy of the new process. No Member suggested that “without just cause or excuse” preserved only notice rights or required proof of discrimination or another independently wrongful purpose before the employer’s reason could be examined (1 August 1968 Debates at cols 765–780 and 786–790).

It is also relevant that Parliament legislated in 1968 against a visible and recently amended body of industrial practice. The IAC’s awards were published; the Government had amended the Ordinance through Act 27 of 1965 in response to continuing controversy over reinstatement; and the 1968 debates referred to dismissal cases, the relief granted, and the institutional experience accumulated. I do not suggest that every Member read every award. A narrower and, I think, sensible inference is that the Government, the organised labour Members, and the industrial-relations officials responsible for the Bill knew the general manner in which dismissal had been treated. If s 34(2) had been intended to reduce that merits-based protection to a contractual claim or a narrow list of forbidden motives, one would expect some trace of so significant a change in the Bill or debate, but no such trace appears.

A connected practical point arises from the process itself. The union ordinarily knew the employer’s asserted reason from negotiation and the IAC could require that reason to be stated, supported, and tested, a feature which s 34 preserved in a different form. The Commissioner could inquire into “the dismissal”; where the Minister decided the matter himself, the employer had to be given an opportunity to explain in writing why dismissal was justified. The scheme therefore contemplated an inquiry in which the employer’s justification would be brought into view. It did not leave an employee who received notice without explanation to prove some unknown wrongful motive while the employer remained silent. In that respect, the older process avoided the peculiar problem of negative proof which now arises under the Guidelines.

Seen in its full setting, therefore, s 34(2) transferred and reconfigured the former protection. Management acquired the freedom from collective bargaining and IAC adjudication which the Government regarded as necessary for economic growth, while the individual employee lost both routes. The surviving public protection restrained dismissal which could not be justified, was administered by the Minister, and was supported by reinstatement, back wages, compensation, finality, and coercive enforcement. Although the balance shifted markedly towards management, the protective standard which survived drew its content and much of its intelligibility from the merits-based industrial justice which preceded it.

A short interjection on the formula

Before following the protection into the EA, I should say something about the origin of the expression “without just cause or excuse”. The short answer is that its provenance in dismissal legislation remains unclear. The words were certainly not coined for that purpose, for the language of doing something “without just cause or excuse” had enjoyed a considerably longer legal life. At common law, it appeared most familiarly in the law of tort and, in particular, descriptions of “malice”. Thus, in Bromage v Prosser (1825) 4 B & C 247 at 255, Bayley J described malice in law, in substance, as the intentional doing of a wrongful act without just cause or excuse.

One possible route by which that vocabulary entered collective labour law runs through the economic torts and, more particularly, the procurement of a breach of contract. Industrial disputes in the 19th and early 20th centuries were often analysed by asking whether a union, official, or combination of workers had induced employees to break their contracts of service. Lumley v Gye (1853) 2 E & B 216, the foundational authority, itself concerned the procurement of a breach of a contract for personal services; Bowen v Hall (1881) 6 QBD 333 generalised the principle; and Quinn v Leathem [1901] AC 495 and South Wales Miners’ Federation v Glamorgan Coal Co Ltd [1905] AC 239 placed the same ideas in or near industrial conflict. The latter concerned a federation whose direction of “stop-days” caused miners to abstain from work in breach of their contracts. Across that line of authority, malice, intentional interference, justification, and the absence of just cause or excuse performed much of the conceptual work.

The same vocabulary appears in a cognate legislative field. In the United Kingdom, the Conspiracy and Protection of Property Act 1875 (UK) dealt with certain wilful and malicious breaches of contracts of service or hiring, while the Trade Disputes Act 1906 (UK) later protected acts done in contemplation or furtherance of a trade dispute from liability merely because they induced a breach of an employment contract. Locally, the present s 2 of the Trade Disputes Act 1941—which began as the Trade Disputes Ordinance 1941—defines “maliciously” as the doing of a wrongful act intentionally without just cause or excuse. The Act also occupies the same general field of wilful and malicious breaches of contracts of service or hiring and combinations in contemplation or furtherance of trade disputes. It may safely be said, therefore, that the expression belonged to an older local vocabulary of collective labour law. That makes its later appearance in the dismissal provision historically plausible, although it does not establish the path by which the draftsman selected it.

Other possible influences arise from the use of “just cause”, which had become familiar in collective agreements and industrial arbitration as a restraint upon discipline and discharge. Chia observed that, where a Singapore collective agreement contained a “just cause” term, a dismissal dispute could be dealt with by its referee, while the IAC developed cognate standards through referrals where no such term existed (Chia in LIRS (1970) at 82). Five years before the amendment, the ILO Termination of Employment Recommendation, 1963 (No 119) stated that employment should be terminated only for a valid reason connected with the worker’s capacity or conduct or the operational requirements of the undertaking; the union and NTUC invoked that Recommendation in Rodrigues. It is therefore possible that the formula drew upon one or more of these industrial sources, the older tortious vocabulary, a shared regional legislative practice, or simply the draftsman’s choice of broad ordinary language.

One can readily understand the attraction of the words in this setting. In ordinary private law, dismissal is an exercise of contractual power; in the life of the employee, it is also an act by which one person deprives another of work, income, status, and security. A formula which asks whether the employer possessed “just cause or excuse” is broad enough to express the idea that such power should not be exercised arbitrarily, capriciously, discriminatorily, punitively, or without a sufficiently legitimate reason. Yet, the ease with which a plausible genealogy can be supplied is itself a reason for caution. The Bill’s Explanatory Statement and Parliamentary debates say nothing about why these words were selected, and the Ministerial decisions which might disclose how they were understood soon after enactment were not published. As the Minister for Manpower acknowledged in 2018, MOM’s dismissal cases had “not been published thus far as there was little need to do so” (2018 Debates at 3.56pm (Mrs Josephine Teo, Minister for Manpower)). Nor have any drafting instructions, departmental minutes, internal legal opinions, or other materials which may reveal how the Labour Department arrived at the formula and architecture been made available. Without such evidence, I do not think it is possible to say with confidence which influence, if any, was actually in the draftsman’s mind.

I should, at this point, correct an error in JGP. At [A.74], I suggested that the Singapore provision was “more likely than not” borrowed from what is now s 20 of the M’sian IRA—a suggestion which was wrong because I overlooked that the equivalent Malaysian provision did not form part of the 1967 Act as first enacted. It was introduced in 1971 as s 16A by the Industrial Relations (Amendment) Act 1971 (Act A92), which came into force on 1 October 1971. Section 16A was repealed and replaced in 1975 by s 17A through the Industrial Relations (Amendment) Act 1975 (Act A299), and s 17A was renumbered as s 20 in the 1976 revision (see Lee Wah Bank Ltd v National Union of Bank Employees [1981] 1 MLJ 169). The architecture is strikingly similar to Singapore’s: a dismissed workman could make written representations for reinstatement, dismissal complaints were placed before the Minister, and available relief included reinstatement or compensation. The chronology, however, means that none of those Malaysian statutory provisions could have supplied the source for Singapore’s enactment in 1968.

The Malaysian emergency regulations preceding Act A92 do not save the error. When introducing the 1971 Bill, the Malaysian Minister for Labour and Manpower explained that it consolidated, with modifications, the Essential (Industrial Relations) Regulations 1969 and the Essential (Industrial Relations) Regulations 1971, making their provisions a permanent part of Malaysian labour law. The former were made on 9 October 1969 and the latter on 3 February 1971. Members likewise treated s 16A as carrying forward the post-1969 emergency arrangement under which dismissal cases had been removed from the Industrial Court and placed before the Minister. Those measures also post-dated the Singapore enactment (Malaysia, Dewan Ra’ayat, Official Report (23 July 1971) at cols 3671–3674; ibid (26 July 1971) at cols 3718–3726). Ironically, therefore, I made the very kind of ahistorical error which, in the same breath, I suggested that the Court of Appeal had made in Lim Tow Peng and another v Singapore Bus Services Ltd [1974–1976] SLR(R) 673 (“Lim Tow Peng”) at [16]–[17] (cf JGP at [A.71]–[A.75]). Little can now be done beyond acknowledging and correcting it.

I do not now adopt the converse proposition that Malaysia probably borrowed its provision from Singapore. There is some contemporaneous academic support for influence in that direction: Tan Ng Chee wrote in 1971 that Malaysian s 16A was “modelled on section 34(2-8) of the Singapore Industrial Relations Act” (Tan Ng Chee, “The Freedom to Picket and the Labour Injunction in Malaysia” (1971) 13(2) Malaya Law Review 408 at 411). The statement is significant because it was made in the year of the Malaysian amendment and points from Singapore to Malaysia, although it identifies no drafting record or official source. The Malaysian debates explain the immediate policy and emergency-law setting in which s 16A was enacted, but likewise do not reveal the provenance of its particular wording or structure. Direct borrowing from Singapore is therefore possible and enjoys contemporaneous secondary support. A common antecedent, the circulation of shared industrial-relations ideas within the region, or independent statutory expression also remain possible. In my view, the available materials permit no confident conclusion as to which explanation is correct, and the origin of Singapore’s formula remains unclear.

Nothing in the present analysis ultimately depends upon identifying that etymological source. The words are general enough to derive meaning from the statutory work which they were enacted to perform, and that work is better evidenced by the provision’s place in the 1968 settlement, its procedure and remedies, the jurisdiction which it replaced, and the understanding disclosed by the debates. Although an exact genealogy would certainly be historically interesting, it would not displace the stronger evidence that s 34(2) carried forward, through a new decision-maker, a substantive inquiry into the justification for dismissal.

The protection extended to non-unionised employees in 1973

I should begin by correcting a chronology which is easily missed. Section 14(2), in its present conception, did not form part of the EA when that Act was enacted in 1968. The original s 14 dealt with discipline for misconduct and allowed an employer, after due inquiry, to dismiss without notice for misconduct inconsistent with fulfilment of the express or implied conditions of service, or to impose the lesser sanctions of immediate downgrading or suspension without salary for up to one week; it also regulated suspension during the inquiry. There was no general representation to the Minister on the ground that a dismissal was without just cause or excuse (see s 14 of the Employment Act 1968 (as enacted)).

That omission produced an evident distinction after the 1968 industrial-relations amendment. A unionised employee could make representations through her union under s 34(2) of the Ordinance, whereas an employee who was not unionised could enforce notice pay and other contractual or statutory entitlements but had no equivalent route through which the substantive justification for dismissal could be examined and reinstatement or compensation ordered. Five years later, the Employment (Amendment) Act 1973 repealed and re-enacted s 14 to remove that distinction. The new subsections were enacted through cl 2 of the Employment (Amendment) Bill 1973, which Parliament considered on 25 July 1973, and the Act came into operation on 1 September 1973.

The Minister for Labour, Mr Ong Pang Boon, explained the object with considerable clarity. The 1968 legislation had restored to management the right to hire and fire. Employers had generally exercised that right fairly, although some employees continued to be dismissed without just cause or excuse. A unionised employee possessed “adequate redress” under the Industrial Relations Ordinance because the Minister could, after due inquiry, order reinstatement or compensation; a non-unionised employee had no such redress. The Commissioner for Labour could award salary in lieu of notice, but that was capped at four weeks and was “clearly no deterrent” against an unjust dismissal. In the Minister’s words, “a non-unionised employee is as much entitled to protection against unjust dismissal as his unionised counterpart”. The amendment would therefore permit the former to make representations to the Minister, who could order reinstatement or compensation (1973 Debates at cols 1158–1159 (Mr Ong Pang Boon, Minister for Labour)).

In every respect material to the substantive protection, the structure enacted through ss 14(2)–(7) followed s 34. An employee who considered that she had been dismissed without just cause or excuse could make written representations directly to the Minister within one month and seek reinstatement. The Minister could request the Commissioner to inquire into the dismissal and report whether it lacked just cause or excuse. If satisfied after considering the report, the Minister could—despite any rule of law or agreement to the contrary—order reinstatement with the wages which would have been earned or direct such amount of wages as he determined by way of compensation. The decision was final and conclusive; a direction barred an action for damages in wrongful dismissal; and non-compliance attracted penal consequences (also see Barker at pp 37–38). The chief procedural difference lay in the absence of a union intermediary. The EA version also omitted s 34(4)’s express provision for written representations by the employer and s 34(9)’s recovery mechanism, while retaining the original s 14 provisions on due inquiry and suspension in renumbered form. Although those differences are real, the gateway, inquiry, substantive standard, principal remedies, finality, and public enforcement are recognisably derived from the same scheme.

The Members understood the relationship between the provisions in closely related, although not identical, terms. Mr N Govindasamy said that “a similar procedure” appeared in s 35(2) of the Industrial Relations legislation and identified “[t]he only difference” as the need to proceed there through a union. He accepted the need to protect employees incapable of unionisation, but objected to allowing an employee eligible to join a union to obtain the same route while remaining outside it. He also emphasised the gravity of dismissal, the employee’s need to know and answer the case, and the risk that an unsatisfactory inquiry would generate distrust and industrial unrest. Mr Chin Harn Tong, for his part, said that the Bill brought the EA into line with the Industrial Relations legislation and described the provisions as “substantially the same”. He proposed that s 14(2) expressly say “whether with or without notice”, lest an employer evade the protection by giving two weeks’ notice. Mr Seah Mui Kok placed the amendment within the settlement which labour had accepted in 1968: managerial functions had been restored to attract investment, while experience of employer abuse showed the need for an effective restraint (1973 Debates at cols 1162–1170).

The Minister did not expressly accept Mr Chin’s proposed words, and the enacted provision remained silent on notice. Yet, the reason which he had already given makes the substantive answer difficult to avoid. The existing power to award up to four weeks’ salary in lieu of notice was inadequate precisely because it did not deter unjust dismissal; Parliament was creating a different protection for that reason. To hold that an employer could defeat the new inquiry merely by giving notice or paying the same salary in lieu which the Minister had described as inadequate would invert that object. The repeated words “notwithstanding any rule of law or agreement to the contrary”, together with the remedies of reinstatement and back wages, confirm the conclusion.

The Minister’s reply also explained how the inherited process worked in practice. A representation under s 34 first went to the “Industrial Relations Section”, which attempted conciliation and asked the employer in writing for the circumstances leading to, and grounds for, dismissal. A case which disclosed no prima facie basis could be rejected. Where management appeared partly at fault, the Department consulted the parties separately with a view to reinstatement or monetary settlement. Only if conciliation failed would the matter proceed to a formal inquiry by the Commissioner under s 35(3) or to written employer submissions for decision by the Minister under s 35(4). From August 1968 until the debate in 1973, 109 representations had been received: 24 were rejected, 52 settled by conciliation, 23 withdrawn, three remained pending, and seven resulted in Ministerial orders—four for reinstatement and three for compensation. The Minister said that the same procedure would apply to the new s 14 (1973 Debates at cols 1172–1174 (Mr Ong Pang Boon)).

That account is particularly revealing for the difficulty considered in these grounds. The employee had to present a case worth investigating, after which the employer was asked for the circumstances and grounds of dismissal. The Department could conciliate upon those reasons and, if necessary, the Commissioner inquired and the Minister decided. Thus, even for an employee who was not unionised, the process did not demand proof of discrimination, retaliation, or some other concealed motive where no reason had been given. It elicited the employer’s explanation and tested whether the dismissal was justified. To that important practical extent, the 1973 process preserved the IAC model within the Ministerial machinery created in 1968.

The conclusion which I draw from the 1973 enactment is therefore reasonably secure. Parliament identified the absence, for non-unionised employees, of a protection already enjoyed by unionised employees and declared them equally entitled to it. It reproduced the language, inquiry, principal remedies, finality, and public enforcement of the 1968 provision, subject to the procedural differences just identified. Section 14(2) consequently inherited the same substantive character and asked whether the employer possessed just cause or excuse for dismissal, irrespective of whether contractual notice had been paid, while supporting that inquiry with reinstatement, lost wages, or compensation extending beyond the remedial limits of the common law.

What was inherited

The preceding history may now be drawn together. Before 1960, an employee’s ordinary protection was principally contractual and statutory: notice pay and other entitlements could be recovered, although the common law supplied no general right to require justification or obtain reinstatement. A union could contest the same dismissal through industrial means, but the pre-1960 machinery could not compel arbitration without consent.

The 1960 Ordinance then placed negotiation, conciliation, and possible compulsory arbitration within an enforceable sequence. As the awards at [283]⁠–⁠[304] show, the IAC could require the employer’s reason to be identified and proved, examine its truth and sufficiency, and determine whether dismissal was the appropriate industrial response. Where genuine redundancy or serious misconduct was established, it upheld the dismissal. Contractual notice remained relevant to termination and relief, although it did not decide the merits inquiry.

In 1968, Parliament withdrew dismissal from collective bargaining and the IAC, while creating s 34(2) within the same arrangement. The new process brought the employer’s justification into view and allowed the Minister to order reinstatement, lost wages, or compensation despite the contract and ordinary law. When Parliament extended materially the same protection to non-unionised employees in 1973, it expressly regarded notice pay as an inadequate deterrent to unjust dismissal. Section 14(2) therefore inherited the industrial protection’s general standard of substantive justification.

That inheritance did not freeze every incident of the IAC’s practice. Parliament remained free in 2018 to confine the protection, alter the burden, or replace the general standard with a taxonomy better suited to judicial administration. The difficulty is that the enacted words disclose no such alteration. Parliament retained the statutory formula and remedies while directing the ECT to have regard to Guidelines which gave the protection a materially different spirit.

The debates in 2018 anticipated that later account by identifying discrimination, benefit deprivation, and retaliation as wrongful, while treating poor performance, misconduct, and redundancy as legitimate, with the employee first required to substantiate a wrongful case following a contractual termination. Yet, the debates did not examine the IAC awards, the paired provisions enacted in 1968, the employer-justification procedure, the remedial continuity, or Parliament’s express extension of the protection in 1973. The legislative materials reveal no collective engagement with what the retained words had historically protected, although individual participants may have appreciated parts of that history.

The divergence may therefore be explained if the Guidelines crystallise a later administrative practice centred upon contractual termination, specified wrongful reasons, and an employee’s initial burden. Such a conception sits uneasily with an inherited protection which elicited and tested the employer’s justification and could reverse a dismissal despite notice. The statutory language and remedial shell survived after much of the conception which had made them intelligible faded, and it is to that change that I now turn.

The faded conception of industrial justice

The preceding section established that the inherited protection permitted public examination and, where industrial justice required it, reversal of an employer’s decision even though the dismissal was contractually effective (see [342]⁠–⁠[347] above). I turn now to how that feature progressively disappeared from the public conception of employment security.

Industrial justice continued to be an object of public policy and found expression in fair wages, worker protection, conciliation, responsible retrenchment, humane adjustment, and the distribution of gains from economic growth, while tripartism became more deeply embedded in the governance of work (cf Woon and Loo at pp 47–52). What changed was the object around which those concerns were organised. As Singapore moved from labour surplus to labour scarcity, from labour-intensive industrialisation through repeated restructuring, and then towards a knowledge- and skills-based economy, employment security increasingly came to mean a worker’s capacity to remain economically active across changes of job, firm, occupation, and industry. Policy accordingly attached growing importance to continuity within the labour market as well as continuity within a particular employment relationship. The development may be traced in three stages: first, the connection drawn between industrial peace, wages, productivity, and flexible adjustment from the early 1970s until the recession of 1985; second, the increasing reliance upon retraining, mobility, and employability from the late 1980s, and their eventual organisation as manpower policy; and, finally, the reform debated in 2018 and the manner in which that later setting informed the conception of dismissal which appeared there. The history assists in explaining that conception; the legal meaning of the words which Parliament retained remains a separate inquiry.

From industrial peace to productivity

The social and economic conditions surrounding the inherited settlement changed rapidly even as its two statutory parts were being assembled. Section 34(2) was enacted against the threatened withdrawal of the British military presence and in an economy which had only recently experienced serious unemployment. Export-oriented industrialisation then proceeded with unusual speed: economic growth averaged 12.6% between 1965 and 1973, and by the beginning of the 1970s Singapore had moved from labour surplus to practical full employment and an emerging shortage. Foreign investment produced more jobs than the resident workforce could fill (Edwin Lee at pp 271–274; cf From Third World to First at pp 62–63 and 87–88). Accordingly, when Parliament extended the protection to non-unionised employees in 1973, labour policy had to address the insecurity of a poorly paid employee alongside the supply, price, training, and allocation of labour in an economy whose workforce was beginning to constrain further expansion (see Woon and Loo at pp 59–60; Kwa Chong Guan et al, Seven Hundred Years: A History of Singapore (National Library Board and Marshall Cavendish Editions, 2019) at pp 268–274; Michael D Barr, Singapore: A Modern History (I.B. Tauris, 2019) (“Barr”) at pp 154–161).

The response to the British withdrawal had already supplied an early illustration of how economic adjustment and worker protection could meet. In 1968, the Government organised three-month courses to retrain redundant base workers for employment in industrial vocations, alongside accelerated public works and construction and an expansion of technical education. The results were uneven: the courses were short, some older workers found the transition difficult, and the retraining programme was eventually discontinued. Even so, Mr Rajaratnam described the displaced base workers as early victims of technological change and warned that machines and computers would later displace white-collar workers as well. Skills and movement into other work were therefore already present in the policy response to displacement when the inherited settlement was made. Their later significance lay in the scale, regularity, and institutional form which they acquired (see Loh Kah Seng, “The British Military Withdrawal from Singapore and the Anatomy of a Catalyst” in Singapore in Global History (“Singapore in Global History”) (Derek Heng & Syed Muhd Khairudin Aljunied eds) (Amsterdam University Press, 2011) 195 at 201 and 203–204; also see From Third World to First at pp 53–55).

The response to the oil and trade shocks of 1973–1975 confirms that preservation of employment remained important. Faced with slower growth and the prospect of substantial job losses, the Government sought further foreign investment, including labour-intensive investment which deferred its preferred restructuring, and undertook reclamation and construction activity to sustain employment. Growth recovered and averaged 9% over the decade (see Woon and Loo at 60–61). Industrial disputes also remained part of ordinary administration: between 1975 and 1980, the Ministry conciliated an annual average of 635 disputes (see Woon and Loo at pp 111–112). Much of the policy effort therefore sought to sustain economic activity and the aggregate supply of jobs, while the statutory representation remained available as a residual means of examining the justice of a particular dismissal. The two forms of protection coexisted, although their institutional scale and public visibility already differed markedly.

The human dimension of productivity then received an institutional programme of its own. On 23 April 1981, the NPB formed the Committee on Productivity to recommend improvements in work attitudes, productivity, and labour-management relations. Its report, published on 11 June 1981, supplied the programme for the national Productivity Movement launched later that year (see Woon and Loo at 119–127; also see From Third World to First at pp 91–92 and 524–525). In examining Japanese practices, the Committee concluded that seniority and lifetime employment might be feasible only for a core of skilled workers in some larger organisations (see Woon and Loo at pp 121–122). Long service, loyalty to the firm, and cooperative labour-management relations retained substantial value, although the model adopted was selective and pragmatic (cf From Third World to First at pp 527–528). Within that setting, the inherited dismissal protection restrained unjustified exercises of power in an existing relationship without converting continued employment into an unconditional promise.

The first period thus ended with a different means of preserving employment. Industrial peace and employment continuity were pursued increasingly through coordinated wages, productivity, enterprise viability, and flexible adjustment. The individual merits inquiry remained legally available, but its Ministerial administration produced no public body of reasons through which the conception of justice could develop. The further step—equipping a worker to move where the work itself could no longer be preserved—had antecedents in the response to the British withdrawal and became increasingly systematic after 1985.

From labour to manpower: flexibility and employability

The NPB programmes introduced between 1986 and 1989 show the beginning of that enlarged and more systematic effort. Launched in 1986, the Core Skills for Effectiveness and Change programme trained workers in communication, personal effectiveness, problem-solving, work economics, computer literacy, and quality, with the stated aim of supplying transferable skills for a rapidly changing workplace. In May 1987, the NPB published Initiatives for Reskilling the Workforce and introduced the Worker Training Plan, encouraging companies to plan and implement systematic training for junior employees. Its Report on Upgrading the Skills of Workers Aged 40–50, published in May 1989, separately addressed training for mature workers (Woon and Loo at 165–167). The common premise was that economic protection increasingly required an ability to perform altered work, whether within the enterprise or after displacement from it.

The plans of the early 1990s extended that premise. The Economic Planning Committee was formed in December 1989 to consider the strategies through which Singapore might attain developed-country status. In March 1990, the National Productivity Council published Productivity 2000, the first formal national productivity plan (see Woon and Loo at p 25), connecting workforce flexibility, multi-skilling, career development, and continuing tripartite cooperation. The Economic Planning Committee’s Strategic Economic Plan: Towards a Developed Nation followed in December 1991 and adopted a 30-to-40-year horizon. Read together, the two plans placed total factor productivity, innovation, higher-value manufacturing and services, and continuing opportunities for education and upgrading within that wider project (see Woon and Loo at pp 203–209). Those plans reflected an increasingly internationalised and service-oriented economy in which knowledge and technology became obsolete more quickly and learning throughout a working career acquired economic importance (see Edwin Lee at pp 560–565; Leong Yew, “A Brief History of the Hub: Navigating between ‘Global’ and ‘Asian’ in Singapore’s Knowledge Economy Discourse” (“Leong”) in Singapore in Global History (Derek Heng & Syed Muhd Khairudin Aljunied eds) at pp 271–274; also see J J Woo, Singapore as an International Financial Centre: History, Policy and Politics (Palgrave Macmillan, 2016) at pp 31–35).

The means by which a worker was expected to make that transition were concrete. A Retraining Voucher Scheme introduced in 1994 trained retrenched workers and matched them with vacancies. The Skills Redevelopment Programme, launched by the NTUC on 20 December 1996 with support from the Singapore Productivity and Standards Board, the Economic Development Board, and the Institute of Technical Education, sought through training and skills certification to improve the employability of workers with lower skills or educational qualifications (see Woon and Loo at p 252). Other programmes attempted to return retirees, economically inactive persons, and retrenched workers to employment (see Woon and Loo at pp 244 and 252). Their work began after, or in anticipation of, displacement: the worker was assisted to enter another relationship once continued reliance upon the former one became insecure.

Two distinct institutional trends help to locate the change in public scale. Referrals to the IAC fell from 122 in 1980 to fewer than 20 in every year from 1993, while training places supported by the SDF rose independently from 32,600 in 1981 to 564,400 in 1998 (see Woon and Loo at p 254). The figures have different denominators and measure different activities, and I rely upon them only for what they show about institutional scale and visibility. By the late 1990s, formal adjudicative referrals had become relatively few, whereas workforce development operated as an ordinary and extensive instrument of economic and social policy.

The institutional vocabulary changed with the field when, on 1 April 1998, the Ministry of Labour became the Ministry of Manpower. The change accompanied an expanded mission which preserved responsibility for harmonious labour relations while adding development of a globally competitive workforce, coordinated national manpower planning, skill upgrading, and the use of foreign manpower to supplement local talent (Speech by Prime Minister Goh Chok Tong at the Singapore Malay Chamber of Commerce and Industry Annual Hari Raya Aidilfitri Dinner (24 February 1998) at paras 40–44; Ministry of Labour, “Ministry of Manpower—Effective 1 April 1998” (Press Release, 31 March 1998)). The former Ministry had long administered wages, productivity, training, foreign labour, safety, and industrial relations. Its new name expressed the enlargement already under way: labour relations had become part of the wider endeavour to plan, develop, and deploy the country’s human resources across the economy.

The Asian financial crisis supplied the immediate setting in which that endeavour became a national manpower strategy. On 31 August 1999, the Government introduced Manpower 21 as its national blueprint for developing what it called “Talent Capital” and securing lifelong employability. Its six strategies concerned integrated manpower planning, lifelong learning, augmentation of the national talent pool, transformation of the workplace environment, development of a manpower industry, and partnership among Government, employers, unions, and the community; a School of Lifelong Learning and one-stop career centres were among the proposed institutions (MOM, “Launch of Manpower 21 Plan” (Press Release, 31 August 1999) at paras 3–11). The plan responded, among other things, to a then-record 29,000 retrenchments in 1998 and the particular difficulty faced by less educated or unskilled workers in finding new work, even while vacancies remained in expanding parts of the electronics sector because available workers lacked the required skills. The crisis exposed a mismatch between the jobs which had disappeared, the capabilities of displaced workers, and the work which remained available (Speech by Deputy Prime Minister Lee Hsien Loong at the Launch of Manpower 21 (31 August 1999) (“Manpower 21 Address”) at paras 4–12; Leong in Singapore in Global History at pp 272–274).

The Lifelong Learning Endowment Fund, announced in August 2000 and established in 2001, supplied a dedicated source of funding for continuing education and training (Edwin Lee at p 579; Paul Vandenberg, “Is Asia adopting flexicurity?” (2010) 149(1) International Labour Review 31 (“Vandenberg”) at 47). The global electronics downturn and recession in 2001 reinforced the need for a permanent administrative response to displacement and structural change. Formed in December 2001 under the chairmanship of Deputy Prime Minister Lee Hsien Loong, the Economic Review Committee reviewed national policy and recommended a strategy for renewed growth. Its report, published in February 2003, anticipated faster restructuring, more frequent displacement, more flexible employment structures, and a growing risk of structural unemployment where workers’ skills became redundant. It proposed continuing education and training under a national body so that Singaporeans could remain employable (see Woon and Loo at pp 261–262; Edwin Lee at pp 572–579; Leong in Singapore in Global History at pp 273–274). The outbreak of SARS later in 2003 added further pressure upon employment and the economy. On 1 September 2003, the Singapore Workforce Development Agency (“WDA”) was formed with a mission to enhance the employability and competitiveness of employees and jobseekers. One arm facilitated employment and industry-specific continuing education and training; another developed the capabilities, curricula, and programmes required by workers vulnerable to unemployment (MOM, “Formation of Singapore Workforce Development Agency” (Press Release, 1 September 2003) at paras 1–4; Woon and Loo at pp 282–283). Labour administration could thereby follow the worker across the boundary between jobs.

Vandenberg supplies a useful, although retrospective, vocabulary for the resulting arrangement. He distinguishes “employment security”, which means security in remaining with a particular employer, from “labour market security”, under which a more moderate degree of employment protection is combined with public support during unemployment and for movement into other work. Employment-protection law seeks in part to prevent job loss, whereas active labour-market measures accept that loss may occur and assist with job search, training, and re-employment (Vandenberg at pp 32 and 34–36). He described Singapore as “flexi-active”: employer flexibility was accompanied by extensive active measures, including the SDF, employment matching, training, and the Lifelong Learning Endowment Fund, although there was no general unemployment insurance and no legislated quantum of severance pay (Vandenberg at 45–47 and 54–55).

By the beginning of the 2010s, slower workforce growth, the physical and social limits of continued foreign-labour expansion, and disappointing productivity growth had altered the problem again. The foreign share of the labour force had risen from 3.2% in 1970 to 34.7% in 2010. Economy-wide productivity growth, measured by value added per worker, then averaged 0.4% annually between 2010 and 2015, against the target of 2% to 3% which the Economic Strategies Committee had set for the decade from 2010 (see Woon and Loo at pp 318–319 and 356–357; also see Barr at pp 160–165). Policy sought to reduce dependence upon further labour inputs and increase the skill, productivity, and wages of each worker (see Woon and Loo at pp 322–323). The Retirement and Re-employment Act 1993, which took effect in January 2012, required eligible older employees to be offered re-employment beyond the minimum retirement age and permitted reasonable adjustment of job scope, wages, and other terms (Tripartite Guidelines on the Re-employment of Older Employees (18 May 2016) at paras 1–2 and 13–18). The statutory design thus preserved participation in work through a relationship capable of adjustment.

The national SkillsFuture effort developed within the same setting. On 17 August 2014, the Prime Minister announced a tripartite committee to develop an integrated system of education, training, and career progression, while promoting advancement according to skill. One month later, on 17 September 2014, the Lifelong Learning Institute opened as a campus for continuing education and training. Deputy Prime Minister Tharman Shanmugaratnam described automation as reshaping tasks across manufacturing, professional work, and services, and treated constant learning as a feature of the good jobs of the future; experience and sustained engagement in a field would remain valuable across changes of role and employer (Speech by Deputy Prime Minister and Minister for Finance Tharman Shanmugaratnam at the Official Opening of the Lifelong Learning Institute (17 September 2014) at paras 5–34). On 5 November 2014, the tripartite committee, subsequently named the SkillsFuture Council, identified four principal objects: informed choices about education, training, and careers; an integrated system responsive to changing industry needs; employer recognition and career development based upon mastery; and a culture of lifelong learning. Further initiatives followed during 2015 and 2016 (see Woon and Loo at pp 359–360).

Those policies preserved continuity through several routes: wage flexibility and training could sustain the same job; retraining within the enterprise and redeployment could sustain the relationship through altered work; re-employment could continue participation on adjusted terms; and employment services could assist movement to another employer. Manpower 21 expected firms to invest in their workers’ capabilities, while much of the SDF supported employees who remained at work. The organising interest was productive participation as the firm, job, and worker adapted, with the existing contract serving as one important site through which that interest could be realised.

The Committee on the Future Economy was convened in January 2016 to consider the strategies required for the next phase of development. Before that Committee reported, the Budget announced on 24 March 2016 an Industry Transformation Programme which carried the integration of productivity and skills to the level of individual sectors. Its Industry Transformation Maps were designed to combine productivity, skills, innovation, and internationalisation, with job redesign and upgrading opportunities forming part of the transformation of 23 industries. A tripartite Council for Skills, Innovation and Productivity was established on 20 May 2016 to oversee that work. On 4 October 2016, WDA was reconstituted as Workforce Singapore to drive workforce transformation, while responsibility for the SkillsFuture movement passed to the newly formed SkillsFuture Singapore (Woon and Loo at p 333 n 7 and pp 359–367). When the Committee on the Future Economy reported on 9 February 2017, it placed deep skills and the Industry Transformation Maps among the strategies for the following decade and described tripartism and good industrial relations as enduring economic strengths (Woon and Loo at 319–323 and 378–380). The re-employment age was raised from 65 to 67 in July 2017 (MOM, “Improving Re-employment Opportunities for Older Workers” (Press Release, 9 January 2017) at paras 3–4).

The conception apparent in 2018

The Bill itself provides the proper starting point. In her opening speech, the Minister described three principal areas of reform: extension of the EA’s core provisions to all managers and executives, thereby covering about 430,000 additional employees; extension of Part IV protection to approximately 100,000 additional non-workmen; and enhancement of the dispute-resolution framework by transferring adjudication of wrongful dismissal from MOM to the ECT. Salary and dismissal disputes could then be resolved through a one-stop service. The cases which MOM had decided were said to disclose broad principles reflecting employer-union consensus. Since those cases had not been published, the principles and illustrations would be placed in future Tripartite Guidelines which the ECT had to consider. The forms of relief—reinstatement with lost wages or compensation—would remain. For managers and executives dismissed with notice, the qualifying period would be reduced from twelve to six months; forced resignation was described as part of the cases already heard by MOM and would be stated expressly in the definition of dismissal (2018 Debates at 3.56pm (Mrs Josephine Teo, Minister for Manpower)). The package was therefore protective as well as institutional: it extended personal coverage, shortened the qualifying period, and moved adjudication into an accessible tribunal.

In her reply after the Members had spoken, the Minister expressly situated that package within continuity at the level of the EA’s general economic purpose. Quoting Mr Rajaratnam’s account of the 1968 Act, she said that the fundamentals had not changed: employment practices were to be regulated in a balanced manner which protected workers while preserving labour-market flexibility so that businesses could thrive and create good jobs. Legislation operated alongside Tripartite Guidelines, Standards, and Advisories, and tripartism was presented as the continuing method through which the balance would be maintained (2018 Debates at 7.03pm (Mrs Josephine Teo)). Mr Patrick Tay, who had served on the tripartite workgroup, similarly placed the EA within the history of industrialisation, productivity, foreign investment, and stable labour conditions. He described universal coverage of managers and executives as both a watershed and a radical advance in employment standards, welcomed greater transparency about dismissal and the one-stop ECT route, raised practical concerns about access and enforcement, and sought an effective role for union assistance (2018 Debates at 4.18pm (Mr Patrick Tay Teck Guan)). Their historical accounts were therefore directed to the general purposes and development of the EA.

The workforce to which the amended Act would apply also placed dismissal in a different social setting. Professionals, managers, executives, and technicians comprised approximately 56% of the local workforce and were projected to comprise 65% by 2030. At different points in the debate, Members referred to regional roles, reskilling, redeployment and possible retrenchment, remote and flexible work, multiple employers and portfolio careers, platform work, automation, and digitalisation (2018 Debates at 3.56pm (Mrs Josephine Teo), 4.43pm (Mr Douglas Foo), 4.59pm (Mr Zainal Sapari), 5.08pm (Assoc Prof Walter Theseira), 5.50pm (Mr Saktiandi Supaat), 6.01pm (Mr Arasu Duraisamy), 6.45pm (Ms Joan Pereira) and 6.49pm (Ms Denise Phua Lay Peng)). Extending dismissal protection to 430,000 managers and executives was a considerable expansion of substantive coverage. The questions concerned the content and administration of that enlarged protection within a labour market in which changes of work, skill, and employment form were familiar premises.

Mr Douglas Foo explained the employer perspective from which some of those questions arose. Speaking as President of the Singapore Manufacturing Federation, Vice-Chairman of the Singapore Business Federation, and Vice-President of the SNEF, he referred to consultation with more than 3,000 employers and human-resource practitioners. He supported the extension of the EA and the tripartite process, although he reported particular employer concern about the extension of s 14, frivolous claims, and the need for clear principles, prompt and even-handed investigation, and stronger human-resource documentation. His broader concern was that firms operating across markets had to remain lean and agile, with the capacity to reskill, redeploy, or retrench as business needs changed, without rigid regulation obstructing managerial flexibility (2018 Debates at 4.43pm (Mr Douglas Foo)). That account helps to explain why clarity, screening, and flexibility occupied a prominent place in the employer response to the reform.

Assoc Prof Daniel Goh raised the questions which came closest to the legal difficulty considered in these grounds. He asked about the legal status of the future Guidelines and whether mandatory consideration might constrain the ECT unduly. He also asked whether the Guidelines would address discrimination, including mental-health discrimination, sexual harassment, disguised retrenchment and a false assertion of redundancy, and raised the available remedy and claim limit (2018 Debates at 5.40pm (Assoc Prof Daniel Goh Pei Siong)). Mr Saktiandi Supaat separately treated resignation induced by bullying or sexual harassment as a matter which should be heard, while calling for fair investigation and safeguards for employers. Dr Intan Azura Mokhtar sought clearer protection in cases concerning discrimination, pregnancy, maternity leave, and forced resignation (2018 Debates at 5.50pm (Mr Saktiandi Supaat) and 6.16pm (Dr Intan Azura Mokhtar)). Taken together, their proposals, questions, and illustrations reveal the breadth of the wrongs which Members wished or asked the new process to address.

Mr Dennis Tan combined support for statutory redress with concern about its operation. He supported coverage of managers and executives and the transfer to the ECT, had previously advocated a tribunal route for unfair-dismissal and discrimination claims, and accepted the object of protecting an employee forced to resign. He nevertheless questioned whether the proposed language was too broad, sought sound evidence and procedural safeguards, and raised representation and the monetary cap. His account of one MOM matter concerned an employer which possessed a record of disciplinary issues and could have dismissed upon that ground, but chose instead to give contractual notice and pay the employee’s entitlements. According to Mr Tan, MOM then told the employer to provide an ex gratia payment. He regarded the contractual compliance and disciplinary record as important and hoped that mediation under the new system would recognise the parties’ contractual rights (2018 Debates at 4.33pm (Mr Dennis Tan Lip Fong)). His speech thus placed a contractual emphasis upon that example, while supporting a legally sound tribunal process for unfair dismissal and discrimination.

The Minister answered those different concerns by identifying substantive categories. In response to Mr Ng and Mr Tan on what constituted dismissal without just cause or excuse, and Assoc Prof Goh on the Guidelines, she said that dismissals on grounds such as discrimination, deprivation of statutory benefits, and punishment for exercising employment rights were wrongful, whereas poor performance, misconduct, and redundancy were not. Since an exhaustive statutory definition was regarded as impracticable, the Guidelines would state principles and parameters and supply illustrations which the ECT had to consider. Claims founded upon forced resignation or discriminatory dismissal could proceed through TADM and, if unresolved, the ECT (2018 Debates at 7.03pm (Mrs Josephine Teo, Minister for Manpower)). The examples were illustrative and left the categories open, although the answer presented the protection through a taxonomy of accepted business grounds and identified wrongful grounds or purposes. It did not address Assoc Prof Goh’s separate questions concerning false or disguised redundancy, or whether the ECT could depart from the Guidelines in an appropriate case; nor did it examine how an employer’s asserted ground would be tested where notice had been given.

Concern about unmeritorious claims produced a separate answer concerning the gateway to mediation. Responding to Mr Foo and Mr Tan, the Minister explained that, where an employer had met its contractual termination obligations, TADM would first require the employee to identify specific incidents substantiating a wrongful case before calling upon the employer to attend mediation. An unresolved case could then proceed to the ECT. In answer to questions from Mr Ng and Assoc Prof Theseira, she said that the ECT could draw adverse inferences where an employer failed to produce documents, such as key employment terms and salary slips, which specified employment laws required it to maintain; she also answered affirmatively a question about an employer’s past behaviour, without explaining the legal basis or limits of that answer (2018 Debates at 7.03pm (Mrs Josephine Teo, Minister for Manpower)). The first part of the answer described an administrative screen before mediation. However, ECT’s adjudicative function is governed by ss 20(2), 20(5), 20(7)(a) and 21(5) of the ECA, which confer powers to identify issues, obtain evidence, and inquire into relevant matters while requiring regard to the Guidelines.

MOM elaborated the burden categories in an implementation guide stated to be correct as at 29 November 2018. The guide should therefore be attributed to the Ministry as a subsequent administrative exposition. It began from the proposition that employer and employee possessed a right of contractual termination. For dismissal without notice, the employer bore the burden of showing cause. Where notice or salary in lieu had been given and the employer stated a reason affecting employability, such as poor performance or misconduct, the employer had to substantiate it. Where notice had been given and no reason stated, the employee bore the burden of proving that the dismissal was wrongful (Tripartite Alliance for Fair & Progressive Employment Practices and MOM, Changes to the Employment Act with effect from 1 April 2019: An implementation guide for employers and HR practitioners (29 November 2018) at 10–12). The guide itself stated that it was not a complete statement of the law. Its immediate importance is administrative and explanatory: it stated with precision the model which MOM intended the new process to apply, subject to the Act, the later Guidelines, and adjudication by the ECT.

The institutional form of the reform reflected the same combination of statutory protection and tripartite administration. The transfer to the ECT offered a one-stop forum for connected salary and dismissal claims; the existing remedies were preserved; and the Minister promised additional TADM and ECT manpower and training. Members supported access and transparency, although their differences concerned claim limits, representation, evidential quality, contractual rights, the risk of abusive claims, and the legal relationship between the Guidelines and adjudication (2018 Debates at 4.18pm (Mr Patrick Tay), 4.33pm (Mr Dennis Tan), 4.43pm (Mr Douglas Foo), 5.40pm (Assoc Prof Daniel Goh), and 7.03pm (Mrs Josephine Teo)). Parliament enacted a duty to have regard to tripartite guidance before the Guidelines themselves had been published. The design thereby placed a general expression inherited from industrial adjudication within an individual-claims forum intended to be expeditious and affordable, while leaving its operational content to be supplied in part by principles and illustrations which, on the Minister’s account, distilled unpublished administrative practice and tripartite agreement.

Read as a whole, the 2018 debate retained substantial concern with unequal bargaining power, access to redress, and discriminatory or abusive dismissal, although it proceeded within a policy setting in which restructuring, reskilling, employability, and management flexibility were familiar premises. The Minister’s taxonomy and MOM’s implementation guide began from contractual mobility, distinguished accepted business grounds from specified wrongful grounds, and placed an initial burden upon an employee who received notice without reasons. That account helps to explain the implementation of the reform, even though the amended Act retained the older words and remedy of reinstatement while enlarging its personal scope.

What had faded, therefore, was the older assumption that the employer’s reason would be elicited and examined whether or not it had been volunteered. Employment security came to be pursued across jobs through adaptable firms, wages, skills, matching, redeployment, and re-employment, while tripartism expanded into wider labour-market governance. Yet, s 14(2) continued to ask whether dismissal was supported by just cause or excuse and to authorise reinstatement. The expression “wrongful dismissal” consequently came to carry the inherited protection, the later administrative conception, and the common law action despite their different sources and purposes. It is to that polyseme that I now turn.

The unhelpful polyseme and the unstable hybrid

The preceding sections disclose three senses in which “wrongful dismissal” has come to be used. At common law, the expression ordinarily denotes a breach of the contract governing termination; under s 14(2), it describes dismissal without just cause or excuse; and, in the Guidelines and the account given in 2018, it refers to a more particular protection organised around notice, communicated reasons, and specified forms of impropriety. As I observed at [39]⁠–⁠[42] and [59]⁠–⁠[61] above, the polyseme causes difficulty because incidents belonging to the contractual and administrative senses have entered the administration of the statutory protection without any settled account of what properly belongs to each. I begin with the contractual sense before drawing together the effects upon notice, proof, and remedy which have already been analysed.

At common law, wrongful dismissal ordinarily means that the employer ended the contract in a manner which the contract did not permit, of which dismissal without the required notice supplies the paradigm. Sufficiently serious misconduct may entitle the employer to dismiss immediately, while an employment of indefinite duration may ordinarily be brought to an end by notice or an authorised payment in lieu. Cause therefore matters principally where the employer seeks to dispense with notice; a termination which complies with notice ordinarily requires no justification by reference to fairness or industrial sufficiency (see Zoe Adams et al, Deakin and Morris’ Labour Law (Hart Publishing, 7th Ed, 2021) at paras 4.12–4.14 and 4.18; Harvey, ch 8 at paras [389]–[391.02] and [409]–[409.01]; also see D’Cruz at 156D).

That conception arose from the need to bring an indefinite hiring to an end. By the middle of the 19th century, the common law implied reasonable notice where the contract stated no fixed end, commonly deriving its length from the wage period or trade custom (see Baxter v Nurse (1844) 6 Man & G 935; Metzner v Bolton (1854) 9 Exch 518 at 521). The reciprocal form of the rule gave each party a lawful route out, although the practical consequences of an employer losing one worker and a worker losing her livelihood were seldom equal. Notice nevertheless came to mark the means of termination without cause and the temporal extent of the employee’s ordinary contractual interest in continued employment (see Michael Ford et al, “Re-Thinking the Notice Rule” (1998) 27(3) Industrial Law Journal 220 at 220 and 227–232).

Addis gave that conception much of its remedial force by limiting recovery to salary and commission referable to notice and excluding injured feelings and later employability loss. The same premise continues in Singapore through the orthodox minimum-obligation rule: the employee ordinarily receives what she would have obtained had the employer adopted its least onerous lawful route to termination, and continued employment is not specifically enforced, so that notice tends to define both breach and loss (see Wee Kim San at [18], [25]–[28], [34] and [36]–[39]; Alexander Proudfoot Productivity Services Co S’pore Pte Ltd v Sim Hua Ngee Alvin [1992] 3 SLR(R) 933 at [13]; Wong Sung Boon v Fuji Xerox Singapore Pte Ltd [2021] SGHC 24 at [120]–[121]).

The contract and wider law may, of course, qualify that ordinary conception. A fixed term, an express requirement of cause, or a binding disciplinary procedure may constrain termination, and exceptional declaratory or injunctive relief may also be available (see McClelland; Gunton v Richmond-upon-Thames London Borough Council [1980] ICR 755 (“Gunton”); Aldabe Fermin v Standard Chartered Bank [2010] 3 SLR 722 (“Aldabe Fermin”) at [90]–[94] and [108]–[116]; Tesco Stores Ltd v Union of Shop, Distributive and Allied Workers [2024] IRLR 998 at [43], [51]–[84] and [122]–[129]). Other jurisdictions have recognised good faith, proportionality, or statutory protection extending beyond notice (see Honda Canada Inc v Keays [2008] 2 SCR 362 at [50]–[60]; McKinley v BC Tel [2001] 2 SCR 161 at [48]–[57]; Wilson v Atomic Energy of Canada Ltd [2016] 1 SCR 770 at [62]–[68]). Although those qualifications confirm that the label carries no universal or indivisible doctrine, its ordinary notice-centred associations remain powerful.

Notice within the statutory protection

Section 14(2), however, proceeded upon a different premise because, as the inherited settlement at [283]⁠–⁠[347] above shows, notice could affect the legal effectiveness of termination, wages, and relief without insulating the employer’s justification from examination. The 1973 protection was enacted because limited notice pay failed to deter unjust dismissal, and the present definition expressly includes termination “with or without notice and for cause or otherwise”. Sections 10, 11, and 14(1) regulate distinct incidents of termination, whereas ss 14(2)–(3) subject dismissal to a separate standard of justification and authorise relief despite contrary law or agreement (see [92]⁠–⁠[95] above).

Once the inherited industrial conception faded, however, the common label made notice a familiar premise around which the protection could be organised. Paragraphs 3 and 7 of the Guidelines begin from the parties’ reciprocal contractual power and presume a dismissal with notice not to be wrongful; para 9 then requires an employee who received no reason to substantiate a positively wrongful one. As I explained at [97]⁠–⁠[116] above, the effect is to move contractual compliance from the manner of termination towards the presumptive answer to the distinct inquiry under s 14(2). Communicated reasons, their truth or falsity, categories such as misconduct and redundancy, and due inquiry then assume consequences which the enacted standard does not invariably assign to them.

Notice also enters the statutory scheme through proof, although s 27(2) gives it only selected effects. As explained at [119]⁠–⁠[171] above, para (a) addresses dismissal without notice under s 14(1), while para (b) addresses notice given or purportedly given for poor performance or misconduct. The gaps become acute where notice is satisfied and no reason given, where another reason is advanced, or where salary in lieu makes the case difficult to classify. A reason may bound the statutory negative and the evidence may call for an answer; nevertheless, the ECT’s active powers and the ordinary movement of an evidential burden cannot create the general duty or burden which Parliament omitted. Notice therefore performs contractual, classificatory, and evidential work, without itself furnishing a coherent substantive test.

The remedial hybrid

The same instability is apparent in the remedial position. At common law, notice pay ordinarily substitutes for the contractual performance denied and compelled continuation is resisted. Under the inherited industrial protection, reinstatement, continuity, and lost wages occupied the centre of an inquiry which could reverse dismissal. Present labour policy may, for its part, favour viable restructuring and movement into other work. Section 14 retains both reinstatement and compensation, although it does not explain how those different remedial premises should be ordered (see [172]⁠–⁠[183] and [348]⁠–⁠[387] above).

In practice, considerations familiar from specific performance—the availability of the role, hardship, clean hands, trust, and the feasibility of continued relations—have consequently entered the approach to reinstatement, although the statutory power was preserved despite contrary rules of law. Those matters may properly bear upon discretion, but the absence of statutory criteria and the rarity described at [173]⁠–⁠[184] leave it uncertain whether reinstatement remains the primary industrial remedy, has become an exceptional response, or survives textually after its practical work has largely passed to compensation.

Compensation is similarly under-specified because the decisions surveyed at [185]⁠–⁠[213] use notice sometimes as the measure of income loss and sometimes as a credit against duplication; they adopt different counterfactuals and assign different meanings and starting points to the base amount for harm. The Second Schedule recognises consequences which Addis excluded, but states neither the counterfactual for income nor the juridical object of harm, and supplies no stable rule for overlap. Contractual compensation, income replacement, a monetary substitute for restoration, and personal or reputational harm therefore remain possible organising ideas without an enacted order between them.

The unstable hybrid

The present scheme therefore combines four conceptions: the inherited merits-based industrial protection; the common law action carried by the shared label; the later manpower conception of security through adaptable firms, skills, redeployment, and employability; and the 2018 arrangement of retained statutory words, notice-based Guidelines, selected burdens, judicial adjudication, and compensation for income and harm. Although each conception may be intelligible on its own terms, their combination is unstable because the scheme does not state their juridical priority.

That instability may work against either party, since a dismissal may escape examination although no sufficient cause has been identified, or attract liability because a communicated reason failed without any inquiry into whether another cause actually existed and justified dismissal. Reinstatement remains first in the statutory sequence but has become rare, while compensation is ordinary without a settled account of what it restores or replaces. An industrial administrator could perhaps absorb some of those tensions by eliciting reasons, conciliating, and applying specialist judgment without developing a publicly articulated rule for each intermediate step. Once the same hybrid was transferred to a court required to interpret enacted words, assign burdens, give reasons, and exercise remedies through legally intelligible criteria, however, the absence of a juridical order could no longer be avoided. It is to that institutional transfer that I now turn.

The institutional transfer

The question posed at the end of the preceding section extends beyond the identity of the decision-maker. Before 1 April 2019, the protection in s 14(2) was administered by the Minister for Manpower through machinery substantially inherited from the industrial-relations settlement described above. After that date, and subject to mandatory pre-claim mediation and the transitional arrangements, a claim invoking the same words had to be heard and determined by the ECT. An executive decision-maker and a court may each be required to apply an open-textured statutory standard, and neither may substitute personal preference for law. The institutions through which they act nevertheless carry different premises, methods, and duties. I therefore begin with the industrial-relations character of the Ministerial decision, before considering what changes when a court must determine the dispute and whether the 2018 Amendment Act supplied the legal and institutional translation which that transfer required.

The administrative decision

Before 1 April 2019, s 14(2) created no ordinary civil action before a court. It permitted written representations to the Minister, who could direct reinstatement with lost wages or compensation if satisfied that dismissal lacked just cause or excuse; the decision was final, a direction barred an ordinary damages action, and non-compliance attracted penal enforcement (see Lim Tow Peng at [10] and [19]–[21]). The closely related route under s 35(3) remains in that form for an eligible unionised employee. As the history above shows, the employee invoked public intervention, while the determination and restorative power belonged to the Minister.

The Minister nevertheless exercised a statutory power bounded by the language and purpose of the provision, relevant considerations, rationality, and natural justice (see Chng Suan Tze v Minister for Home Affairs [1988] 2 SLR(R) 525 at [86] and [119]; Tan Seet Eng v Attorney-General [2016] 1 SLR 779 at [97]–[99]; SGB Starkstrom Pte Ltd v Commissioner for Labour [2016] 3 SLR 598 at [55]–[58]). Stansfield Business International Pte Ltd v Minister for Manpower [1999] 2 SLR(R) 866 confirms that the former s 14 process remained reviewable despite the finality clause, and required the employer to be informed of the material allegations and afforded a fair opportunity to answer them (at [21]–[22] and [26]–[34]). I use the descriptions “administrative” and “industrial” only to identify the setting and lawful range of judgment within which the statutory criterion operated.

Within those limits, the criterion entrusted an open-textured judgment to an administrator who combined responsibility for labour policy with access to departmental inquiry and conciliation. The Minister had to be “satisfied” that dismissal was without just cause or excuse, but the provision defined no permissible causes, prescribed no closed sequence of legal tests, and assigned no relative weight among managerial need, the employee’s conduct and record, proportionality, the continuing viability of the relationship, and the consequences of dismissal. The available orders required further evaluative choices, principally whether reinstatement remained appropriate and what compensation in wages should replace it if it did not. The Minister therefore had to apply the broad statutory standard through specialist experience and such lawful policy considerations as bore upon that standard and its remedy.

That judgment operated within the field of industrial relations because s 35 appears in legislation governing collective bargaining, conciliation, and settlement of trade disputes, while the representation was created as part of the 1968 arrangement which removed dismissal from collective bargaining and the IAC. The materially similar protection extended to non-unionised employees in 1973 was to be administered through the same procedure (see [337]⁠–⁠[345] above). Its industrial character was therefore historical and institutional, despite the appearance of s 14 in an individual-employment statute.

The procedure described at [337]⁠–⁠[345] above gave practical expression to that industrial-relations setting. The Industrial Relations Section sought the employer’s grounds, screened the representation, and attempted conciliation before any formal inquiry or Ministerial determination. Of the 109 representations recorded between August 1968 and July 1973, 52 settled by conciliation and only seven produced orders. I cannot assume that every later case followed the same course. The available account nevertheless shows a process which elicited an explanation and sought an industrial adjustment before finally determining the statutory criterion.

The IAC awards discussed at [287]⁠–⁠[308] above supplied an available industrial background. They examined the employer’s actual reason, its truth and gravity, proportionality, the employee’s record, operational needs, and the suitability of relief, while contemporary commentary described a process directed towards compromise and improved industrial relations (Chia in LIRS (1970) at 82–85). The Minister was not bound by those awards, and the substitute process did not reproduce the IAC’s institutional features. Even so, it was enacted and first administered in their immediate shadow.

The Minister’s understanding of industrial justice could therefore operate through an industrially informed evaluative judgment of the statutory question. An administrator entrusted with labour relations might conclude, from accumulated cases and experience, that habitual lateness justified discipline but not dismissal in the particular circumstances; that an operational redundancy was genuine despite its severity for the employee; that an accusation of dishonesty had to be proved on the available evidence; or that reinstatement had become impracticable despite an unjustified dismissal. Such conclusions involved judgment about the proper accommodation of managerial authority, employee security, and workplace discipline. Industrial peace and wider labour policy could bear only so far as they were lawfully connected to just cause or excuse or the authorised remedy. So confined, the industrial premise occupied a natural place: the phrase was general enough to be applied in light of specialist administrative experience without first being reduced to a comprehensive taxonomy.

Those qualifications do not displace the institutional point. The former arrangement could apparently operate without a publicly articulated rule for each intermediate question because it was structured to obtain the employer’s account, seek agreement, use formal inquiry where necessary, and entrust an industrially informed executive judgment to the Minister. Conciliation could produce a pragmatic accommodation which restored the relationship or settled the dispute. Where the Minister made an order instead, he still had to be satisfied that the statutory criterion was met before selecting an authorised remedy, although he did not thereby have to create a publicly precedential doctrine for the next case. The old conception, and its later admixture with contractual practice, may therefore have been workable within an institution whose purposes included conciliation and industrial settlement, albeit without assurance of transparency or consistency.

The judicial decision

The transfer to the ECT changed the office in which the dispute was decided, for employment remained the subject matter of the controversy while a court became the institution which had to decide it. The ECT may facilitate conciliation or refer the dispute for mediation, and the parties remain free to compromise. Once adjudication becomes necessary, however, the ECT must determine the statutory right asserted and the order which the law authorises.

The ECT bears the central marks of a court exercising judicial power. It is a subordinate court within the State Courts system and is presided over by a legally qualified Tribunal Magistrate. Sections 2(1), 8–9, 12, 20–23, and 27 of the ECA confer jurisdiction to hear and determine claims, require the ECT to identify relevant issues and ensure that evidence is adduced, permit it to summon witnesses and documents and inquire into relevant matters, assign burdens of proof, and authorise coercive orders enforceable as District Court orders; subject to permission, an appeal lies upon a question of law or jurisdiction. In Qureshi Mohamed Aslam v Majlis Ugama Islam, Singapura [2026] SGCA 19 (“Qureshi”), the Court of Appeal explained that a judicial function characteristically involves a controversy, findings on facts as they stand, application of the relevant law, and a binding determination of the parties’ rights and obligations through a process involving evidence and legal argument (at [47]–[50], applying Mohammad Faizal bin Sabtu v Public Prosecutor [2012] 4 SLR 947 at [27]), an account which accords with the function entrusted to the ECT.

Fuller’s account helps to explain why the form of participation matters. The parties offer proofs and reasoned arguments, and their participation creates a demand that the resulting decision withstand scrutiny for rationality and principle (Lon L Fuller, “The Forms and Limits of Adjudication” (1978) 92 Harvard Law Review 353 (“Fuller (Adjudication)”) at pp 364–370). Accordingly, even where a judge has facilitated settlement, an eventual adjudication must answer the legal claim by reasons capable of explaining why like cases should receive like treatment.

Cranston’s recent account begins from substantially the same institutional distinction. Judicial work combines impartial dispute resolution with maintenance of the rule of law: judges find facts, apply legislation and precedent, and give reasons which expose the legal route to the result. Independence and impartiality allow decision without improper influence, while accountability follows through decision according to law, reasoned judgments, review, and—subject to lawful qualifications—public scrutiny. The law remains controlling even where the judge considers its policy unsatisfactory. Fidelity to enacted legislation and precedent protects litigants from an idiosyncratic or forensic lottery, while defects which the statutory words cannot bear remain for Parliament to cure (Ross Cranston, Judging (Oxford University Press, 2025) (“Cranston”) at pp 46–54, 67–70, 240–242, 273–275, and 279–283). Those requirements supply the conditions upon which a court may authoritatively determine one party’s rights against another.

First, fidelity to the governing law requires the ECT to construe “without just cause or excuse” in its statutory context, identify the purpose at the appropriate level, and select the textually permissible meaning which best advances that purpose (see Tan Cheng Bock at [35]–[54]; Re Blackstone Asia Real Estate Partners Ltd [2026] SGCA 12 at [23]–[26] and [51]–[55]). History, industrial setting, and labour policy may assist within that exercise. The court must also have regard to the Guidelines under s 20(7)(a), while preserving the hierarchy which prevents them from amending s 14(2) or displacing the court’s own construction.

Second, although the ECT is freed from the ordinary rules of evidence and may identify issues, question parties, summon evidence, and inquire into relevant matters, fidelity to facts and fairness requires its findings to rest upon a rational evidential basis and observe the burdens which Parliament prescribed. Material obtained upon the ECT’s initiative must be disclosed; each party must have a meaningful opportunity to meet a decisive case; and the active powers cannot alter a legal burden under s 27 (see Cooperatieve at [30]–[31]; ss 20(2), 20(4)–(5), 21(1), and 21(5)–(6) of the ECA; Soh Beng Tee & Co Pte Ltd v Fairmount Development Pte Ltd [2007] 3 SLR(R) 86 at [65(a)] and [65(d)]–[65(e)]; Bintai Kindenko (at [34]–[47]). Those obligations assume particular importance where parties appear without lawyers and the employee may not know the employer’s reason.

Third, the ECT must give an explanation adequate to the decision, although its content will depend upon the circumstances. In Qureshi at [42]–[51], the Court of Appeal identified the enduring functions of that duty: relevant submissions and evidence are confronted; the parties understand why they won or lost; like cases can be anticipated; appellate review becomes possible; and the exercise of judicial power may be debated and held accountable. The same functions were explained in Thong Ah Fat v Public Prosecutor [2012] 1 SLR 676 at [20]–[25] and [41]–[45]. The ECT’s high-volume and informal setting may justify concise or oral reasons in a straightforward case, and the ECA does not require a published written judgment in every claim. Nevertheless, s 21(8) requires a summary of the facts and issues and notes of the proceedings, s 22(10) contemplates a statement of the ECT’s grounds, and r 29(1) of the Employment Claims Rules 2017 requires certified written grounds after a notice of appeal. Since an appeal is confined to law and jurisdiction and the appellate court may neither reverse a factual finding nor receive further evidence, the need to identify the material findings, legal test, and decisive route to the result is particularly acute.

Fourth, consistency within the legal order requires binding appellate authority to be followed and materially similar cases to be treated alike unless a legally relevant distinction is explained (see See Toh Siew Kee v Ho Ah Lam Ferrocement (Pte) Ltd [2013] 3 SLR 284 at [35]). Published decisions of coordinate ECTs may carry less authority than an appellate decision, although they still make agreement, distinction, or reasoned disagreement possible. They also allow the law to develop incrementally as successive cases disclose the operation of a general standard. Cranston describes legislation and precedent as constraints against forensic lottery, and Fuller explains why a claim of right necessarily invokes a principle capable of extending beyond the immediate demand (Cranston at 240–242 and 279–283; Fuller (Adjudication) at pp 368–370). An unpublished Ministerial disposition could settle the individual representation without performing that systemic work. A judicial decision should therefore supply reasons capable of principled application even where no public administrative doctrine exists.

Finally, remedial choice is also a judicial act because ss 14(2)–(3) of the EA authorise the employee to claim either reinstatement together with an amount equivalent to the wages which would have been earned but for dismissal, or compensation, while s 22 of the ECA supplies the basis for the ECT’s orders and s 20(7)(b) requires compensation to be calculated in accordance with the EC Regs. The ECT must identify the statutory wrong, the remedial purposes and limits supplied by the enactments, the evidence bearing upon the available orders, and the considerations which justify its choice. The common law’s reluctance to compel a continuing employment relationship cannot silently determine a statutory power which expressly authorises reinstatement; nor can an appeal to industrial justice support an order beyond the statute. The legal basis of the choice must be capable of explanation, particularly because an error concerning the scope of the power or a mandatory consideration may constitute a question of law even where the evaluative weight assigned to a fact does not.

Open-textured standards are familiar subjects of judicial determination, and the ECT may apply one through the informal, private, lawyer-free, and judge-led procedure which Parliament deliberately created. Industrial practice and labour policy may inform the meaning or application of s 14(2) where the statute permits. The court must, however, express their relevance through reasons connected with the enacted standard, the facts proved, and a result capable of principled application beyond the immediate case.

The significance of the transfer

The 2018 Amendment Act converted the former right to make representations to an executive Minister into a claim which a court had to hear and determine, while leaving substantially intact the substantive formula and principal restorative remedies. The final decisional stage was thereby removed from the industrial-relations administrator and placed within an institution whose ordinary work required the general standard to be expressed as law. The new scheme preserved an administrative front end through mandatory TADM mediation, fact-gathering, and officer participation, although it imposed no general statutory requirement that the employer’s operative reason be elicited and communicated. If mediation failed, the ECT had to determine the right asserted, the facts which engaged it, the burdens governing those facts, and the legal purpose of the available remedies. The protection had thus acquired judicial incidents, whether or not Parliament used the word “judicialisation”.

The history of the ECT makes the change more conspicuous. When the ECA was enacted in 2016, the Minister distinguished salary-related disputes—such as unpaid salary, allowances, commissions, and salary in lieu of notice—from non-salary disputes, including unfair dismissal and employee grievances. The consultation proposal expressly excluded unfair dismissal and discrimination because the ECT was intended as an expeditious salary-claims mechanism, although the consultation response and Parliamentary debate contemplated a later expansion to non-salary disputes (MOM, “Proposed Establishment of an Employment Claims Tribunal” (25 February 2016) at paras 7–8; “Summary of Responses to Key Feedback from Public Consultation on Employment Claims Tribunals (25 February–23 March 2016)” (“2016 Consultation Response”) at para 5). The Minister accordingly said that the ECT would “take over the role of the Labour Court and not beyond that”, and repeatedly described it as a cheaper, better, and faster means of deciding statutory and contractual monetary entitlements. Tribunal Magistrates would be legally qualified and decide according to legislation, case law, and legal principles, while the judge-led and lawyer-free design would make relatively straightforward claims accessible (Singapore Parliamentary Debates, Official Report (16 August 2016) vol 94 (“2016 Debates”) (Mr Lim Swee Say, Minister for Manpower)). Later expansion may have been contemplated, but the ECT’s original design centred upon relatively straightforward salary claims. An inherited standard governing the justification of dismissal, with the possible restoration of employment, presented a materially different undertaking.

The public rationale advanced in 2018 was principally practical: the consultation described an employee with both salary and dismissal disputes as having to approach “two different parties” and asked how the process could be made “more streamlined”, while the response paper recorded that most respondents considered it “more expedient” for the ECT to hear both and that the tripartite partners agreed to the transfer (MOM, “Public Consultation on the Employment Act Review” (18 January 2018) at para 6; MOM, “Summary of Responses to Key Feedback from Public Consultation on the Employment Act (18 January–15 February 2018)” (2018) (“2018 Consultation Response”) at paras 13–14). The Second Reading account proceeded in the same terms: salary and dismissal disputes were often related, and moving adjudication from MOM to the ECT would give employees and employers a convenient “one-stop service”, although TADM mediation, the ECT proceeding, and the separate claims remained distinct steps. MOM’s accumulated cases were said to put into practice broad principles reflecting employer-union consensus, which would be carried forward through Tripartite Guidelines containing illustrations which the ECT had to consider. Reinstatement and compensation remained available, and the scope of the remedies was said not to change (2018 Debates at 3.56pm (Mrs Josephine Teo, Minister for Manpower)). Those legitimate aims may therefore explain the consolidation of the routes, although they did not work out the substantive consequences of converting an administrative representation into a judicial claim.

Parliament plainly appreciated that the new forum carried substantial consequences. The original design materials had expressly addressed the application of legislation, case law, and legal principles, judge-led receipt and compulsion of evidence, grounds of decision, records, and appeals (2016 Consultation Response at paras 19–20; 2016 Debates). The amended scheme drew upon the ECA’s existing assessor and judge-led powers, claim limits, permission-based appeal, general prohibition on legal representation, and private-hearing rule. It added, among other things, participation by a specified EA officer, selected adverse-inference and burden provisions, discretionary publication, and directions that the ECT and High Court have regard to the Guidelines and calculate compensation in accordance with the regulations; the amended EA retained reinstatement, lost wages, and compensation despite contrary law or agreement. Under r 6 of the Employment Claims (Amendment) Rules 2019, a report on the findings of the approved mediator or authorised person had to accompany the wrongful dismissal claim, while rr 11–12 of the Employment Claims (Amendment) Regulations 2019 prescribed the compensation mechanics. Tribunal Magistrates remained legally qualified and appointed within the State Courts structure. In reply to concerns about implementation, the Minister said that the ECT and TADM would expand their manpower and that the mediators and investigation officers involved would receive relevant training (2018 Debates at 7.03pm (Mrs Josephine Teo, Minister for Manpower)). Parliament therefore recognised the judicial character of the new decision and made substantial procedural adaptations for it.

The difficulty lies in the matters which those adaptations left untranslated. The 2018 Amendment Act did not define “just cause or excuse”, identify the elements of the statutory wrong, state the causal or justificatory inquiry, explain the relationship between notice and justification, or prescribe how an employer’s uncommunicated reason should be treated. It did not say whether the inquiry concerned the actual reason, the reason communicated, a reason established later, the range of responses open to a reasonable employer, or the ECT’s own judgment of sufficiency. Nor did it identify the continuing role of the earlier IAC practice, the history of industrial relations, or the manpower policies which had influenced MOM. Legislative silence on those matters does not, without more, show failure, since open-textured standards are often deliberately left for judicial elaboration. The cumulative difficulty is that the ECT was asked to elaborate an old provision whose operative administrative tradition had never been published, while having regard to Guidelines intended to carry that tradition forward without any explanation of how their propositions ranked against the statutory text.

The unpublished corpus is the clearest sign of that incomplete bridge. The Minister said that MOM had accumulated a considerable body of cases, that those cases put broad principles into practice, and that they had not been published because “there was little need to do so”. Yet, the cases themselves—the facts found, reasons accepted or rejected, intermediate standards applied, and remedial choices made—were not brought into a publicly accessible decisional corpus. Administrative information could reach the ECT through the findings report, officer participation, assessors, and the ECT’s inquiry powers, although those case-specific means could not substitute for reasons capable of being tested, distinguished, reconciled, or developed. Continuity was consequently presented chiefly through abstracted Guidelines and particular materials; there was no accessible body of prior decisions which explained why one result followed rather than another.

This matters because broad principles may perform different work in administration and adjudication. An administrator may use a policy to organise recurring cases, provided the policy is lawful and genuinely permits exceptions. A court must first establish the legal rule which governs the parties’ rights and then determine the weight which a non-binding guideline may bear within it. As I explained under “Binding force” above, ss 20(7)(a) and 35 of the ECA assign the Guidelines an express legal place, and s 20(7)(a) requires the ECT to have regard to them, but those provisions do not confer upon their contents the force of primary or subsidiary legislation. During the 2018 Debates, Assoc Prof Daniel Goh asked directly about the legal status of the Guidelines, whether the ECT had to remain within them, how it should take guidance from them, and whether it could depart in a particular case. The Minister reiterated that the ECT had to take their principles and parameters into account, but did not answer whether or why those contents could control the statutory inquiry (2018 Debates at 5.40pm (Assoc Prof Daniel Goh Pei Siong) and 7.03pm (Mrs Josephine Teo, Minister for Manpower)). Parliament therefore required consideration of the Guidelines while leaving their particular weight and relationship with the enacted words for judicial reconciliation. That flexible formulation may preserve judicial independence; it does not explain what should occur where a proposition in the Guidelines cannot be reconciled with s 14(2).

The consequences of the incomplete bridge appear in the earlier analyses. The Guidelines carry MOM’s abstracted conclusions without the interpretive reasoning needed to reconcile them with s 14(2) (see [62]⁠–⁠[118] above); the former practice of eliciting the employer’s explanation was replaced with selected burdens and active evidential powers which leave ordinary configurations unresolved (see [119]⁠–⁠[171] above); and the retained remedies were placed within a judicial compensation scheme which does not fully identify their purposes (see [185]⁠–⁠[213] above).

The consultation in 2018 also rejected a general duty upon employers to provide reasons, relying upon the contractual symmetry under which an employee may resign on notice without explanation (2018 Consultation Response at paras 23–25). That premise left the court to determine a reason-based statutory prohibition where the material explanation could remain solely with the employer. Section 27(2), TADM’s findings report, officer participation, adverse inferences, and ss 20–21 of the ECA supplied partial bridges, although they did not produce a complete account of the legal burden in an unexplained, mixed, or later-explained dismissal.

The remedial transfer was similarly substantive as well as institutional. The employee now elects her relief; monetary orders are capped and calculated under the Second Schedule; ECT orders are enforceable as District Court orders; and an appeal lies, with permission, on law or jurisdiction. Those features differ from the former Ministerial order and require the court to identify the interest which reinstatement or compensation vindicates. Yet, as explained at [185]⁠–⁠[213] above, the legislation supplies caps and factors without fully stating those purposes.

There is, moreover, an unresolved duality within the legislation itself. The 2018 Amendment Act transferred the former EA route while leaving s 35(3) of the IRA as a Ministerial representation for an eligible unionised employee. Sections 16(5) and 16(6) of the ECA (formerly ss 16(4A) and 16(4B)) treat an ECT claim and a s 35(3) representation as mutually exclusive avenues, although the same event of dismissal and the same formula “without just cause or excuse” open both. One route remains embedded in the administrative sequence described above; the other is adjudicated by a court under the ECA; and the express statutory direction to have regard to the Guidelines applies to the ECT and High Court, but does not apply to the Minister. As the discussion at [588]⁠–⁠[598] below also shows, Parliament restated and enlarged the EA definition of dismissal without making an equivalent textual amendment to the IRA. These choices may reflect an intention to preserve different routes for different constituencies. Parliament did not, however, say whether the decision-makers were expected to apply an identical legal test, whether industrial considerations could bear differently upon each, or how divergence between an unpublished Ministerial approach and published judicial law should be resolved. Their coexistence makes particularly plain that institutional identity was not an inevitable incident of the words and that the transfer required an account of which incidents travelled with the protection.

Training and specialisation contribute to the transfer through s 9 of the ECA, which requires legal qualification, and s 18(2), which permits assistance from an assessor with relevant experience, thereby giving a judge-led employment tribunal the benefit of workplace expertise. That expertise must nevertheless be expressed through the governing sources, evidence, reasons, and statutory power. It cannot reconcile the Guidelines with s 14(2) or supply an unexpressed remedial purpose unless the law identifies how it bears upon the decision.

On balance, I think the public record supports a narrower inference. Parliament plainly recognised the ECT as a court and made substantial procedural and informational adaptations for judicial adjudication. Its attention appears, however, to have been directed principally towards continuity, access, consolidation, and procedure. The materials do not show lawmakers confronting in full how MOM’s broad principles were to be reconciled with statutory text, judicial interpretation, precedent, reasons, and appellate supervision, or how far the former scheme’s apparent workability depended upon its institutional premises. The rationale of a one-stop service, and the use of a tribunal designed principally for salary claims, show how the reform was conceived. The absence of an accessible Ministerial corpus, the uncertain relationship between the Guidelines and s 14(2), the partial burden provisions, the under-specified remedial scheme, and the continuing parallel Ministerial route disclose what remained unresolved.

The comparison may therefore be stated shortly. The Ministerial arrangement applied the standard through an industrial-relations administration which elicited reasons, sought conciliation, and drew upon specialist experience and lawful policy, whereas the ECT applies the same words as a court and must construe the statute, determine facts under legally assigned burdens, give reasons, and exercise remedies through purposes capable of consistent application. Although the 2018 legislation created several procedural links between the institutions, it left the substantive and jurisprudential translation incomplete. The uncertain relationship between the Tripartite Guidelines and s 14(2), including the departures already identified, made the intended bridge unstable and completes the fourth failure.

Pulling the four failures together

The four failures arose at different stages of the history, although they form a sequence in which each made the next possible. Their combined effect was to leave a statutory merits protection overlaid with contractual incidents and motive-based categories, while its restorative remedies came to be influenced by notice-limited reasoning. Before reconnecting those matters to the major troubles identified at [57]⁠–⁠[241] above, I draw the sequence together.

The first is apparent from the history beginning at [253]⁠ above. The protection emerged from an industrial jurisdiction which examined the employer’s reason and could reverse a dismissal notwithstanding contractual notice. When Parliament removed dismissal from the IAC in 1968, it transferred and reconfigured that inquiry through Ministerial machinery; when it extended materially the same protection to non-unionised employees in 1973, it did so because notice pay did not deter unjust dismissal. The reform in 2018 did not recover that settlement, and the retained words were therefore left to appear as a generality upon which a more familiar account could be placed.

That inheritance was then obscured as the prevailing conception of industrial justice changed. For sound economic reasons, industrial justice came increasingly to be expressed through viable firms, adaptable wages, skills, redeployment, matching, and employability across jobs (see [352]⁠–⁠[391] above). Yet, s 14(2) was left unamended and the relationship among labour-market mobility, the retained general standard, and reinstatement was not explained. The inherited law and prevailing policy therefore came to point in different directions without either having been expressly abandoned.

The common label supplied the means by which those different conceptions could be combined. The common law label carried a notice-centred account of lawful termination, cause, damages, and specific enforcement, while the later manpower conception supplied categories of legitimate business grounds and identified abuses. Once both came to travel under “wrongful dismissal”, the Guidelines could place them within one practical account even though the statutory protection derived neither its force nor its content from either conception. As explained at [392]⁠–⁠[404] above, notice, communicated reasons, categories, process, and remedy consequently acquired several inconsistent functions without any rule determining their priority.

Finally, the transfer to the ECT exposed that combination to the demands of adjudication. The Ministerial process could elicit an explanation, seek compromise, and make an industrial judgment without publishing a doctrine for each intermediate step. The ECT, once called upon to adjudicate, must state the governing rule, assign burdens, find facts, explain liability, and exercise remedies through legally intelligible criteria. The procedural links which the scheme created between the institutions in 2018 did not complete the substantive translation of unpublished administrative practice and the Guidelines into propositions which a court could lawfully apply (see [405]⁠–⁠[439] above).

Taken together, these stages explain how the inherited character went unrecognised, a later conception of work and security came to appear natural, the shared label permitted that conception and common law incidents to enter the statutory protection, and the resulting hybrid was placed before a court without a coherent hierarchy of sources, elements, burdens, and remedies. Each stage could have been addressed expressly: by recovering the inherited settlement; amending the statutory standard; adopting distinct terminology; or enacting a judicially complete scheme. Parliament adopted elements of the last course, although not enough to bring order to the whole.

The history thus explains the trouble concerning juridical foundation and content. Section 14(2) remains the binding source; the Guidelines must receive the regard directed by the ECA; and the common law continues to govern any separate contractual right. Yet, the practical prominence which the Guidelines were expected to assume has allowed their altered formula, notice presumption, motive requirement, categories, and treatment of reasons and process to resemble the legal test. Those circumstances may bear evidentially upon the inquiry under s 14(2), but they cannot replace the inquiry which the enacted words require.

The difficulty concerning proof follows from the same incomplete movement. The inherited process elicited the employer’s ground; the later conception accepted silence after contractual termination; and s 27(2), the findings report, officer participation, adverse inferences, and the ECT’s active powers supply only partial bridges between the two. As I explained at [119]⁠–⁠[171] above, those arrangements permit principled adjudication in an individual case, although the enacted scheme leaves several ordinary configurations unresolved.

The remedies likewise draw upon several conceptions, for reinstatement and back wages express restoration of the protected relationship, contractual damages express notice-limited performance, and later manpower policy favours compensation and movement into other work. The EC Regs divide compensation into income and harm without fully stating the interest represented, the relevant counterfactual, or the relationship among those ideas. The remedial uncertainty identified at [185]⁠–⁠[222] above is, accordingly, connected to uncertainty concerning the wrong which the remedy is intended to answer.

The resulting instability may operate against either party: contractual compliance can coexist with statutory liability; an employee may be unable to prove a hidden wrongful motive despite evidence that no sufficient justification existed; a true allegation may be treated as decisive without examining its causal role or gravity; and a false allegation may be treated as conclusive despite another operative ground. A claimant who succeeds may then receive no relief, or compensation assessed by a measure whose connection with the wrong remains uncertain. Those difficulties arise even before one reaches the hard facts at the margins, because the available premises themselves can direct the inquiry towards opposing outcomes.

The ECT must therefore keep the IAC awards, contemporary policy, the Guidelines, and industrial fairness within the juridical place which the legislation gives to each. The difficulty with s 14(2) is that an inherited industrial protection survived after its history and animating conception had faded, acquired notice-centred and administrative incidents under a shared label, and entered judicial adjudication without a complete translation. I therefore turn to reconstructing the law in the order which a judicial decision requires: beginning with the enacted provisions, reading them in their statutory and historical setting, assigning the Guidelines their lawful place, and deriving a coherent account of the wrong, its proof, and its consequences.

Reconstructing the law in its best light

The preceding analysis has identified the juridical, evidential, and remedial troubles attending a claim under s 14(2), and traced how an inherited industrial protection came to be administered through a framework which draws upon contractual, administrative, and judicial conceptions of dismissal. I now bring those conclusions together and state the present law in the form which, so far as the enacted words allow, best reconciles its history, modern institutional setting, and practical operation.

I begin with the juridical foundations of s 14(2) and its relationship with contract, before turning to proof and remedy. Some of the remaining uncertainties arise from legislative choices or omissions which only Parliament can resolve conclusively. Subject to them, however, I think the account which follows identifies the questions which the ECT should decide and the order in which it should decide them.

The juridical foundations of section 14(2)

As I concluded at [223]⁠–⁠[241] above, s 14(2) creates a statutory relation between employee and employer: the employee has a claim that the employer refrain from dismissing her without just cause or excuse, and the employer bears the correlative duty. The employer nevertheless retains the legal power to bring the relationship to an end, so that a dismissal may be legally effective even though its exercise contravenes that statutory duty.

Section 14(2) regulates, in that sense, the basis upon which the employer exercises the power to terminate. Where the employer does so without the justification which the statute requires, the ECA gives the employee the means to seek reinstatement or compensation, notwithstanding that the dismissal has already taken effect.

It follows that the statutory duty exists before any claim is lodged. The claim route, together with the rules which govern proof and remedy, operates later by supplying the means through which that pre-existing duty is enforced and the consequences of its breach are addressed.

The content of the inquiry follows from the analysis at [90]⁠–⁠[118] above. The ECT must identify the cause or excuse asserted, determine whether the material facts existed and bore upon the dismissal, and then decide whether that basis was sufficiently just in the employment and industrial setting. Proof that an adverse fact concerning the employee existed answers only part of the inquiry, because the fact must also have operated upon the dismissal and carried sufficient weight to justify it. Correspondingly, the employee may establish the absence of just cause or excuse without proving discrimination, retaliation, deprivation of a benefit, or some other positively improper purpose.

The Guidelines occupy the subordinate place explained at [62]⁠–⁠[87] above. The ECT must genuinely consider their principles and illustrations, whose legal significance derives from the direction to have regard to them and which may illuminate the statutory standard and assist its application. Where a proposition cannot be reconciled with s 14(2), however, the statutory provision must prevail.

The relationship between section 14(2) and contract

As the analysis at [63]⁠–⁠[65] and [392]⁠–⁠[404] above has shown, the shared label “wrongful dismissal” cannot be allowed to collapse the contractual and statutory protections. Contract, read where applicable with ss 10 and 11, ordinarily determines the notice required, the availability and amount of salary in lieu, and the contractual payments due upon termination. Section 14(2) asks the further and distinct question whether the employer possessed just cause or excuse for dismissal. That distinction is confirmed by the definition of “dismiss” in s 2(1), which expressly includes termination with notice and for cause or otherwise.

Contract remains relevant to the statutory inquiry, although for different reasons. It may identify the employee’s duties, the governing standard of performance or conduct, an agreed disciplinary process, and the notice position which bears upon proof, related claims, or double recovery. A contractual breach may therefore form part of the cause relied upon. Its existence does not, however, dispense with the statutory inquiry into whether it operated upon the dismissal and carried sufficient gravity to justify that response.

The juridical foundations have been practically altered

That juridical account must be applied within the scheme which now governs proof and adjudication. The former Ministerial machinery elicited the employer’s ground, attempted conciliation, and examined its industrial justification. Under the present scheme, mandatory mediation precedes a judicial claim in which no general reason-giving duty exists and s 27(2) reallocates the legal burden only in specified circumstances (see [119]⁠–⁠[171] above). An employee in a residual case may therefore have to establish the absence of just cause or excuse without knowing which possible explanation she must meet.

The Guidelines have, moreover, shaped the practical routes through which such a claim is ordinarily presented, by organising dismissals around notice, communicated reasons, and stated categories. Those routes may assist the inquiry, although, for the reasons given at [62]⁠–⁠[118] and [392]⁠–⁠[404] above, they must remain within the limits of the statutory test.

Section 27(2) differentiates the inquiry further, since para (a) places upon the employer the burden of proving that a qualifying dismissal without notice was with just cause or excuse, while para (b) places upon it the burden of proving the poor-performance or misconduct ground for which notice is or purports to be given. In a residual case, the employee bears the legal burden, subject to proof by circumstantial evidence, the ordinary movement of an evidential burden, and the confined operation of s 108. The form of termination and the reason attached to it therefore affect the risk of non-persuasion, although the statutory question to which the evidence is directed remains the same.

The modern industrial setting also bears upon the evaluative content of that question. As I explained at [352]⁠–⁠[391] above, present policy attaches greater weight to viable enterprises, flexibility, mobility, skills, and employability across jobs. Those considerations may inform whether a business, operational, performance, or conduct reason is sufficiently just in its setting. The enacted standard itself remains unchanged, however, and the ECT must articulate their application through the judicial reasons required for the purposes described at [415]⁠–⁠[424] above.

The practical alterations can be rendered coherent

In my view, the practical alterations can be made substantially coherent if four questions are kept in their proper order. The ECT should ask, first, whether the employee was dismissed within the applicable statutory meaning; second, what basis the employer says constituted cause or excuse; third, whether the facts comprising that basis existed and operated, at least materially, upon the decision to dismiss; and fourth, whether the proved and operative basis carried sufficient weight, in the employment and industrial setting, to make dismissal a just response. The party bearing the legal burden determines how an unresolved doubt is answered, while the questions to which the evidence is directed remain unchanged.

Section 14(2) imposes no general requirement that the employer communicate its reason at the time of dismissal. A contemporaneous explanation may, however, identify what actually caused the decision, define a factual case which can be tested, and expose inconsistency if a materially different account emerges only in the proceedings. The role of later evidence is confined to proving that a cause or excuse existed and operated at the material time; a historical cause which did not then exist cannot be created after the event.

The truth or falsity of a communicated reason therefore bears the evidential significance explained at [107]⁠–⁠[110] above. Falsity may support an inference of concealment and aggravate the harm caused, while proof of truth establishes the fact asserted but leaves its operation upon the dismissal and its justificatory weight to be decided.

The categories used in the Guidelines serve a similar illustrative function. Discrimination, retaliation, or deliberate deprivation of a benefit will ordinarily point strongly towards liability, while misconduct, poor performance, redundancy, or genuine operational requirements are capable of supplying justification. Whichever category is invoked, however, the ECT must still determine the facts, their causal operation, and their gravity in the particular case.

In evaluating that gravity, the ECT should allow the employer a real measure of latitude in assessing operational needs, organisational fit, performance, and workplace conduct. That latitude depends upon proof that the material facts existed, played a part in the decision, and possessed sufficient weight to justify termination. Relevant considerations may include the nature of the employment, the employee’s duties and length of service, the importance of the standard engaged, the seriousness and persistence of the conduct or deficiency, the employer’s operational needs, consistency of treatment, and the alternatives realistically available. I mention those considerations as a non-exhaustive group whose relevance lies in what they reveal about whether dismissal was just within the particular employment setting.

Process ordinarily bears upon the inquiry through the evidence which it produces and the reliability which it lends to the asserted cause. I address the fuller uncertainties concerning due inquiry and contractual process from [545]⁠ below. For present purposes, a warning, investigation, or opportunity to answer may strengthen the employer’s account, while its absence may weaken an allegation which was contestable or readily answered. JGP gives due inquiry substantive work where misconduct is relied upon despite notice or salary in lieu, although that first-instance holding has not been resolved on appeal. Its general function under s 14(2) is therefore evidential; any substantive requirement depends upon an independently applicable statutory or contractual obligation.

Where the employee bears the legal burden, she may establish the statutory negative through circumstantial evidence in the manner described at [138]⁠–⁠[169] above. A satisfactory record, recent confirmation or reward, continued deployment, the absence of contemporaneous concern, inconsistent explanations, or withheld evidence may cumulatively call for an answer. An evidential burden may then fall upon the employer while the legal burden remains with the employee. Section 108, if applicable, may allocate an identified fact especially within the employer’s knowledge, while an adverse inference may assist in deciding what missing evidence would have shown. Each performs that confined work within the existing allocation of the composite burden.

The resulting inquiry is therefore whether a factually established and operative cause or excuse was sufficiently just to warrant dismissal. Notice, communication, truth, category, and process bear evidentially in the ways just described, while s 27(2) determines the legal burden in the cases which it identifies. In a residual case, the employee may discharge her burden through the circumstantial record as a whole, assisted where appropriate by a moving evidential burden and the ECT’s statutory powers. In that way, the Guidelines perform meaningful work, the latitude required by contemporary employment relations is preserved, and both remain within the standard which Parliament enacted.

The remedial framework has been practically altered

A comparable practical alteration may be seen in the remedial framework, where reinstatement and wages upon restoration gave the older industrial conception its most direct remedial expression. Although s 14 continues to place reinstatement first and authorises it despite any contrary rule of law or agreement, the remedy has become exceedingly rare for the reasons given at [173]⁠–⁠[184] and [352]⁠–⁠[391] above, with compensation accordingly becoming the ordinary response.

That compensation now takes a structured judicial form under reg 17A and the Second Schedule, which divide it into loss of income and harm, prescribe ceilings, and identify aggravating and mitigating factors. The claimant must elect the remedy sought; any monetary order remains subject to the applicable claim limit; and the ECT must explain its exercise of discretion through reasons capable of examination within the restricted appellate route.

As I explained at [185]⁠–⁠[222] above, however, the translation remains incomplete. The legislation identifies neither the counterfactual for loss of income, the interest represented by the base amount for harm, nor the relationship between either component and reinstatement. It also leaves uncertainty concerning proof of non-pecuniary injury, the separation of the base from later adjustments, and the treatment of overlapping financial and reputational consequences. The court must address those omissions within the remedial scheme which Parliament has enacted.

The practical alterations can also be rendered coherent

Even so, the remedies can be placed in a coherent order. Reinstatement is the restorative remedy and should be considered by asking whether the protected relationship can sensibly and fairly be restored. Relevant matters include the employee’s wish to return, the continued existence of the position or a materially equivalent one, the parties’ practical ability to work together, the effect upon other employees and operations, the employee’s own conduct, and whether restoration would answer the wrong established. Those considerations derive their relevance from the statutory object of restoration. The common law’s aversion to specific performance cannot supply a presumption against the very remedy which s 14(3) expressly authorises.

Compensation supplies the monetary alternative where restoration is not sought or should not be ordered, and, within it, loss of income addresses financial loss proved to have been caused by dismissal without just cause or excuse. The question is therefore factual: what income would the employee probably have earned had that statutory wrong not occurred, subject to mitigation, intervening events, and the three-month ceiling? If the evidence shows that a justified dismissal, resignation, expiry of a fixed term, redundancy, or another event would probably have ended the income stream later, that prospect may be reflected in the award. The employer’s bare contractual ability to give the shortest notice cannot define the counterfactual, because that would assume away the additional statutory restraint and reduce the remedy to the contractual claim which s 14(2) was enacted to supplement.

Salary in lieu of notice should, accordingly, be treated as a separate entitlement and, where the same period is engaged, a credit which prevents double recovery, while setting neither a ceiling nor a presumptive measure for statutory loss. The employee must nevertheless prove that the income claimed was probably lost because of the wrongful dismissal and must take reasonable steps to mitigate it. Ordinary difficulties in obtaining alternative work may support an award where the evidence permits causal and quantitative findings, whereas speculation that any period of unemployment must have resulted from dismissal does not.

The harm component is best understood as monetary recognition of the non-income injury occasioned by the wrongful severance of the protected relationship. It may include the affront, disruption, personal consequences, and, where proved, reputational or physical effects which do not belong within the income calculation. The employee need not prove a recognised psychiatric injury before any award can be made, although the existence of the statutory wrong does not require the maximum base in every case. The ECT must select a base within the two-month ceiling which reflects the nature and seriousness of the harm established or properly inferred, and then apply the aggravating and mitigating factors without using the same circumstance twice.

The factors listed in paras 4, 6, and 7 of the Second Schedule should be applied according to their function. The manner of dismissal, a false accusation, physical injury, or deliberate impairment of later employment may aggravate the employer’s wrong and its consequences. Misconduct, poor performance, or insubordination which contributed to the events but fell short of just cause or excuse may mitigate the employee’s claim to the full conventional response. Where a fact has already determined the base, particular care is required before it is used again as an adjustment; and where a reputational consequence has produced measurable income loss, the ECT should ensure that the same loss is not compensated again under harm.

Finally, the word “may” in s 14(3) preserves a remedial discretion which must be exercised judicially and in furtherance of the statutory remedies. A successful claim does not create an automatic entitlement to the employee’s preferred order, and a remedy may be reduced or refused where the conditions for its proper exercise are not established. The discretion should not, however, be used to nullify an established right without explanation. The ECT should identify the wrong found, the remedial object engaged, the facts bearing upon that object, and the reasons why the order made—or declined—follows from them.

I fully recognise that this reconstruction leaves seams in the law. The ECT must continue to apply an industrially evaluative standard within a modern judicial and manpower setting, using evidential and remedial provisions which do not address every ordinary case. Even so, the account provides a workable order: s 14(2) protects against dismissal unsupported by a factually established, operative, and sufficiently just cause or excuse; the surrounding rules govern proof and remedy; contract performs its separate work; and the Guidelines inform the statutory inquiry. On that basis, I turn to the remaining troubles and, ultimately, Mr C’s claim.

Addressing a miscellany of lesser troubles

As I foreshadowed at [49] above, the reconstruction just proposed does not exhaust the difficulties in the surrounding scheme. I have sought through that reconstruction to minimise the major troubles and to state the law coherently in the light of the contemporary employment relationship. Several lesser troubles nevertheless remain and, for the purposes of analysis, I have arranged them in the following five groups:

(a) The first concerns institutional drift—that is, the drifting of dismissal claims out of their proper forum.

(b) The second concerns institutional fragmentation—that is, the splitting of the statutory and contractual claims arising from the same dismissal between different adjudicative routes.

(c) The third concerns statutory overlap—that is, how ss 14(2) and 84(2) of the EA relate to ss 35(2), 35(3) and 82 of the IRA.

(d) The fourth concerns process before dismissal—that is, the legal significance, if any, of the process followed by an employer before dismissing an employee.

(e) The fifth concerns the meaning of “dismiss”—that is, the reach of the definition in s 2(1) of the EA, including its application beyond s 14(2) and its relationship with forced resignation and constructive dismissal.

I take each of these groups in turn.

Institutional drift of the statutory actions under the EA

The first lesser trouble concerns the institution authorised to hear claims brought under ss 14(2) and 84(2). Each provision directs the employee to lodge a claim under s 13 of the ECA and makes that route subject to the mandatory mediation process in s 3 of the ECA. The question is whether this machinery supplies the exclusive route for the statutory claim or merely an additional forum alongside the ordinary civil courts.

To answer this question, I will, in this section, examine: (a) the general statutory scheme supporting claims under ss 14(2) and 84(2); (b) three decisions of the District Court, the Magistrates’ Court, and the High Court Registry which bear on this question; (c) the policy arguments for and against construing the ECT as having exclusive jurisdiction over ss 14(2) and 84(2) claims; and (d) the power of transfer conferred by s 17 of the ECA.

Ultimately, my view is that the rights created by ss 14(2) and 84(2) and claim-route expressly prescribed by those provisions form a single, inseparable scheme. On this view, only the ECT has jurisdiction to adjudicate ss 14(2) and 84(2) claims within the limits prescribed by the ECA and EC Regs. The ordinary civil courts are not seised of original jurisdiction over these two statutory claims save in situations where the power of transfer under s 17 of the ECA is exercised. Even then, upon such transfer, those courts are bound by the same limits prescribed by the ECA and EC Regs—most significantly, the awards which may be made in respect of those claims may not exceed the sum the ECT would have been able to award—$20,000 or $30,000 (see s 12(7)(b) of the ECA read with regs 17(1)(b) and 17(2)(a) of the EC Regs)—had the matter not been transferred out of the ECT.

The general statutory scheme

I begin with the words of the provisions. Section 14(2) has been set out at [91] above and I do not reproduce it again. Sections 84(1) and (2) state:

Reading s 14(2), one will see that it describes a protection against dismissal without just cause or excuse and, in the same sentence, prescribes the means by which that protection is to be enforced. It identifies the employee who may complain, the event of which she may complain, the two remedies which she may seek, and the claim-route by which she may seek them. Section 84(2) does the same for a female employee in the circumstances identified in ss 84(1)(a), (b), or (c) who considers that the notice of dismissal given to her was not given for sufficient cause. In each case, the employee “may lodge a claim, under section 13 of [the ECA]”, and may do so for reinstatement or compensation. The protection, remedies, and claim-route are thus expressed as parts of one statutory arrangement.

The word “may” does not, in my view, allow those parts to be separated. Of course, it leaves the employee free to decide whether to pursue the statutory claim at all. However, if an employee does decide to pursue the claim, the words “under section 13” plainly identify how that right is to be invoked. Had Parliament meant to confer on employees the right to bring claims under ss 14(2) and 84(2) either before the ECT or in the ordinary civil courts, it almost certainly would not have expressly specified only one of those two routes, while leaving the other entirely unstated.

The degree of specificity with which Parliament seems to have allocated the forum available for ss 14(2) and 84(2) claims is also readily apparent from the fact that both provisions are expressly made “subject to section 3” of the ECA. Before a claim may be lodged in the ECT, s 3 requires the underlying “wrongful dismissal dispute” to be submitted to mediation. If it is not settled, the claim referral framework in s 6 leads to the lodging of a claim under s 13, subject to the jurisdictional conditions in s 12. The ensuing proceeding is informal and judge-led, legal representation is excluded, the ECT must have regard to the Guidelines, and s 27 allocates particular burdens of proof. The available orders and monetary limits are likewise prescribed, and any appeal follows the restricted route in ss 23 to 26. These provisions prescribe the institutional framework by which the rights under ss 14(2) and 84(2) are to be enforced and, taken together, it is clear that they constitute the means by which Parliament made the statutory protection privately enforceable.

The remedial provisions in the EA point in the same way. Sections 14(3) and 84(3) confer upon the ECT the power to order the employer to reinstate the employee or pay compensation once the statutory criterion is satisfied. No corresponding provision authorises an ordinary civil court, upon an independently commenced action, to grant those remedies. This matters because breach of a statutory duty does not, without more, generate a private civil action; the question is whether Parliament intended to confer one and, if so, on what terms (see Tan Juay Pah v Kimly Construction Pte Ltd and others [2012] 2 SLR 549 at [54]). Where the statute expressly creates a cause of action, its scope has to be determined first and foremost by construing the statute as a whole (see Reed, Michael v Bellingham, Alex [2022] 2 SLR 1156 at [63]–[65]). Here, the intention to permit private enforcement is express, and that intention is accompanied by clear provision for the form of the claim, the remedies available, and the institution to which it is to be brought.

The only textual indication going in the other direction is s 16 of the ECA. Section 16(1)(a) prevents a claim relating to a “specified employment dispute” from being lodged in the ECT where proceedings “relating to that claim” are pending in, or have been heard and determined by, another court or the IAC. Section 16(2) then prevents proceedings relating to an ECT claim from being commenced elsewhere after it has been lodged, subject to stated exceptions. However, these provisions only contemplate that related proceedings may exist in another court. Nothing about them suggests that they were intended to confer jurisdiction to the ordinary civil courts, or create a cause of action which no other provision supplies. As I see it, their function is simply to prevent duplication across the different contractual and statutory disputes brought within the ECA, many of which may arise alongside independently enforceable civil rights. A rule governing priority between proceedings does not thereby create a parallel civil form of every claim to which it applies.

For completeness, ss 122 and 132 of the EA do not, in my view, bear on what I have just said. Section 122 provides only that nothing in Part 15 limits or affects the jurisdiction of any court. It thus preserves court jurisdiction from the claims and inquiry machinery in that Part; it cannot be sensibly read as creating a cause of action under provisions found elsewhere in the EA or displacing the specific route prescribed by ss 14(2) and 84(2). Section 132 is more specific. In the circumstances which it identifies, it preserves the civil rights and remedies of an employer or employee for breach or non-performance of a contract of service. However, as its premise is contractual—and, as I have explained, the rights conferred by ss 14(2) and 84(2) are decidedly not contractual in nature—s 132 cannot reasonably be said to have anything to do with the fora which may adjudicate claims under ss 14(2) and 84(2).

In sum, the general statutory scheme makes the following clear. An employee may sue in an ordinary civil court for inadequate notice, unpaid contractual benefits, breach of another term of the employment contract, or an independently actionable tort. A pregnant employee may also have an enforceable payment entitlement under s 84(1)(b), which performs work distinct from the reinstatement-or-compensation claim in s 84(2). What the employee may not do is take the statutory complaint and remedies constituted by ss 14(2) or 84(2), detach them from the specified route and limits with which Parliament coupled them, and commence them as an ordinary civil action. It is with this in mind that the decisions discussed in the next section must be read.

The three decisions

The first decision is Longitude 101 Pte Ltd v Navinea Kanapathy Pillai and another matter [2025] SLR(StC) 403 (“Longitude 101”). The employee’s counterclaim was brought under s 84(1)(b) (at [16]). The District Court identified the issue as whether the notice of dismissal had been given “without sufficient cause” such that it could not deprive her of the payments protected by that subsection (at [114]). The court was not asked to order reinstatement or compensation under s 84(2). However, in deciding the sufficient-cause question, the court considered the Guidelines and reasoned that declining to do so would produce different rules for the “same wrong” in the civil courts and ECT and encourage forum-shopping (at [117(c)]). It then held that reliance upon contractual notice or salary in lieu of notice did not, without more, establish sufficient cause (at [118(a)]).

Reading Longitude 101 very precisely (cf my reading of the case in JGP at [A.109]–[A.115]), it can be said that—given the specific statutory basis on which the counterclaim was pleaded (s 84(1)(b) of the EA)—the court did not strictly allow the action under s 84(2) to drift out of the ECT into the ordinary civil courts. Such a reading is arguably fairer than that which I adopted in JGP. After all, s 84 attaches more than one consequence to a notice of dismissal given without sufficient cause. Section 84(1) protects specified maternity-related payments, whereas s 84(2) creates a claim for reinstatement or compensation and directs that claim to the ECT. Though both consequences depend upon the same inquiry into the sufficiency of cause for which the notice of dismissal was given, that does not mean they are the same claim. This, in turn, may be used to explain why the court in Longitude 101 thought that the Guidelines were relevant to the s 84(1)(b) inquiry even though the Guidelines are—in terms—more properly connected with a claim under s 84(2).

All that being said, Longitude 101 itself did not state the distinction in the terms I just did. Its description of the two inquiries as concerning the “same wrong”—coupled with its concern about forum-shopping—risks obscuring the fact that the words “sufficient cause” relates to two legally distinct claims—one arising from s 84(1)(b) and another from s 84(2). That risk, in my view, can be said to have manifested in Cisilia Oktavia Lim v Reins International (Singapore) Pte Ltd [2024] SGMC 68 (“Cisilia”).

In Cisilia, the employee had initially pleaded a contractual claim for wrongful dismissal (at [9] and [24]). However, by the trial and closing submissions, her case appeared also to invoke s 14(2) and the Guidelines (at [25]). The Magistrates’ Court recognised that s 14(2) expressly directed the claim to the ECT and, quite correctly, observed that it was “not immediately clear” whether that provision and the Guidelines applied to the civil dispute before it (at [27]). Despite that doubt, the court proceeded on the basis that an action could be brought in the Magistrates’ Court according to principles similar to those governing a s 14(2) claim. It did so in reliance upon: (a) Longitude 101; (b) what it regarded as the “strange” difference between the monetary limits of the ECT and Magistrates’ Court; and (c) the parties’ common position (at [28]–[30]). In the result, the court found that the employee’s poor performance had been substantiated and, on that basis, dismissed her claim.

None of the three matters relied on by the court, however, seems to answer the jurisdictional question. First, despite the statements in Longitude 101 which may have obscured the point (see [495]–[497] above), the case—as I just explained—was one concerned with a claim under s 84(1)(b), not s 84(2). Thus, relying on Longitude 101 as support for the proposition that a claim under s 14(2) may be prosecuted in and determined by the Magistrates’ Court was two steps removed from what the case actually determined. Second, the difference in monetary limits would only be strange if it were first established that the same statutory claim was available in both the ECT and the civil courts. If, as I think, the ECT limit forms part of the conditions attached to the claims under ss 14(2) and 84(2), what Cisilia described as strange would simply be part of the statutory design. Finally, the parties’ common position could neither create the statutory cause of action nor answer whether that claim was independently invocable in the Magistrates’ Court.

I do not say any of this to cast doubt upon the factual conclusion reached in Cisilia. The employee’s contractual claim—as opposed to her claim that she had been dismissed “without just cause or excuse”—may well also have failed on the court’s findings. My point is confined only to the issue of which institution is authorised to adjudicate claims arising from ss 14(2) and 84(2) of the EA. Having identified the uncertainty arising from s 14(2), the Magistrates’ Court—in my view—erroneously proceeded on, amongst other things, the parties’ common position without first deciding whether the legislation allowed that claim to be brought before it at all.

In a way, both Longitude 101 and Cisilia are reflective of the kinds of problems that can be produced by polysemes. In both cases, the courts seemed to have reached the views they did, at least in part, because the label of “wrongful dismissal” was attached to both the breach of contract actions with which the ordinary civil courts are very familiar, and the statutory actions arising from ss 14(2) and 84(2) coupled with the Guidelines. As noted in Polysemy and the Law, familiar legal labels can provide decision-makers with useful heuristics which reduce the work required to decide a case, but the associations carried by those labels may also misdirect where their meanings are related yet distinct; and once a label is chosen, it may itself influence the cases in which the labelled concept is later applied (at 1084–1085, 1103 and 1116–1117). That, in my view, helps—at least in part—to explain what occurred here. The common label made the connection between the contractual and statutory actions easy to grasp, but also made it easier for the attributes of one to be carried across to the other. However, the shared label did not mean that the two actions had the same juridical source, content, remedies, or claim-route. Those matters had to be answered by construing the provisions which created the statutory actions and, in Cisilia, it was that anterior inquiry which remained unanswered.

This brings me to the third case, which only arose after I gave judgment in this case on 26 January 2026: Daniel Seng Hock Chye v Denso International Asia Pte Ltd [2026] SGHCR 14 (“Daniel Seng”). To be absolutely clear, Daniel Seng says nothing which bears on the decision I made in the instant case. I address it only because it considers both Longitude 101 and Cisilia and, in doing so, correctly identifies an important limit on the use of the Guidelines in ordinary civil litigation. Although I agree with that part of its reasoning, its treatment of Cisilia, in my view, requires some qualification.

Daniel Seng concerned an employee whose employment had been terminated after approximately 19 years and four months of service. He was paid salary in lieu of notice, but commenced a civil action in the High Court alleging, amongst other things, that he had been wrongfully dismissed; that the termination had been carried out in bad faith as a disguised retrenchment to deprive him of retrenchment benefits; that the termination was contrary to the Guidelines; and that he had suffered psychiatric harm, distress, humiliation, and loss of reputation as a result (at [1]–[4] and [12]–[13]). The employer applied to strike out the statement of claim. The court allowed the application, holding that the claim disclosed no reasonable cause of action and was, in any event, legally and factually unsustainable.

The starting point of the court’s reasoning was the absence of any pleaded legal basis for the employee’s claim. Though he had described the termination as a “wrongful dismissal” and a breach of duties owed to him “without just cause or excuse”, he had not alleged a breach of contract, any statutory provision, or any identified contractual, tortious, or statutory duty (at [14]–[17]). The court observed that the words “without just cause or excuse” appeared to have been borrowed from s 14(2), but that the employee had not pleaded a claim under that or any other statutory provision. It then added that the EA framework “operates through the [ECT], not the civil courts” (at [17]).

When the court dealt with the employee’s reliance on the Guidelines, it held that they did not, by themselves, create a civil cause of action (at [48]–[53], [56], and [103]). That conclusion is sound because, whatever significance the Guidelines acquire because the ECA requires the ECT to have regard to them, they do not purport to confer an independent private right or specify any civil remedy for their contravention. Their relevance must therefore come from some legal right or cause of action which exists independently of the Guidelines; they cannot supply that right or cause of action by their own force.

The court’s treatment of Longitude 101 can also fairly be read consistently with the narrower reading I have adopted at [496] above (as opposed to that in JGP). At [54], Daniel Seng accurately identified the claim in Longitude 101 as one brought under s 84(1)(b). When it then emphasised that the Guidelines had been considered “in the context of applying a statutory provision”, the natural referent was that same provision, rather than s 84(2). On this reading, Daniel Seng did not treat Longitude 101 as a case in which a claim under s 84(2) had been brought in the ordinary civil courts. It relied on the case for the more limited proposition that the Guidelines had been used to assist in determining a right which s 84(1)(b) itself conferred, instead of supplying the juridical source of a freestanding action.

The difficulty lies instead in its treatment of Cisilia. At [55], Daniel Seng described that decision as “[e]ven more telling” because the Magistrates’ Court had expressly questioned whether s 14(2) and the Guidelines applied outside the ECT. That description is accurate so far as it concerns the doubt expressed at [27] of Cisilia. However, the court in Cisilia did not stop there. As explained at [498]⁠–⁠[500] above, it went on at [28]–[30] to proceed on the basis that principles similar to those governing a s 14(2) claim could be applied in the Magistrates’ Court, relying upon Longitude 101, the difference between the monetary limits of the two institutions, and the parties’ common position. Its expression of doubt therefore did not resolve the institutional question. It preceded the very assumption which made the decision problematic.

It is, of course, true—as Daniel Seng stated at [56]—that neither Longitude 101 nor Cisilia held that the Guidelines “alone” created a standalone civil cause of action. However, that proposition does not answer the criticism of Cisilia. The difficulty there was that the court appeared to treat s 14(2), or principles derived from it, as applicable in an ordinary civil action without first determining whether the statutory claim could lawfully be brought in that forum. The source and character of the action were never satisfactorily identified. Thus, the doubt expressed at [27] could not justify the course taken at [28]–[30] of Cisilia. Insofar as Daniel Seng relied upon Cisilia as supporting the use of the Guidelines only within their “proper statutory framework”, that reliance was not entirely well placed, because the proper framework was precisely the question which Cisilia left unanswered.

The three decisions can therefore be brought together in this way. Read fairly and precisely, Longitude 101 concerned the statutory payment protection in s 84(1)(b), and did not decide that a claim under s 84(2) could originate in the civil courts. Cisilia brought s 14(2) directly into view, but proceeded upon an assumption after identifying the jurisdictional difficulty and supplied no sound basis for that assumption. Daniel Seng correctly held that the Guidelines cannot independently create a civil cause of action and contains observations which support the view that the EA framework operates through the ECT. Its discussion of Cisilia does not, however, cure the difficulty in that decision. Accordingly, none of the three cases supplies a reason to depart from the conclusion which the general statutory scheme otherwise requires.

The competing policy considerations

At first sight, some support for that view may also be found in the legislative history of the ECA. When the ECT was established in 2016, the Minister for Manpower explained that all employees would continue to have access to the civil courts, and that such recourse was especially important for complex claims requiring legal representation and more time to resolve (Singapore Parliamentary Debates, Official Report (16 August 2016) vol 94 (Mr Lim Swee Say, Minister for Manpower)). However, those remarks concerned the salary-related disputes within the original ECA. Many of those disputes rested upon contractual or independently enforceable statutory entitlements which were already cognisable in the civil courts. When Parliament later transferred the adjudication of claims under ss 14(2) and 84(2) from the Minister to the ECT, it simply placed those claims within the existing institutional framework without any indication—either in the ECA or the debates—that claims under those provisions were also actionable in the civil courts.

It must also be said that construing the ECT route as exclusive does not mean that every dispute arising from a dismissal must be heard there. There are many kinds of dismissal disputes heard outside the ECT. For example: (a) an employee may bring an ordinary civil claim for inadequate notice, unpaid contractual benefits, breach of some other contractual obligation, or an independently actionable tort; (b) a pregnant employee may—as in Longitude 101—rely upon the payment protection in s 84(1)(b); and (c) an eligible employee may instead pursue the Ministerial route under s 35(3) of the IRA through the trade union of which she is a member, subject to ss 16(3)–(6) of the ECA. Though these routes may arise from the same dismissal, they remain different legal actions, each with its own source, conditions, procedure, and remedies. The exclusivity with which I am presently concerned is confined to the claims for reinstatement or compensation which ss 14(2) and 84(2) themselves create.

These alternative routes answer much of the practical objection stated at [510] above and, once the focus returns to the same statutory claim, the argument in favour of exclusivity seems to me considerably stronger. Parliament coupled the reasons-based protection against dismissal with mandatory mediation, an informal and judge-led hearing, restrictions on legal representation, limited rights of appeal, prescribed remedies, and capped monetary exposure. If the same claim could instead be commenced in an ordinary civil court, each of those features would become optional at the employee’s election. Most significantly, the $20,000 or $30,000 limit would cease to define the maximum monetary liability attached to the statutory claim and would become merely a limit accepted by employees who elect to proceed in the simpler, faster forum. That, as I suggested, would be an odd scheme for Parliament to have created by provisions which expressly identify the ECT route, but say nothing at all about a concurrent and potentially uncapped civil action.

In that connection, I would add that the general purpose of protecting employees from unjustified dismissals cannot supply a sufficient reason to take the contrary view. Such purpose does not itself answer which forum Parliament selected or which procedure, remedies, and limits govern the claim. Those matters must be answered by examining how Parliament gave effect to the purpose (see Bennion at 399, 405–406, 414–416, 419–420, and 440–443). Here, Parliament did so by creating a claim for reinstatement or compensation, making it subject to mediation, directing it to the ECT, and placing it within prescribed procedural and monetary limits. An uncapped civil action would remove those conditions from the right which Parliament created and, in my view, produce a materially different statutory right.

Taking these considerations together, I do not think the policy arguments justify treating the ordinary civil courts as possessing concurrent original jurisdiction. I accept that exclusivity may require different claims arising from the same dismissal to be pursued in different institutions, and may leave an employee unable to recover every loss she attributes to the infringement of ss 14(2) or 84(2). However, allowing the same statutory claim to be brought elsewhere would detach it from the process, remedies, and monetary limits which define its operation under the ECA. In any event, the ECA contains a narrower mechanism by which connected and sufficiently complex matters may be combined and heard together in an ordinary civil court. That is the power of transfer conferred by s 17 of the ECA, to which I now turn.

The power of transfer under section 17 of the ECA

Section 17(1) provides that, where it appears to an “appropriate court”, on the application of a party to any proceedings before the ECT, that there is sufficient reason for those proceedings or a counterclaim in them to be dealt with by that court, the court may order their transfer to itself. For this purpose, the “appropriate court” may be the High Court, a District Court, or a Magistrates’ Court (s 17(5)). The sequence prescribed by these words is important: there must first be proceedings before the ECT, and it is a party to those proceedings who may apply for the transfer. Section 17 does not provide an alternative means of commencing the statutory claim in a civil court. It provides the means by which that court may become seised of an existing ECT claim where sufficient reason for the transfer has been shown.

A further feature of Rida Global is material to the question of what happens to the ECT’s monetary limit after transfer. The employee had originally sought $35,000, but abandoned $15,000 so that his claim would comply with the applicable limit of $20,000 (at [6]). Section 15(3) of the ECA expressly provides that a claimant who abandons an excess amount cannot recover it “in a tribunal or any other court”. Accordingly, whatever other effect the transfer may have had, it could not revive the abandoned $15,000. That said, the judgment in Rida Global contains no express decision on whether the receiving court was otherwise bound by the ECT limit and, relatedly, whether it would have been possible for the employee to amend his claim—following its transfer to the High Court—to seek a sum exceeding that limit.

On this broader question, neither s 17 nor any decision of which I am aware provides an express answer. Indeed, there is some textual basis for hesitation because s 12(7)(b), which prescribes the claim limit, and s 22(5), which limits the total amount of money that may be ordered, both speak in terms of the jurisdiction or orders of the tribunal. A court receiving proceedings under s 17 possesses a wider general monetary jurisdiction. Even so, s 17 authorises the transfer of “those proceedings”. It does not say that the statutory claim is enlarged upon transfer, that the receiving court may award remedies unavailable in the ECT, or that the conditions which had to be satisfied when the claim was lodged cease to apply. There is genuine ambiguity, but on balance—given everything said above—the better reading is that the court receives the same claim which entered the ECA machinery and adjudicates it subject to the same limit of $20,000 or $30,000, as the case may be.

Just for completeness, I would add that the power in s 17(3) to make the transfer order “on such terms as the appropriate court thinks fit” does not, in my view, point to a different conclusion. Those terms may regulate the manner of transfer and the subsequent conduct of the proceedings. There is no indication that Parliament intended those general words to empower the court to reconstitute the statutory claim, enlarge the remedies created by ss 14(2) and 84(2), or reverse the express extinguishing effect of s 15(3). Had that been intended, one would expect s 17 to say something about the enlargement of the claim or the disapplication of the limits which otherwise govern it. Instead, the repeated reference is to transferring the existing “proceedings”. The order therefore changes the institution which decides the claim without enlarging either the remedies available or the amount which may be awarded.

I therefore return to the conclusion stated at [494] above. A claim for reinstatement or compensation under ss 14(2) or 84(2) must originate through the mediation and claim-route prescribed by the ECA. The ordinary civil courts remain seised of jurisdiction over independent contractual, tortious, and independently enforceable statutory claims arising from the same dismissal, and they may acquire jurisdiction over an existing ECT claim if an order under s 17 is made. On the construction I consider proper, the receiving court adjudicates the claim which was lodged in the ECT, subject to the same remedies and monetary limits. This gives effect to the protections which Parliament created without detaching them from the institutional, remedial, and monetary limits which Parliament attached to their enforcement.

Institutional fragmentation: Contract claims in the ECT

The second lesser trouble arises from a different form of institutional division. As the preceding section shows, a claim under s 14(2) or s 84(2) is bound to the ECT. Yet, the dismissal which gives rise to such a claim may also give rise to a claim for salary in lieu of notice under s 11(1) of the EA or the contract. The ECA expressly accommodates both: it classifies the former as a wrongful dismissal dispute and the latter, where it falls within the prescribed description, as a salary-related employment dispute, while s 12(7) applies a separate aggregate claim limit to each class. Thus, an employee entitled to invoke both protections may lodge both claims and recover upon each, subject always to the jurisdictional limit applicable to the particular claim and the rule against double recovery.

The difficulty, however, is that wrongful dismissal at common law is convenient shorthand for a breach of the employment contract connected with dismissal, and the claim which results may extend beyond salary in lieu of notice. A failure to give the notice required by the contract supplies the ordinary example. Since the common law measure will commonly confine the practical value of the action to the remuneration which would have been earned during that period, the wider possibility may easily be overlooked. Even so, another contractual term may have been broken. A disciplinary or dismissal procedure may, for example, possess contractual force, with the consequence that non-compliance may found a contractual claim even where the employer has paid notice in full (see Gunton; Aldabe Fermin at [90]–[112]).

As I explained in JGP, a contractual disciplinary framework may, depending upon its language and incorporation, confer rights which extend beyond a claim for salary in lieu under s 11(1) and which arise separately from the statutory protection against dismissal without just cause or excuse. The existence and limits of that possibility are illustrated by Aldabe Fermin. There, the High Court accepted the principle in Gunton and held that the bank’s disciplinary materials had been incorporated, though, upon the proper construction of that particular contract, it concluded that the express clauses authorising immediate termination prevailed over the inconsistent incorporated procedure (at [99]–[112]). The point which matters for present purposes is that the wider contractual inquiry existed and was not exhausted by asking whether salary in lieu had been paid.

That wider contractual inquiry does not, however, fall generally within the ECT’s jurisdiction. The ECA does not purport to confer upon the ECT jurisdiction over every claim for damages arising from breach of an employment contract. Its salary-related jurisdiction is instead attached to the contractual disputes prescribed in the First Schedule and the statutory disputes prescribed in the Second Schedule. Thus, a contractual claim for salary in lieu falls within the former, while ss 11(1) and 16 of the EA supply notice-related statutory claims within the latter. Section 16 is significant because it supplies a notice-equivalent sum where either party terminates in breach of contract. Still, it does not turn the ECT into a court of general contractual jurisdiction (see JIS v JIT [2026] SGECT 5 (“JIS”) at [9]–[12]). It follows that, although an employee may be able to place both the statutory dismissal claim and the prescribed notice-pay claim before the ECT, a distinct contractual claim arising from the same termination must be pursued, if at all, in an ordinary civil court.

The same jurisdictional limit is illustrated by Goh Hui En Rebecca. The court there emphasised that the ECT may determine only those employment disputes which the ECA places within its jurisdiction, and that the statutory bar on other proceedings applies to “that claim”, rather than to every legally distinct claim arising from the same employment relationship (at [19]–[29] and [67]–[79]). The ordinary doctrines of issue estoppel, res judicata and abuse of process, as well as the rule against double recovery, may nevertheless bear upon any later contractual action. Those qualifications do not affect the narrower point which matters here: factual overlap with an ECT claim cannot bring within the ECT’s jurisdiction a contractual cause of action which the ECA does not empower it to determine.

In practice, I think that the trouble will often be considerably smaller than its conceptual form suggests. The common law minimum-obligation rule ordinarily confines damages to the position which would have obtained had the employer performed the contract in the least onerous lawful manner which it could and would have adopted. In the usual wrongful dismissal case, this will mean the remuneration and benefits which would have accrued during the contractual notice period. Where a binding disciplinary procedure ought to have preceded dismissal, the period may extend to the time which proper observance of that procedure would have occupied, though it will not ordinarily support damages upon the premise that employment would have continued indefinitely (see Wee Kim San at [25] and [37]–[39]; Aldabe Fermin at [90]–[95] and [116]; Long Kim Wing v LTX-Credence Singapore Pte Ltd [2017] SGHC 151 (“Long Kim Wing”) at [176]–[186]; Barker at pp 52–54). As such, salary in lieu will exhaust, or substantially exhaust, the ordinary contractual recovery in many cases.

Therefore, although the ECA does not consolidate the whole common law action, the practical duplication is unlikely to be substantial. A claimant seeking more must still identify a distinct contractual promise, establish its breach, and prove recoverable loss which has not already been answered by notice pay or another award. For present purposes, I recognise that conceptual possibility but do not pursue it further.

The relationship between sections 14(2) and 35(3)

The third group of lesser troubles concerns statutory overlap. I begin with the closest relationship, which is that between s 14(2) of the EA and s 35(3) of the IRA. Their history has been examined from [309]⁠–⁠[351] above. What is now s 35(3) formed the Ministerial substitute, introduced in 1968, for the dismissal jurisdiction which Parliament withdrew from collective bargaining and the IAC, while s 14 was enacted in 1973 to extend substantially the same protection to non-unionised employees. Each provision continues to use the same substantive formula, permitting an employee who considers that she has been dismissed “without just cause or excuse” to invoke the route which it creates.

The ECA confirms that the two routes are alternatives in respect of the same statutory grievance. Sections 16(5) and 16(6) (formerly ss 16(4A) and 16(4B)) prevent an employee from maintaining both an ECT claim and representations under s 35(3) in respect of the same dismissal. Thus, a claimant who has lodged an ECT claim is taken to discontinue it upon making the representations, while the making of those representations prevents the corresponding ECT claim from being pursued in the manner prescribed. The provisions do more than prevent double recovery. They treat the two avenues as mutually exclusive responses to the same event and protection.

Against that background, one would expect the phrase “without just cause or excuse” to carry the same substantive meaning in each Act. The ordinary presumption that identical language used for closely related protections should be construed coherently points in that direction, as do the provisions’ common history and the ECA’s treatment of the two routes as equivalents. Indeed, it would be difficult to justify a position under which the same proved reason was sufficient before the Minister but insufficient before the ECT simply because an eligible employee had selected one route instead of the other.

There are, however, limits to the degree of institutional coherence which can presently be achieved. The ECT is a court which exercises the adjudicative powers and duties explained and described at [405]⁠–⁠[439] above, whereas the Minister remains an administrative decision-maker acting within the industrial-relations framework of the IRA. The ECT must have regard to the Guidelines, applies the particular burden provisions in s 27 of the ECA, and is confined by the ECA and EC Regs as to the form and monetary amount of relief. Those provisions do not govern the Minister’s decision under s 35. It follows that the incidents of the two proceedings cannot be made identical merely because the substantive words coincide.

There is, in addition, a material difference in the evidential position from which the Ministerial machinery begins. Section 35 permits the Minister to request the Commissioner in writing to inquire into the dismissal and report. If the Minister decides to deal with the representations himself, he must, before deciding them, give the employer an opportunity to make written representations explaining why it considered the dismissal justified. As such, the statutory sequence brings the employer’s asserted justification into view. The residual position before the ECT is different. Subject to s 27(2) and the qualifications discussed at [119]⁠–⁠[171] above, the employee may bear the legal burden of proving the absence of just cause or excuse.

I therefore think that, so far as their shared words and purpose permit, the substantive standard should carry the same meaning. Even so, the identity of the decision-maker and the procedure prescribed for each route affect the evidence which will be elicited, the burden borne by either party, the remedial choices available, and the manner in which reasons and precedent develop. Parliament has supplied no express rule explaining how a published judicial construction and an unpublished Ministerial practice are to be kept coherent if they diverge. For that reason, the two positions cannot presently be made to coincide in every respect.

That disjunction may deserve legislative attention when an appropriate occasion arises, so that Parliament may state whether the substantive elements are intended to be identical and identify the incidents which should properly differ because of the institution concerned. I do not, however, regard the present difference as inherently intolerable. Over time, the two protections have acquired different constituencies, procedures, and remedial settings, and the ECA expressly requires an eligible employee to choose between them. Thus, a unionised employee whose trade union is willing and entitled to represent her may elect the Ministerial route, while another may proceed under s 14(2). The ECA may therefore tolerate different procedures and remedies for the alternative routes, even though their precise juridical relationship remains under-explained.

The relationship between sections 14(2) and 35(2)

The possible overlap with s 35(2) is narrower. Section 35(2) preserves the IAC’s jurisdiction where an employee is dismissed in circumstances arising out of a contravention of s 82 of the IRA. Section 82, in turn, protects employees against dismissal or threatened dismissal, injury or threatened injury in employment, and alteration or threatened alteration of position to their prejudice, where the employer acts by reason of a specified trade-union circumstance. Its concern, as such, is victimisation within the field of collective labour relations, which is appreciably narrower than dismissal without just cause or excuse generally.

A dismissal for protected trade-union activity may, in principle, engage both provisions. The prohibited reason under s 82 would ordinarily be incapable of supplying just cause or excuse under s 14(2), while the same facts may found the victimisation dispute over which s 35(2) leaves the IAC seised. Yet, that coincidence does not erase the difference in juridical structure. Section 14(2) states a general reasons-based restraint enforced through the ECT. Sections 35(2) and 82, by comparison, identify specified trade-union reasons, operate through the IRA’s industrial-dispute machinery, and leave adjudication with the IAC.

The reported awards compiled for these grounds also show that victimisation references have been exceptionally uncommon for a considerable period. Though the older cases illustrate the work of the protection, they disclose no recurring modern problem of concurrent litigation. Further, as Arokiasamy Joseph v Singapore Airlines Staff Union [2000] 1 SLR 473 makes clear, the statutory route depends upon the circumstances specified in s 82 and the industrial-relations machinery by which the matter reaches the IAC. Union membership, without the necessary causal reason and procedural foundation, does not suffice. I therefore think that the theoretical overlap may be acknowledged without constructing an elaborate reconciliation for a difficulty which has not materially troubled present practice.

The different formulations of sections 14(2) and 84(2)

The remaining statutory overlap arises from the different words which the EA uses for two closely related protections. Section 14(2) applies where an employee considers that she has been “dismissed without just cause or excuse”. Section 84(2), read with s 84(1), applies where the qualifying female employee considers that a “notice of dismissal” was “not given for sufficient cause”. Thus, the former speaks of just cause or excuse for dismissal, while the latter speaks of sufficient cause for giving notice.

There are, of course, literal differences between the formulations. “Just” directs attention to the normative quality of the cause, “sufficient” to its weight, and “excuse” may extend beyond a positive cause in the strict sense. Section 84 is also tied to notice given during the protected pregnancy period and to the payment protection in s 84(1), whereas s 14 applies to dismissal more generally. Those differences explain, at least in part, why the pregnancy provision retains language which originated in the statutory protection of maternity payments, while the later general provision carried over the industrial formula from the IRA.

However, when the provisions are read in their statutory context and interpreted purposively in the ordinary manner required by s 9A of the Interpretation Act, their substantive inquiries converge (see Tan Cheng Bock at [37]–[54]). In each provision, Parliament made the employer’s reason for bringing the relationship to an end the object of inquiry. In each, the employee may seek reinstatement or compensation through the ECT. The ECA also classifies both as wrongful dismissal disputes in respect of which the ECT must have regard to the same Guidelines. In these circumstances, a cause sufficient in weight but unjust could hardly answer a statutory protection directed to pregnancy; likewise, a cause just in character but insufficiently weighty could hardly answer s 14(2).

The personal and temporal gateways remain different, and s 84(1) produces a payment consequence which s 14 does not. The burdens of proof also differ. Section 27(2)(c), and para (d) where the Child Development Co-Savings Act 2001 (Act 13 of 2001) (“CDCSA”) route is engaged, place the merits burden in a s 84(2) claim upon the employer, whereas a residual s 14(2) claim may leave the legal burden upon the employee. Once the merits inquiry is reached, however, I do not think that the different formulations prescribe a salient difference in the intensity of substantive justification. The allocation of the risk of non-persuasion does not make a cause more or less sufficient. As the analysis of “dismiss” at [573]⁠–⁠[587] below will show, the provisions should likewise be read coherently, so far as their words reasonably permit, as to the event which engages them.

There is no provision which makes the two claims mutually exclusive in the manner that ss 16(5) and 16(6) govern the Ministerial route. A qualifying pregnant employee may therefore be eligible to invoke both, though each claim must satisfy its own gateway and burden and any relief must be coordinated to avoid double recovery for the same wrong. The linguistic seam is, as such, unnecessary and capable of generating argument. When an appropriate opportunity arises, it would be useful for Parliament to state expressly how the protections relate, whether either was intended to confer an additional substantive standard, and how concurrent eligibility should be treated. Until then, I think that the better construction gives genuine effect to their different gateways, burdens, and historical forms, while treating each as requiring a factually established and sufficiently weighty reason which is just in the circumstances.

Various uncertainties about process prior to dismissal

The fourth group of lesser troubles concerns the significance of what an employer does before dismissing an employee. There may be a temptation to use “due process” as though it described a single and general obligation. The legislation, however, does not take that form. Section 14(1) expressly requires “due inquiry” before a particular kind of summary dismissal; the contract may separately confer a procedural right; and the facts generated by an investigation, warning, consultation, or hearing may bear evidentially upon one or more statutory claims. Before considering their effect, it is necessary to keep those sources and functions apart.

The bearing of “due inquiry” on section 14(2) claims

Section 14(1) provides that an employer may, “after due inquiry”, dismiss without notice an employee on the ground of misconduct inconsistent with the fulfilment of the express or implied conditions of service. Section 14(2), however, operates “[d]espite subsection (1)” and asks whether the employee was dismissed without just cause or excuse. Since Parliament used different formulations, one cannot safely treat the express procedural condition in s 14(1) as an unstated and invariably dispositive element of s 14(2).

This much follows from the analysis at [113]⁠–⁠[116] above. A failure to conduct due inquiry may show that the employer was not entitled to invoke s 14(1) to withhold notice, though whether salary in lieu then becomes due depends upon the contract, ss 10 and 11, and any other termination route upon which the employer can properly rely. By itself, however, such a failure does not necessarily establish the absence of a substantively sufficient cause or excuse for dismissal. Poor performance may, for example, justify ending the relationship upon notice without amounting to misconduct which permits notice to be withheld. Conversely, serious misconduct may be proved at trial despite an inadequate pre-dismissal inquiry. The notice question, merits question, and evidential question must therefore remain distinct.

There is now a direct, though qualified, first-instance answer. In JGP, I held that an employer which relies upon misconduct as just cause or excuse must satisfy the safeguard of due inquiry even where the dismissal was effected with notice or salary in lieu (at [82]–[90] and Annex A at [A.134]). That holding gives process substantive work beyond the express circumstances in s 14(1). However, it is an ECT decision, its extension beyond the enacted words of s 14(1) is contestable for the reasons which follow, and no reported appellate decision has resolved the point. It is therefore necessary to state both the holding and the uncertainty which attends it.

The more difficult question concerns the work which due inquiry should perform within the merits inquiry. One possibility is that its significance is evidential. A contemporaneous investigation which fairly identifies the allegation, permits an answer, and tests the available material may lend weight to the employer’s account that the misconduct existed and actually caused the dismissal. Conversely, its absence may weaken that account, especially where the fact was contestable, depended upon an inference, or could readily have been answered by the employee. On this view, process affects the reliability and weight of the proof without itself constituting the statutory wrong.

The other possibility is that process bears directly upon justification. The word “just” may be thought broad enough to include the fairness of the means by which the employer reached its decision, particularly where dismissal carries a disciplinary judgment upon the employee. The inherited IAC practice and the Guidelines also contain procedural considerations which may support that understanding. If that is correct, a reason which was factually sufficient may nevertheless fail to constitute “just cause or excuse” where the employer acted upon it without giving the employee a fair opportunity to answer it.

The text supplies no obvious resolution. The express inclusion of “due inquiry” in s 14(1), coupled with its omission from s 14(2), supports the evidential conception. At the same time, the breadth and evaluative character of “just cause or excuse”, read with the industrial history, may support the substantive conception. Nor does the phrase “[d]espite subsection (1)” resolve the matter. Though it establishes that s 14(2) remains applicable despite the particular summary-dismissal power, it does not explain whether the process prescribed by s 14(1) bears only upon that power or also upon the quality of the employer’s justification.

For present purposes, I think that the soundest minimum position is the evidential one adopted in the reconstruction above. Due inquiry may bear powerfully upon whether the asserted ground existed, operated, and was sufficiently just. I treated it in JGP as a substantive safeguard where misconduct is relied upon, though the uncertain statutory basis of that holding counsels against enlarging it into a universal procedural element of every s 14(2) claim. There may also be cases in which the nature of another asserted cause makes an opportunity to answer practically inseparable from the conclusion that dismissal was just. Even so, whether process can more generally supply a separate and freestanding basis of liability remains uncertain. Legislative clarification would be valuable, particularly because the Guidelines presently encourage the procedural and substantive questions to be treated as one.

The bearing of “due inquiry” on other statutory claims

The expression “due inquiry” does not appear in the other dismissal protections considered in these grounds. Section 84 asks whether notice was given for sufficient cause; s 35(3), whether dismissal was without just cause or excuse; and ss 35(2) and 82, whether specified trade-union activity supplied the reason for prejudicial treatment. None of those provisions directly incorporates the procedural condition enacted in s 14(1).

Accordingly, the condition should remain confined to the source which enacts it. Nevertheless, the procedure followed may bear evidentially upon another statutory inquiry. A reliable investigation may support the existence and operative force of the reason asserted, whereas the absence of contemporaneous inquiry may expose a later account as conjectural, mistaken, or pretextual. The weight which either circumstance bears will depend upon the nature of the allegation and the institution deciding it.

In this connection, the Minister’s power under s 35 to direct the Commissioner to inquire must be distinguished from the employer’s due inquiry under s 14(1). The former belongs to the statutory process by which representations are investigated after dismissal. The latter is a condition attached to the employer’s power to dismiss summarily for misconduct. Though the Ministerial inquiry may develop the evidential record by obtaining material, it cannot retrospectively mean that the employer conducted the pre-dismissal inquiry described in s 14(1).

The bearing of “due inquiry” on contractual claims

At common law, there is no general rule requiring an employer to conduct a fair inquiry before exercising a contractual power to terminate upon notice. That is consistent with the absence of any ordinary general requirement of natural justice whenever one contracting party exercises a right prejudicial to another (see Leiman, Ricardo v Noble Resources Ltd [2020] 2 SLR 386 at [133]–[134]). A contractual claim must therefore begin with the terms agreed between the parties, since s 14(1) does not supply a contractual procedure.

The contract may, however, require an inquiry, disciplinary hearing, warning sequence, appeal, or some other process. That obligation may be express or arise through the incorporation of a sufficiently certain staff policy or collective instrument. Where it does, a failure to follow the procedure is capable of constituting a breach distinct from the failure to give notice. The principle in Gunton permits damages, in an appropriate case, for the remuneration which would have accrued during the time required to complete the contractual procedure. Aldabe Fermin accepted that analysis as a matter of principle, though it held that the express immediate-termination clauses in the contract before the court prevailed over the inconsistent incorporated procedure (at [90]–[112] and [116]). Long Kim Wing, in turn, supplies a local illustration of an award for the reasonable period needed to complete a contractually required inquiry (at [176]–[186]).

There is no necessary identity between such a contractual procedure and “due inquiry” under s 14(1), because the contract may require more, less, or something different. Whether its terms are incorporated, mandatory, and applicable to the dismissal is a matter of construction. Equally, s 14(1) cannot, without more, convert a non-contractual human-resources policy into a contractual promise. The statutory and contractual procedures may overlap in fact, but each draws its legal force from a different source.

The relationship between sections 11(2) and 14(1)

A further difficulty arises because ss 11(2) and 14(1) may both be invoked when an employer terminates without notice. Section 11(2) is bilateral: either party may terminate without notice if the other has wilfully breached a condition of the contract of service. Section 14(1), by contrast, is unilateral: after due inquiry, the employer may dismiss without notice for misconduct inconsistent with the fulfilment of an express or implied condition of service, or impose one of the lesser disciplinary sanctions identified there.

There is substantial factual overlap. Wilful disobedience, dishonesty, or another serious breach by an employee may amount both to a wilful breach of a contractual condition and to misconduct inconsistent with its fulfilment. Yet, the provisions are not coextensive. Section 11(2) requires a wilful breach and is available to either party, whereas s 14(1) speaks of misconduct, applies only to the employer, requires due inquiry, and creates a disciplinary power which extends beyond dismissal. Wilfulness and misconduct may therefore coincide in many cases without becoming synonymous (also see Barker at pp 33–35).

The Guidelines’ statement that misconduct is the only legitimate reason for dismissal without notice does not resolve the relationship, since it would leave s 11(2) without work in the dismissal setting. At the same time, to treat s 11(2) as an unrestricted alternative whenever the employee commits a breach would allow the employer to avoid the due-inquiry condition by describing the same conduct as wilful contractual breach. Each available construction therefore presents a difficulty. One treats s 14(1) as the more specific provision governing an employee’s disciplinary dismissal and s 11(2) as the general contractual termination provision; the other allows both provisions to operate concurrently according to their distinct elements and consequences.

For present purposes, it is enough to say that an employer relying upon either route must prove the facts which engage it, and that a failure to justify withholding notice does not automatically determine the separate inquiry under s 14(2). Nevertheless, the coexistence of the provisions—and, in particular, the possibility that s 11(2) may be used to avoid the express process in s 14(1)—would benefit from legislative clarification.

The relationship between contract and section 14(1)

Much the same difficulty appears in the relationship between an express contractual power of summary dismissal and s 14(1). A contract may specify acts which entitle the employer to terminate immediately, incorporate a disciplinary code, or reserve an apparently broad power to dismiss for misconduct. At common law, the construction and operation of those terms determine the contractual entitlement. Subject to an effective limiting term, an employer may generally rely upon repudiatory misconduct existing at dismissal even if it discovered that misconduct only later (see Boston Deep Sea Fishing; Aldabe Fermin at [44]–[50]; Phosagro Asia at [42]–[48]; Blackstone’s at paras 25.02–25.03).

Section 14(1), however, is a statutory power and condition applicable within the EA. To the extent that the Act applies and the contract offers terms less favourable than its requirements, s 8 of the EA prevents the contract from removing the statutory protection. An express clause cannot therefore be assumed to dispense with due inquiry simply because it permits “immediate” dismissal, though the precise interaction may depend upon whether the employer invokes s 14(1), s 11(2), the contract, or common law repudiation. The recent decision in Kanan Packrisamy v Herbal Pharm Pte Ltd and others [2026] SGHC 2 illustrates the need to keep those sources distinct, but no more should be attributed to it. The court decided contractual and common law issues, rejected an attempt to rely upon an unpleaded implied contractual term, and considered the MOM guidelines and Long Kim Wing in that setting. It did not determine the operation of s 14(1) as a statutory cause of action (at [124] and [131]–[135]).

Conversely, a contract may confer a procedural protection more favourable than the statutory minimum. Where that procedure applies, compliance with s 14(1) does not answer the contractual claim; nor does a contractual breach necessarily determine whether the statutory condition was satisfied. Thus, an employer may comply with one source and breach the other. This supplies another reason why the loose expression “due process” should be avoided unless the duty, its source, and its consequence are identified.

The bearing of contractual due process on any statutory claim

Where an employer has undertaken a contractual disciplinary procedure, its compliance or non-compliance could potentially bear upon a statutory claim. The most direct legal consequence would, of course, remain contractual. Nevertheless, the existence of the promise may show what the parties regarded as an appropriate means of establishing misconduct or performance failure, while the records produced through the procedure may prove or disprove the cause later advanced. Further, a deliberate departure from it may support an inference that the stated ground was not the true operative reason, or that the employer lacked confidence in its factual basis.

That evidential relevance does not permit the contract to define the statutory standard. A procedure drafted for a particular organisation may be more generous than s 14(2), s 84(2), s 35(3), or s 82 requires, and a minor contractual irregularity may have no rational bearing upon the sufficiency of the justification for dismissal. Conversely, exact compliance with a procedure cannot make a false, irrelevant, discriminatory, or insufficient reason just. The statutory decision-maker must therefore ask what the contractual process reveals about the facts and justification, while giving neither breach nor compliance conclusive effect.

The significance of process on dismissals not for misconduct

Finally, process may matter even where the employer alleges no misconduct. Poor performance is commonly assessed over time against communicated standards, so warning, feedback, support, and an opportunity to improve may assist in proving that a deficiency was real, persistent, understood, and sufficiently serious. In the same way, consultation and selection materials may show whether a redundancy or reorganisation was genuine, while medical assessment and dialogue may bear upon incapacity. Since the causes differ, the precise relevance of the process must differ with them.

Nothing in s 14(2) creates a general statutory requirement of a hearing, warning, consultation, or improvement plan for every category of dismissal. To derive one would add a substantial positive obligation which Parliament did not express and whose appropriate form would vary across employment settings. Even so, the absence of the process which a particular asserted reason would ordinarily be expected to generate may leave the employer unable to prove that reason, its operative connection, or its sufficiency. Similarly, a process undertaken only after the decision was effectively made may carry little weight.

Drawing these subsections together, the safest conclusion is, in my view, a modest one. Process has direct legal force where an applicable statute or contract gives it that force. Otherwise, it ordinarily supplies evidence bearing upon the existence, causal operation, and justice of the reason for dismissal. The present enactments do not clearly answer whether some forms of procedural unfairness may independently make a dismissal unjust under the statutory merits protections. Legislative clarification upon that question would improve the law more satisfactorily than the unstructured migration of “due inquiry” from one provision to another.

Various uncertainties relating to the definition of “dismiss”

As I mentioned near the outset of these grounds (see [4] above), the starting point is the following definition of “dismiss” contained in the EA:

That definition gives rise to three uncertainties. The first concerns whether it applies to the claim which a pregnant employee may bring under s 84(2) of the EA. The second concerns the meaning of “dismiss” in s 35(3) of the IRA. The third concerns the relationship between the limb directed to a “forced” resignation and the common law conception of “constructive dismissal”.

The applicability of the definition to section 84(2)

Under s 84(2), a female employee may bring the claim where she considers that “a notice of dismissal given to her” was not given for sufficient cause. Section 14(2), however, permits a claim where the employee considers that she has been “dismissed without just cause or excuse” by her employer. Once the definition of “dismiss” in s 2(1) is borne in mind, the difference between the two formulations appears somewhat curious.

There is no obvious reason why the general protection against dismissal “without just cause or excuse” and the narrower pregnancy-related protection against dismissal “without sufficient cause” should be engaged by different events. Both concern an employer-initiated termination; both permit the employee to seek reinstatement or compensation; both are now routed, in most cases, through the ECT; and the ECA treats both as “wrongful dismissal disputes”. Their statutory language is nevertheless different. It is therefore necessary to consider whether that difference carries any significance.

The position under s 14(2) is relatively straightforward because its operative word is “dismissed”. Since “dismiss” is defined in s 2(1), the natural question is whether the employee was dismissed within that definition. She is dismissed if the employer terminates the contract of service at its initiative, whether with or without notice and whether for cause or otherwise. She is also dismissed where she resigns but proves, on the balance of probabilities, that the resignation was involuntary and that the employer’s conduct, omission, or course of conduct or omissions forced it.

Section 84(2), by contrast, makes “given” the operative verb in the phrase “a notice of dismissal given to her”. Read literally, the provision applies where the employer gives the female employee a notice of dismissal, and that sits uneasily with the present definition of “dismiss”. Suppose, for example, that an employer gives no notice but engages in a course of conduct which causes the employee to resign. The resignation may fall within s 2(1) if she proves that it was involuntary and forced by the employer’s conduct. Yet, upon a literal reading of s 84(2), the absence of a notice would leave that forced resignation within s 14(2) alone.

That would be an odd result. In my view, the literal drafting should not be accepted without considering how it came about. The substantive protection now found in s 84 has existed since the EA was enacted in 1968. It then appeared in s 104, within Part X, titled “Employment of Women”. Section 104(1) provided that “[n]o notice of dismissal given without sufficient cause” by an employer to a female workman within three months before confinement would deprive her of maternity allowance. Section 104(2) then required any question whether “any notice of dismissal given” under s 104(1) was or was not given for sufficient cause to be referred to the Minister, whose decision was final. The operative verb “given” has, as such, been present from the beginning.

The formulation now found in s 14(2), which uses “dismissed” directly, did not appear in the EA upon enactment. As explained at [333]⁠–⁠[341] above, s 14 was added only in 1973. Even so, the formulation was known to Parliament. During the same legislative season in which it enacted the EA, Parliament considered the Industrial Relations (Amendment) Bill 1968, which inserted into the 1960 Ordinance the predecessor of the present s 35(3) of the IRA. That provision used the language now familiar through s 14(2) of the EA: “where an employee considers that he has been dismissed without just cause or excuse by his employer”. It may therefore be tempting to infer that Parliament’s use of “dismissed” in one setting and “notice of dismissal given” in another reflected an intention to confer materially different threshold scopes.

I do not think that would be a sound inference, because it attributes too much significance to a textual difference created within a legal world materially different from ours. The Second Reading debates upon the Employment Bill 1968 were concerned overwhelmingly with its broad economic purpose, the urgent need for industrialisation, the preservation of industrial peace, and specific controversies over retirement age, working hours, bonus, sick leave, and related matters (see, generally, Singapore Parliamentary Debates, Official Report (10 July 1968) vol 27 at cols 470–523; ibid (11 July 1968) vol 27 at cols 524–568; ibid (12 July 1968) vol 27 at cols 569–620; and ibid (15 July 1968) vol 27 at cols 621–686). The only express point about maternity benefits appears to have concerned why they were confined to female workmen instead of extending to female employees generally (Singapore Parliamentary Debates, Official Report (15 July 1968) vol 27 at col 651 (Mr S Rajaratnam, Minister for Foreign Affairs and Minister for Labour)). Nothing in those debates suggests that Parliament had drawn, still less consciously intended, a fine distinction between an employee being “dismissed” and a “notice of dismissal” being “given” to her.

More importantly, the difference between ss 14(2) and 84(2) only acquired its present significance when the involuntary-resignation limb was added to the definition of “dismiss” in s 2(1) in 2018. It would, as such, be anachronistic to treat the phrase “notice of dismissal given” in the original s 104 as though Parliament intended it to exclude, or to sit apart from, a statutory conception of involuntary resignation which did not then exist.

The history reinforces that point. The employment-law concept now called “constructive dismissal”—under which the employer’s repudiatory conduct causes the employee to resign and the law treats that resignation as a dismissal—entered the mainstream common law vocabulary only later. Its modern contractual formulation is usually attributed to Western Excavating (E.C.C.) Ltd v Sharp [1978] QB 761 (“Western Excavating”), where Lord Denning MR, giving the lead judgment for the English Court of Appeal, rejected a broad reasonableness test and asked whether the employer had committed a significant breach going to the root of the contract, or shown an intention no longer to be bound by one or more essential terms, such that the employee was entitled to leave. I do not suggest that the ordinary law of contract was unaware, before Western Excavating, of termination in response to a counterparty’s repudiatory breach or repudiation. My narrower point is that the employment-specific idea by which an apparent resignation could be regarded in law as a dismissal was unlikely to have formed part of the general legal grammar against which s 104 was drafted in 1968.

Seen in that light, it is most unlikely that Parliament, when it used the phrase “notice of dismissal given” in 1968, contemplated that those words might one day sit awkwardly with a later statutory conception of involuntary resignation. The likelier explanation seems to me more prosaic. Section 104 was concerned with a female workman’s entitlement to maternity allowance and with preventing an employer from defeating it by issuing a notice of dismissal shortly before confinement. The drafting therefore focused upon the usual instrument by which the mischief would occur: the notice of dismissal. As the EA later accumulated amendments—including the 2008 definition of “dismiss” and the 2018 addition of the involuntary-resignation limb (see [592]⁠–⁠[594] below)—the statute was not rewritten as a fully integrated conceptual code. Older phrases remained while newer concepts were added, leaving linguistic seams behind.

Against that history, I think that legislative accretion provides the better explanation for the difference in wording. The wording supplies too uncertain a basis for attributing to Parliament a policy of excluding forced resignations from s 84(2), though the text would plainly benefit from correction.

For those reasons, I do not think that the difference should be allowed to produce an irrational distinction between ss 14(2) and 84(2). Quite apart from the history, s 9A of the Interpretation Act requires a construction which promotes the purpose or object of the written law to be preferred to one which does not. In Tan Cheng Bock, the apex court described that exercise in three steps: first, one identifies the possible interpretations of the provision, having regard to its text and statutory context; second, one ascertains the legislative purpose; and third, one compares the possible interpretations against that purpose (at [37]–[54]).

In undertaking that exercise, primacy must be given to the text and statutory context. Extraneous material may nevertheless be used in the situations identified by s 9A(2) and with the caution stated in Tan Cheng Bock (at [43]–[47] and [54(c)]). That approach accords with the general modern conception of purposive interpretation, under which a court seeks to give effect to the purpose of the enactment without doing violence to its words (see, eg, Craies at paras 18-001–18-004; Bennion at pp 398–402).

At the relevant level of generality, the purpose of s 84 is not difficult to identify. It protects female employees from being deprived of maternity-related protection by dismissal during pregnancy, save where sufficient cause exists. Were s 84(2) confined to a formal notice issued by the employer, the protection could be avoided by doing indirectly what could not safely be done directly. Instead of issuing a dismissal notice to a pregnant employee, an employer could make her working conditions intolerable, withdraw essential support, isolate or threaten her, or otherwise pressure her to resign. Such a construction would permit evasion of the very protection which s 84 was enacted to supply (cf Bennion at pp 421–425, 436–437), and strongly tells against it.

Thus, even without the perspective which history supplies, the purposive method required by s 9A(1) of the Interpretation Act leads me to the view that the material scopes of ss 14(2) and 84(2) should be read consistently. Section 84(2) plainly applies where the employer has issued a formal notice of dismissal to a pregnant employee. Where she alleges that she was forced to resign during the protected period, the textual position is less neat. Nevertheless, the definition of “dismiss” in s 2(1), the purpose of s 84, and the policy against evasion point in the same direction. A forced resignation which satisfies the s 2(1) definition should, in principle, be capable of engaging s 84(2), although no formal notice of dismissal was issued.

The meaning of “dismissed” in section 35(3)

Like s 14(2) of the EA, s 35(3) of the IRA may be invoked where “an employee considers that he has been dismissed without just cause or excuse by his employer”. However, the IRA contains no definition of “dismiss” equivalent to that found in s 2(1) of the EA. It follows, at least as an immediate matter, that the definition in s 2(1) cannot, within the ordinary rules of statutory interpretation, be said to apply directly to s 35(3). It is prefaced by the words “In this Act”, and no provision of the IRA incorporates it. The word “dismissed” must, as such, be construed within the IRA’s own statutory setting.

That does not, however, render the EA definition irrelevant. The identical statutory language, the closely related history of the two provisions, and their subsequent treatment as alternative routes of redress may properly inform whether an independent construction of s 35(3) produces materially the same result. For this purpose, it is useful to distinguish an employer-initiated termination from a resignation which is said to have been forced.

An employer-initiated termination presents little difficulty. In its ordinary sense, an employee is dismissed by her employer where the employer brings the employment relationship to an end. Whether notice was given concerns the manner and timing of the termination, while whether there was cause concerns its justification. Neither alters the identity of the party which brought the relationship to an end. Section 35(3) itself maintains that distinction. It first requires a dismissal “by [the] employer” and then asks whether that dismissal was “without just cause or excuse”. The two inquiries would be improperly collapsed if “dismissed” were confined to summary dismissal, dismissal for misconduct, or some other termination which was itself contrary to the contract.

That conclusion is consistent with the wider statutory setting. Under s 35(6), the Minister may direct reinstatement, payment of lost wages, or compensation “notwithstanding any rule of law or agreement to the contrary”. It would sit oddly with that provision if an employer could prevent the statutory inquiry from arising merely by complying with a contractual notice clause. The contractual right to terminate with notice may determine whether the termination was wrongful at common law. It does not determine whether the employer dismissed the employee for the purpose of a statutory protection enacted to impose an additional restraint upon the exercise of that right.

It also accords with the history of the two provisions. As explained at [309]⁠–⁠[345] above, s 14(2) of the EA and what is now s 35(3) of the IRA formed parts of the same legislative settlement, though in two stages: the latter was enacted in 1968, and the same substantive protection was extended to non-unionised employees through s 14 of the EA in 1973. When introducing the former protection in 1968, the Minister for Labour described the employer’s right to dismiss as being “subject to certain restraints” and explained that directing individual dismissal cases to the Minister would provide speedier machinery for their resolution (31 July 1968 Debates at cols 739–740 (Mr S Rajaratnam, Minister for Foreign Affairs and Minister for Labour); also see OHL (2018) at pp 366–369). When the first limb of the present EA definition was enacted in 2008, the Acting Minister for Manpower similarly explained that an unduly narrow interpretation of “dismissal” would defeat the mechanism for redress against unjustified dismissals (Singapore Parliamentary Debates, Official Report (18 November 2008) vol 85 at cols 953–954 (Mr Gan Kim Yong, Acting Minister for Manpower)). Though that explanation concerned the EA, the interpretive difficulty it identified would arise with equal force under s 35(3). There is no apparent reason why the older Ministerial route should reproduce the very limitation which Parliament considered inimical to the redress mechanism under s 14(2).

The omission of an equivalent definition from the IRA is nevertheless material. Parliament inserted the first limb of the definition into the EA in 2008 and added the limb concerning forced resignations in 2018, but made no corresponding amendment to s 35(3). The latter omission is especially conspicuous because the 2018 legislation amended the IRA in other respects and expressly addressed the relationship between s 35(3) representations and ECT claims. It would therefore be wrong to proceed as though Parliament had reproduced the s 2(1) definition in the IRA. The omission cautions against assuming that every marginal application of the two provisions must be identical. Still, its effect should not be carried further than that. It identifies the Act within which the enacted definition operates; without more, it does not prescribe a contrary meaning for the undefined word in s 35(3).

Indeed, two features of the 2018 legislation tell against an inference that Parliament intended a materially narrower gateway under s 35(3). First, the Explanatory Statement described cl 2(a) of the Employment (Amendment) Bill 2018 as “restat[ing] what constitutes a dismissal”. At the Second Reading, the Minister for Manpower likewise described the body of wrongful dismissal cases which MOM had already been hearing as including cases in which employees had “resigned involuntarily” (2018 Debates at 3.56pm (Mrs Josephine Teo, Minister for Manpower)). Those statements cannot determine what s 35(3) had always meant, but they make it unsafe to infer that involuntary resignation was regarded as a wholly new form of dismissal confined to the amended EA. Second, ss 16(5) and 16(6) of the ECA (formerly ss 16(4A) and 16(4B)) treat an ECT wrongful dismissal claim and representations under s 35(3) as mutually exclusive routes. An employee who pursues the latter cannot concurrently maintain the former. That structure does not require every procedural and remedial feature of the two routes to be identical, but it strongly favours a common understanding of the event which opens either route. Otherwise, the legal character of the same termination could depend upon which of the two alternative avenues an eligible employee selected.

The position concerning a forced resignation is more difficult because a resignation is ordinarily an act by which the employee, rather than the employer, terminates the employment. The express words added to s 2(1) in 2018 therefore perform real work. Even so, the language of s 35(3) is capable of looking beyond the formal instrument of termination. The words “dismissed … by his employer” ultimately require responsibility for the termination to be attributed to one party or the other. Where an employee signs a resignation letter but the employer’s conduct has deprived her of any genuine choice whether to remain employed, the existence of that letter does not necessarily prevent the employer from being, in substance, the party which brought the employment to an end. A resignation tendered in response to a final ultimatum that the employee resign or be dismissed provides the clearest example. A sufficiently compelling course of conduct or omission may produce the same conclusion even without an express ultimatum.

A contrary construction would, in addition, make the availability of the protection turn upon the employer’s choice of form. An employer determined to end the relationship could avoid the Ministerial inquiry by procuring a resignation instead of issuing a dismissal letter. Oun Hean Loh, writing before the 2018 amendment, identified precisely that danger and suggested that a resignation obtained through duress or compulsion could be treated as a dismissal for the purpose of an appeal to the Minister (OHL (2018) at p 377). That observation is, of course, not legal authority for the construction of s 35(3). Its narrower significance is that the danger of engineered resignation, and the possibility of looking beyond its form, had already been recognised within the older Ministerial setting. Thus, the problem did not originate with the 2018 definition.

Of course, none of this means that every resignation prompted by employer conduct amounts to a dismissal. It is insufficient that the conduct supplied the employee with a reason or occasion to resign, or that resignation was a reasonable response to it. The employee must establish that the resignation was not voluntary and was forced by the employer’s conduct, omission, or course of conduct. In this respect, the language now found in s 2(1) of the EA supplies a useful and principled articulation of the necessary threshold. It does so because it identifies when substantive responsibility for an apparent resignation should be attributed to the employer, and not because the EA definition governs the IRA.

I therefore conclude that, for the purposes material here, ‘dismissed’ in s 35(3) includes an employer-initiated termination, with or without notice and whether purportedly for cause or otherwise, as well as a resignation which the employee establishes was not voluntary and was forced by the employer’s conduct or omission. That conclusion arises from s 35(3) itself; the definition in s 2(1) of the EA confirms and usefully expresses it, but is not its juridical source. It answers only the anterior question whether s 35(3) is engaged, leaving for separate inquiry whether the dismissal was without just cause or excuse.

The relationship with common law constructive dismissal

The final uncertainty concerning “dismiss” is whether the forced-resignation limb in s 2(1) of the EA simply enacts the common law of constructive dismissal. Though the concepts may cover many of the same events, their juridical language and the inquiries which they require are different.

At common law, constructive dismissal is a contractual conclusion. The employer must commit a repudiatory breach of the employment contract; the employee must accept that breach by terminating; the breach must cause the employee to leave; and the employee must not have affirmed the contract before doing so. This is the framework stated in Western Excavating and applied locally in Cheah Peng Hock v Luzhou Bio-Chem Technology Ltd [2013] 2 SLR 577 at [35]. A sufficiently serious breach of an express term may qualify, as may a repudiatory breach of an applicable implied term. However, no Singapore appellate court has yet adopted a general implied term of mutual trust and confidence. Dong Wei left the question open (at [69]–[82]); Dabbs, Matthew Edward v AAM Advisory Pte Ltd [2024] SGHC 260 (“Matthew Dabbs”) declined to imply the term in law (at [88]–[91]); while Prashant Mudgal v SAP Asia Pte Ltd [2026] 3 SLR 914 (“Prashant Mudgal”) accepted its existence (at [106]–[121]), subject to the important qualification that the term cannot contradict an express termination provision (at [157]–[162]). The conflict between the two High Court decisions therefore remains unresolved at the appellate level.

The statutory definition, however, directs attention to compulsion arising from the employer’s conduct or omission. It requires the employee to prove, on the balance of probabilities, that she “did not resign voluntarily but was forced to do so” because of the employer’s conduct, omission, or course of conduct or omissions. It does not express the common law requirements of repudiatory breach, acceptance, and affirmation. As Prashant Mudgal observed when comparing the EA with s 95(1)(c) of the UK ERA, the EA definition does not, in terms, require the same inquiry into an entitlement to terminate for repudiatory breach (at [146]–[147]).

That threshold is not satisfied merely because the employee had a good or reasonable reason to leave. The word “forced” imposes a demanding threshold. The employer’s conduct must have deprived the resignation of its voluntary character in the relevant legal sense and caused the employee to terminate. Dissatisfaction, an unattractive change, workplace difficulty, or a reasonable preference to leave may therefore fall short. An ultimatum to resign or be dismissed supplies the clearest case, though a sustained course which leaves continuation as no genuine choice may also qualify. The intensity and persistence of the conduct, the alternatives realistically available, the employee’s response, and the temporal connection will bear upon that conclusion.

Common law principles may nevertheless assist by identifying circumstances in which the employer’s conduct was sufficiently grave and causative. Thus, a proved repudiatory breach followed by a causative resignation is relevant evidence, although acceptance remains an election and may still be voluntary in the statutory sense, as Prashant Mudgal illustrates (at [146]–[147] and [186]). The statutory element must therefore be decided separately, and the familiar label “constructive dismissal” cannot confer additional statutory force upon the contractual conditions. Conversely, a forced resignation within s 2(1) establishes only the statutory gateway. An employee who wishes to sue at common law must still identify a contractual breach and satisfy the rules governing acceptance and remedy.

The institutional consequence is that the ECT decides whether a resignation constitutes a statutory dismissal for the purpose of the prescribed claim, whereas an ordinary civil court may decide whether it amounted to an acceptance of repudiation and what contractual damages follow. The same facts may satisfy both doctrines, only one, or neither. For that reason, the safest terminology is to reserve “constructive dismissal” for the common law doctrine and to use “forced resignation” for the statutory limb, unless the context makes the intended conception plain.

A restatement of the law on dismissals

Having examined separately the protections and the points of uncertainty which remain, I now draw the law together. I begin with the foundational distinctions before stating each protection according to its source, institution, scope, procedure, proof, and remedy. Where the reasoning has already been given, I refer back to it instead of repeating it.

The law’s foundational premises

I begin with the two premises already established at [57]⁠–⁠[118], [223]⁠–⁠[241], and [453]⁠–⁠[463] above, which organise the restatement that follows. The first is that a single termination of the employment relationship may engage more than one legal protection. The second is that each protection remains bound to the institution, procedure, burden, and remedy which the law has assigned to it. As I explained at [392]⁠–⁠[404] above, the equivocation in the common label “wrongful dismissal” has tended to obscure both premises.

Protections under contract and statute are separate but connected

I begin with the protection supplied by contract, which is directed to the bargain made between employer and employee. It therefore determines, amongst other things, the work and remuneration promised, the duration of the relationship, the notice required to bring it to an end, any power to terminate immediately by payment in lieu, and any substantive or procedural restriction which the parties incorporated. A contractual claim accordingly asks which term applied, whether it was broken, and what position proper performance would have produced.

The statutory protections considered in these grounds perform a different juridical function. Section 14(2) of the EA protects a relevant employee against dismissal without just cause or excuse, while s 84(2) confers a closely related protection upon a qualifying pregnant employee. Section 35(3) of the IRA permits an eligible employee, through her trade union, to invoke Ministerial intervention upon the same general statutory formula; ss 35(2) and 82 preserve the narrower industrial route where dismissal arose from specified trade-union activity. Though the contract supplies the relationship and termination upon which the statute operates, each of those protections draws its binding force from statute.

Their factual connection nevertheless remains important because the same matters may bear upon more than one claim. An allegation of misconduct may justify summary dismissal under the contract, engage s 11(2) or s 14(1), and be advanced as the just cause or excuse which answers a claim under s 14(2). A contractual disciplinary process may generate evidence which assists the statutory inquiry, just as salary in lieu may satisfy the contractual entitlement while leaving the statutory merits of dismissal unresolved. The protections must therefore be kept conceptually distinct without losing sight of those connections, and the question should be what juridical work each fact performs within each applicable source.

Protections are institution-bound

That brings me to the institutional consequence examined at [405]⁠–⁠[439] and [484]⁠–⁠[539] above, under which a legal protection remains connected to the institution which the law authorises to enforce it. The ECT possesses the jurisdiction which the ECA confers over scheduled salary-related disputes and statutory wrongful dismissal disputes. The ordinary civil courts exercise their general jurisdiction over contractual and other civil claims, subject to statutory limits and to the procedural bars created when an ECT claim is pursued. The Minister determines representations under s 35(3), while the IAC retains its confined dismissal jurisdiction in the circumstances arising from s 82. A decision-maker may, of course, determine an incidental question necessary to decide a matter within its jurisdiction. That power does not confer an omitted cause of action.

The contractual protections against dismissal

As I explained at [458]⁠–⁠[459], [523]⁠–⁠[529], and [556]⁠–⁠[567] above, the contractual protection is constituted by the terms of the particular employment contract, together with any statute which supplies or controls those terms and the ordinary law of contract. Its familiar centre is notice, though a fixed duration, a restriction to specified causes, an incorporated disciplinary procedure, or an independently operating express or implied obligation may qualify the employer’s power to terminate. The task, therefore, is to identify the promise which the employee says was broken.

The institutions in which contractual claims may be brought

The institutional allocation follows from [484]⁠–⁠[529] above. Within their general monetary and subject-matter limits, the ordinary civil courts possess jurisdiction over claims for breach of an employment contract. That jurisdiction includes claims for inadequate notice or salary in lieu, unsupported summary dismissal, premature termination of a fixed-term promise where no lawful early route existed, breach of an incorporated dismissal procedure, and breach of another contractual term which caused independent loss. The appropriate level and availability of the civil forum will, of course, depend upon the relief and amount claimed.

The ECT, however, possesses only the contractual jurisdiction which the ECA schedules confer. The First Schedule includes specified contractual salary-related disputes, amongst them contractual salary in lieu of notice, while the Second Schedule includes the employee’s statutory entitlement under s 11(1) of the EA and the payment arising under s 16 where the employer breaks, or is deemed to break, the contract. That allocation does not turn the ECT into a general employment-contract court. A claim for damages arising from breach of an incorporated disciplinary code, an alleged implied term, or a promise of continued employment does not, for example, become an ECT claim merely because the employee seeks money as the remedy.

As to a contractual salary-related claim, the ECA supplies an accessible tribunal route without requiring that every contractual employment claim be brought there. Subject to the ordinary jurisdictional requirements, an employee may bring her civil claims together in the civil courts from the outset. If she instead elects to pursue a scheduled claim in the ECT, the claim limit, abandonment provisions, bar upon the same claim proceeding elsewhere, and ordinary doctrines of finality will operate according to their terms. Those propositions should be distinguished from my conclusion above that a claim under s 14(2) or s 84(2) was created together with, and remains bound to, the ECT route (see Goh Hui En Rebecca at [19]–[45] and [67]–[79]).

The typical content of the protection

I have discussed the relationship between contract and the statutory protection at [458]⁠–⁠[459] and [523]⁠–⁠[529] above. The starting point is the contract construed as a whole. From it, the court must identify the duration of employment, the express notice and payment-in-lieu provisions, any summary-dismissal clause, the remuneration and benefits promised, and whether a handbook, code, or policy was incorporated. Statute may nevertheless imply or control part of the bargain. Section 8 of the EA, for example, prevents a contractual term from providing conditions less favourable than those secured by the Act, while ss 9–11 regulate termination for contracts within their scope. Subject to anything in the contract to the contrary, s 16 also makes the party which breaks, or is deemed to break, the contract liable for a notice-equivalent sum. JIS treated that provision as implying an agreed-damages term into the contract (at [9]–[12]), though that decision does not determine whether s 16 exhausts every wider civil claim founded upon a distinct contractual promise and independently caused loss.

Employment for an indefinite duration will ordinarily be terminable by the agreed notice or, where the contract is silent and the EA applies, the statutory minimum notice. Section 10 permits either party to give notice at any time, while s 11(1) permits termination without waiting for that notice to expire upon payment of the prescribed sum in lieu. Where the contract itself confers an immediate power to terminate by payment in lieu, the valid exercise of that power ends the relationship in accordance with the bargain. Unless some other contractual term or statute requires a reason, each route ordinarily permits termination without proof of one.

Notice and payment in lieu should, however, not be treated as juridically identical in every contract. A notice keeps the contract in existence until it expires, unless the contract otherwise ends, and the employee ordinarily remains entitled and bound accordingly. An express power to terminate immediately upon payment in lieu, however, ends the contract at once and causes the payment obligation to accrue as a debt. Where no such power exists and the employer simply refuses further performance without notice, the contract is broken and the notice value ordinarily appears as damages. These differences may affect benefits, mitigation, and whether later-discovered misconduct can answer the monetary claim (Harvey, ch 8 at paras [421]–[423.01]; Phosagro Asia at [41]–[47]).

Garden leave differs from both. It ordinarily leaves the contract subsisting throughout the notice period while relieving the employee of the obligation to perform active duties, and the employer’s power to require it must arise from the contract properly construed. The arrangement is, as such, distinct from payment in lieu because the contractual relationship and its reciprocal incidents continue unless the contract provides otherwise (see Man Financial (S) Pte Ltd v Wong Bark Chuan David [2008] 1 SLR(R) 663; BGC Partners (Singapore) Ltd v Sumit Grover [2024] SGHC 206 at [95]). Any negative relief sought to enforce obligations during that period nevertheless remains exceptional and is subject to the ordinary equitable and restraint-of-trade principles.

A contract for a fixed period or specified work ordinarily expires upon the agreed event. If it contains no lawful route for early termination, premature termination may support damages measured by the unexpired promise, subject to the ordinary rules. Some caution is nevertheless required where ss 9–11 of the EA apply to the employee, since the statute permits notice at any time. Whether an apparently fixed term excludes, qualifies, or coexists with that statutory route is therefore a question of statutory operation and contractual construction, which should not be answered by importing the English common law position without analysis.

The relationship between summary dismissal, ss 11(2) and 14(1), and contract was examined at [113]⁠–⁠[116] and [559]⁠–⁠[562] above. Where no express term supplies a wider power, an employer may dismiss without notice at common law only in response to a repudiatory breach by the employee. Serious dishonesty, deliberate disobedience, grave insubordination, or other conduct going to the root of the relationship may suffice, while a minor breach, mistake, or ordinary negligence will not automatically do so. The question in every case is whether the employee’s breach, viewed in its context and consequences, was sufficiently serious to entitle the employer to accept the repudiation and terminate immediately (see Chandran (Lexis) at paras 6.116–6.124).

The parties may, however, agree upon greater job security than notice alone supplies by providing that dismissal will occur only for specified causes, after warnings, upon completion of an investigation or appeal, or by a particular decision-maker. Whether an employment handbook or disciplinary policy has contractual force is a question of construction. Language of promise and incorporation points towards legal obligation, whereas language of guidance, a retained power of unilateral variation, and an express statement that the policy is non-contractual point away from it. Further, an employee’s promise to obey policies does not necessarily produce a reciprocal promise that every policy binds the employer (see Aldabe Fermin at [90]–[116]; Long Kim Wing at [126]–[143] and [161]–[186]).

Any incorporated procedure must then be read with the express termination provisions. A disciplinary route may govern dismissal for misconduct while coexisting with a separate and unqualified right to terminate upon notice. Conversely, the contract may make completion of the procedure a condition which precedes every termination of the relevant kind. The answer must be found by construing the bargain as a whole. In doing so, the court should avoid a construction by which the incorporated code defeats an express power with which it cannot reasonably coexist, or by which the express notice clause empties a specific procedural promise of content.

Singapore law does not presently contain a settled appellate holding that every employment contract includes a term prohibiting conduct calculated or likely to destroy mutual trust and confidence without reasonable and proper cause. The Appellate Division made clear in Dong Wei that the Court of Appeal did not endorse such a term merely by reciting its formulation in an earlier case (at [69]–[82] and [87]–[93]). At first instance, Matthew Dabbs rejected the term (at [69]–[91]), while Prashant Mudgal accepted it (at [96]–[121]). That conflict remains unresolved at the appellate level. Even upon the accepting view, however, the term cannot contradict an express right to terminate upon notice, turn that right into one exercisable only for cause, or create a universal duty of fair investigation (see Prashant Mudgal at [226]–[255]).

Contractual constructive dismissal occurs where the employee accepts the employer’s repudiatory breach by resigning in response before affirming the contract. To establish it, the employee must identify the express or implied term breached, show that the breach was repudiatory, establish that it caused the resignation, and demonstrate acceptance, with each of those contractual elements being considered upon its own terms. The forced-resignation limb in s 2(1) poses the different statutory question examined at [599]⁠–⁠[604] above.

Finally, serious misconduct discovered only after termination may be relied upon to answer a contractual damages claim for wrongful summary dismissal (see Chandran (Lexis) at paras 6.130–6.131). The reason is that an employee cannot ordinarily recover upon the premise that continued contractual performance was owed where she had already committed a repudiatory breach which entitled the employer to terminate. Later discovery does not, however, retrospectively extinguish a payment in lieu which accrued as a debt through the valid exercise of an express termination power. Without a separate statutory analysis, the contractual doctrine likewise does not determine whether the dismissal was without just cause or excuse (Phosagro Asia at [41]–[64]).

The remedies that may be granted if the protection is engaged

As distinguished from statutory compensation, the contractual measure is expectation damages. For an indefinite contract terminable upon notice, the measure will commonly comprise the net salary and contractual benefits which would have accrued during the required notice period, less sums earned or reasonably obtainable through mitigation (see, eg, Harvey, ch 9 at paras [510]–[511]). Where the contract would have supplied them, the recoverable expectation may include allowances, leave, pension or Central Provident Fund (“CPF”) consequences, insurance, commission, or bonus rights. However, a discretionary or conditional benefit does not become unconditional merely because dismissal occurred (see Wee Kim San at [25] and [33]–[44]). Further, because this measure concerns damages, a payment in lieu which accrued as a debt through the valid exercise of an express termination power is not converted into damages and subjected to mitigation merely because both sums may resemble the notice value (see Phosagro Asia at [41]–[47]).

The minimum-obligation rule identifies the counterfactual lawful performance by asking what the employee would have received if the employer had performed through the least onerous route which it could and would have taken. That qualification matters because an employer cannot assume a termination power which did not exist, bypass a procedure which was a condition of lawful termination, or invoke a route which the evidence shows it would not have used. A binding procedure may therefore add the time reasonably required for its performance before the notice period could begin. An independently loss-causing breach before termination may also support a separate award if causation, remoteness, mitigation, and the avoidance of double recovery are satisfied (see Aldabe Fermin at [76]–[95] and [116]; Long Kim Wing at [176]–[186]).

The ordinary action does not compensate injured feelings, humiliation, reputational harm, or difficulty obtaining employment merely because a dismissal was harsh. Nor does it generally award damages upon the premise that employment would have continued indefinitely where the employer could lawfully have brought it to an end upon notice. Loss caused by a genuinely independent anterior breach remains analytically possible. Even then, the employee must show that it is not dismissal loss under another description and must prove its causal and financial consequences.

Specific performance and reinstatement are not ordinary private-law remedies for breach of an employment contract, since a court will ordinarily decline to compel the continuance of a personal relationship which requires cooperation and confidence (see Wee Kim San at [39]–[44]; Lim Tow Peng at [17] and [20]; Arokiasamy Joseph Clement Louis at [50]; Harvey on Industrial Relations and Employment Law (Ian Smith gen ed) (LexisNexis, Issue 312, January 2024) (“Harvey”), ch 9 at paras [521]–[523.01]). A declaration is likewise discretionary and must resolve a real controversy or serve some useful practical purpose; it is not granted merely as abstract vindication or as an indirect means of reinstatement. That is not to say that one will not find examples of courts effectively compelling the continued performance of an employment contract (see, eg, Jones v Gwent County Council [1992] IRLR 521, where the court granted a permanent injunction restraining the dismissal of an employee pursuant to a particular dismissal letter, and further granting interim injunctive relief restraining dismissal otherwise except where proper grounds existed and the proper contract procedures were followed). One will, however, find that these cases are exceptional rather than normal, especially where unremarkable employment contracts—without potent contractual protections against dismissal—are concerned.

The protection against dismissal under section 14(2)

For the reasons given at [449]⁠–⁠[467] above, s 14(2) requires an employer-initiated termination within s 2(1), including a qualifying forced resignation, to be supported by a factually established, operative, and sufficiently just cause or excuse. The protection regulates the merits of the employer’s exercise of its power to terminate and remains distinct from the contractual action.

The institution to which a claim under section 14(2) is bound

The exclusive character of the ECA route follows from [484]⁠–⁠[522] above. A claim under s 14(2) must originate through the mediation and claim route prescribed by the ECA and, if it remains unresolved, be adjudicated by the ECT. The ordinary civil courts do not possess concurrent original jurisdiction over that statutory cause merely because they possess jurisdiction over a contractual claim arising from the same dismissal. They may, however, receive an existing ECT proceeding through the statutory transfer power, and the General Division may, with leave, hear an appeal upon a question of law or jurisdiction in the manner prescribed by the ECA. Upon the construction adopted above, the receiving court determines the same statutory claim with the same authorised remedies and $20,000 or $30,000 limit. Transfer does not enlarge the cause, while any excess abandoned under s 15(3) cannot be revived in another tribunal or court. Neither route therefore gives rise to an independently originating civil action under s 14(2).

The ECT is a court constituted within the State Courts. Though its procedure is deliberately informal, ordinarily excludes legal representation, and permits the tribunal to inform itself without being bound by the ordinary rules of evidence, the ECT must decide according to law, have regard to the Guidelines, apply the statutory burden provisions, afford the parties a fair opportunity to answer the material upon which it relies, and remain within the remedies and monetary jurisdiction conferred upon it.

The personal scope of protection

By “personal scope”, I mean simply who may invoke the protection and against whom. For an ordinary claim under s 14(2), the employee must be a “relevant employee”, and s 14(2A) divides employees into three groups. First, an employee in a managerial or executive position who is dismissed with notice or salary in lieu is protected only if she has served the employer for at least six months in any position. Second, an employee in a managerial or executive position who is dismissed without notice and without salary in lieu is protected regardless of her length of service. Third, an employee outside a managerial or executive position is likewise protected irrespective of her tenure.

There are few authorities upon what it means to be employed in a managerial or executive position. The issue arose in Brightway Petrochemical Group Singapore Pte Ltd v Ang Lily [2007] 4 SLR(R) 729 (“Brightway”), where the employer contended that an employee engaged as an “Accountant” occupied a managerial, executive or confidential position and therefore fell outside the then definition of “employee” in the EA (see s 2(1) of the Employment Act (Cap 91) (1996 Rev Ed) (version in force from 1 October 2004 to 30 June 2007)). In support of that contention, it relied upon her designation and salary, her responsibility for setting up the company’s accounts, the fact that she was the only accountant in a growing business, and the likelihood that she had access to financial and business information (Brightway at [7]–[10]).

The High Court rejected that submission. Whether the employee’s position was “managerial, executive or confidential” had to be ascertained from all relevant circumstances, including her terms of employment, remuneration, designation, responsibilities, and qualifications. On the facts, the employee reported to the finance manager, performed a supportive role, had no supervisory functions or staff under her charge, and discharged duties which were predominantly administrative. Though setting up the company’s accounts formed part of her work, she likely did so under the finance manager’s direct supervision. There was also no evidence that she dealt with salary or budgetary matters, or had access to classified information or trade secrets (Brightway at [11]–[15]).

The cognate phrase “managerial or executive position” arose more recently in Hasan Shofiqul v China Civil (Singapore) Pte Ltd [2018] 5 SLR 511 (“Hasan Shofiqul”), in the context of Part 4 of the EA. The claimant had been promoted from construction worker to “site supervisor”. After his employment ended, he sought payment for work done upon rest days and public holidays, and that claim depended upon whether Part 4 applied to him. The Assistant Commissioner for Labour held that it did not because he was employed in an “executive” position. She relied upon his role as a site supervisor, responsibility for a team, day-to-day operational decisions, confirmation of safety compliance before works proceeded, influence upon workers’ appraisals, and recommendations to promote or dismiss workers (Hasan Shofiqul at [58], [66]–[67]).

The High Court disagreed on appeal. The title “site supervisor” could not, by itself, make the claimant an executive, just as some degree of supervision did not necessarily remove him from the category of “workman”. Much depended upon the nature and level of his supervisory powers and all the other relevant circumstances. His supervision was hands-on and on-site: he maintained bored pile records, conducted toolbox meetings, applied for daily permits to work, and assisted with communication and coordination between workers and management. Though important, those were routine on-site administrative and coordination functions which required no specialised expertise or training (Hasan Shofiqul at [68]–[69]).

The High Court also held that the claimant’s ability to provide feedback about the workers whom he supervised did not make him managerial or executive. Actual authority over hiring, dismissal, promotion, transfer, reward, and discipline remained with his superiors. The ability to give feedback which might influence a superior’s decision was different from exercising the decision-making power itself. The claimant was therefore not employed in an executive position. For completeness, the court added that no argument had been addressed to whether “manager” and “executive” were essentially the same concept for this purpose, but observed that the claimant was most unlikely to be a manager for substantially the same reasons (Hasan Shofiqul at [70]–[73]).

When read together, Brightway and Hasan Shofiqul establish that the inquiry is substantive and cannot be answered by title, designation, salary, or a general assertion that the employee was expected to take responsibility. Relevant considerations include the employee’s duties, the degree of independent judgment she could and did exercise, her reporting lines, whether she had staff under her charge, whether she possessed power over employment decisions, her qualifications, and her remuneration. However, those cases principally show what was insufficient and provide few positive examples from which one may confidently draw a line between managerial and non-managerial, or executive and non-executive, employees.

There is some further local material from a neighbouring context. In Kim Seah Teck Kim, “Eligibility of Employees to Join Rank-and-File Unions: The Ministry of Labour’s 1992 Guidelines” [1992] Singapore Journal of Legal Studies 496 (“Kim”), the author considered the then-new Ministry of Labour guidelines upon whether employees in “managerial, executive or confidential” positions should be eligible to join rank-and-file unions. Although that is a different question, the discussion is useful because it draws together IAC awards in which the same family of descriptions—“managerial, executive and confidential”—had to be applied in concrete employment settings. The inquiry in those awards was likewise substantive and concerned, in particular, whether representation in a rank-and-file union would place the employee in conflict between duty and interest (see Kim at 498–502, 508–512).

Upon Kim’s reading of those awards, designation was not decisive. Thus, in Singapore Manual & Mercantile Workers’ Union v Business System & Equipment Ltd (IAC Case No 144 of 1964), the IAC held that the terminology used to describe an employee’s position should not be treated as the ultimate criterion. Likewise, in Sime Darby Holdings Ltd v Sime Darby Holdings Local Executives Union (IAC Case No 73 of 1966), the IAC focused principally upon the duties and responsibilities of the employees concerned, the possibility of a conflict between duty and interest if their positions became representative and bargainable, and, to some extent, their salaries. As in Brightway and Hasan Shofiqul, the concern lay in what the employee actually did and where those functions placed her within the employer’s organisation (see Kim at 508–512).

The awards surveyed by Kim further suggest that supervision, technical responsibility, or some influence upon workplace matters was not, by itself, sufficient. In Pioneer Industries Employees’ Union v Far East Levingston Shipbuilding Ltd (IAC Case No 26 of 1971), chargehands who supervised 10 to 12 employees but performed broadly similar duties were not treated as managerial or executive. In Singapore Air Transport Workers’ Union v Saber Air Pte Ltd (IAC Case No 35 of 1971), supervisors responsible for ensuring that maintenance work was efficiently carried out were likewise not excluded from rank-and-file representation for that reason alone. Nor was it enough that employees might be consulted upon company policy or informed of management decisions before others, as E.M.I. (Singapore) Pte Ltd v Singapore Industrial Labour Organisation (IAC Case No 198 of 1973) illustrates (see Kim at 506–508).

The same substantive approach applied to employees in training or positions of possible advancement. In Evatt & Co v Singapore Manual & Mercantile Workers’ Union (IAC Case No 87 of 1973), trainee audit assistants remained eligible for rank-and-file representation while they were still in training, even though they might later enter higher, non-bargainable positions. Similarly, in Singapore Industrial Labour Organisation v Shinsei (Singapore) Pte Ltd (IAC Case No 62 of 1980), the IAC looked to present duties and responsibilities rather than future potential alone. Salary and fringe benefits were relevant, though not conclusive. Unsurprisingly, Kim characterises the IAC’s approach as one driven by common sense and the facts of each case, with less regard to labels than to whether the substance of the employee’s position made rank-and-file representation inappropriate (see Kim at 508–512).

These IAC awards reinforce the general approach in Brightway and Hasan Shofiqul, though they were concerned with a different question and should not be transposed uncritically to s 14(2A) of the EA. If handled with the necessary care, however, they probably provide useful data points by which parties addressing a s 14(2A) issue may chart a line of best fit between managerial and non-managerial, or executive and non-executive, employees.

Even where the requirements of s 14(2A) are satisfied, they do not exhaust the threshold conditions for bringing the claim in the ECT. Under s 12(2)(a) of the ECA read with reg 14(5) of the EC Regs, an ordinary s 14(2) claim may be made only by an employee who satisfies the prescribed connection and status requirements and who falls outside the exclusions. Broadly stated, she must have worked under a contract of service with an employer who, if an individual, resides or carries on business in Singapore or, if not an individual, is incorporated or formed under written law or carries on business in Singapore. She must also be a Singapore citizen, permanent resident, or foreigner entitled to work for that employer when the dispute arose in one of the ways recognised by reg 14(5)(b)(iii). Finally, she must not be a seafarer, domestic worker, a person excluded from the EA by Ministerial notification under para (d) of the EA definition of “employee”, or a public officer.

The material scope of protection

The protection is engaged only where there has been a “dismissal” within s 2(1). This requires either that the employer terminated the contract of service at its initiative, whether with or without notice and whether for cause or otherwise, or that the employee prove, upon the balance of probabilities, that her resignation was involuntary and was forced by the employer’s conduct, omission, or course of conduct or omissions. The expiry of a contract according to its own term is not necessarily a dismissal, though an employer’s decision or conduct which brings the relationship to an earlier end may be.

Once a dismissal is established, the question is whether it was “without just cause or excuse”. That question directs attention to substantive justification. The cause must have existed in fact, actually operated or borne materially upon the decision, and possessed sufficient weight and justice to warrant dismissal in the circumstances. A reason invented only after the event cannot become the historical cause, though evidence obtained later may prove a cause which genuinely existed and operated when the decision was made.

Misconduct, poor performance, genuine redundancy, incapacity, loss of a necessary qualification, or a real operational requirement may provide just cause or excuse. Conversely, discrimination, retaliation for exercising an employment right, the deliberate deprivation of a benefit, or a false ground may demonstrate its absence. Those examples are illustrative, and the statutory standard must still be applied to the facts irrespective of the label which either party attaches to them. Equally, the existence of some shortfall cannot establish justification without attention to its gravity and causal operation.

As I have explained, contractual validity does not answer the material inquiry. Though notice or salary in lieu may satisfy the contract and affect proof or remedy, s 2(1) expressly includes such a termination and s 14(3) authorises relief notwithstanding an agreement to the contrary. The Guidelines occupy a similar position. The ECT must consider their account genuinely, though their notice presumption, positive-wrongful-reason formulation, and categorical consequences cannot replace the enacted test where they differ from it.

The procedure needed to invoke section 14(2)

The employee must first submit the “wrongful dismissal dispute” to the approved mediation process no later than one month after the date of dismissal. If the dispute is not resolved, a claim referral certificate permits her to lodge the ECT claim through the prescribed process. This must be done within four weeks after the certificate is issued and in compliance with the ECA, the EC Regs, and the applicable procedural rules concerning form, fee, service, response, and attendance.

The claim must seek either reinstatement or compensation and remains subject to the applicable claim limit. At present, the ordinary limit for a wrongful dismissal claim is $20,000, increasing to $30,000 where the prescribed tripartite or union-assisted mediation condition is satisfied; a separate limit applies to a salary-related claim arising from the same employment relationship. An employee may abandon an excess amount to bring the claim within jurisdiction, though the abandonment has the extinguishing effect prescribed by the ECA.

The employee cannot simultaneously pursue an ECT claim and representations under s 35(3) of the IRA concerning the same dismissal, because the ECA makes those routes mutually exclusive in the manner which it prescribes. Proceedings elsewhere relating to the same ECT claim are likewise controlled by s 16, while s 17 supplies the means by which intertwined proceedings may be transferred to an appropriate court when its conditions are met.

The content of the protection and the proof required to engage it

The content of the protection, the allocation of proof, and the limits of that allocation, were examined at [90]⁠–⁠[171] above. The claimant must establish a dismissal within the personal and material scope of the protection and, subject to the statutory burden provisions, prove that it was without just cause or excuse. Section 27(2)(a) places the burden upon the employer where the employee was dismissed without notice under s 14(1), while para (b) does so where the employee was dismissed with notice and the notice is or purports to be given on the ground of poor performance or misconduct. An immediate termination upon payment of salary in lieu under s 11(1) is a termination without notice. It falls outside para (b) and, unless the dismissal was under s 14(1), outside para (a). The limits of both allocations, including the treatment of salary in lieu and reasons first advanced in the proceedings, remain as described at [119]⁠–⁠[170] above.

Outside those categories, the legal burden ordinarily rests upon the employee. Though that burden concerns the absence of sufficient justification, she may discharge it without direct access to the employer’s mind and without proving one positively prohibited motive. The statutory negative may be proved circumstantially by reference to the objective course of employment, the reason communicated or withheld, the timing and sequence of events, the work entrusted to her, the presence or absence of contemporaneous concerns, any inconsistency or comparative treatment, and the employer’s response to evidence calling for an explanation.

An evidential burden may move while the legal burden remains where Parliament placed it. If the employee adduces evidence which, if unanswered, supports the absence of just cause or excuse, ordinary reasoning may call upon the employer to produce the records and witnesses capable of identifying and substantiating the cause. Where the employer supplies a coherent and evidenced answer, the employee must meet it while retaining the ultimate risk of non-persuasion. Conversely, if material peculiarly available to the employer is withheld without adequate explanation, the ECT may draw the inference rationally supported by the circumstances. Silence, however, cannot be treated as automatic liability (see Yap Son On at [79]–[80]; Leong Kin Man v Chantrise Investment Pte Ltd [2021] 1 SLR 918 at [28] and [30]).

The ECT’s powers to identify issues, question the parties, and obtain existing material assist that inquiry. They do not create a general legal burden upon the employer where Parliament imposed one only in specified cases, nor do they relieve the claimant of establishing the claim which remains hers. The tribunal must disclose material upon which it may rely and afford a fair opportunity to answer it. Thus, the ECT’s freedom from technical evidence rules permits a practical inquiry, but it does not permit a decision by intuition or upon material which cannot rationally support the finding.

The employer’s cause should be examined as a composite proposition. The ECT must determine what the asserted ground means in the case, whether its material facts are proved, whether it actually bore upon the dismissal, and whether it was sufficiently serious and just to warrant that response. The falsity of a stated ground may provide strong evidence that there was no justification or that another reason operated, though it does not, as a matter of logic, disprove every possible cause. Equally, an uncommunicated cause cannot be accepted merely because it would have justified dismissal in the abstract. There must be proof that it existed and bore upon the decision.

As I explained at [545]⁠–⁠[558] above, process ordinarily bears upon the merits inquiry through its evidential effect. A genuine investigation, clear warning, opportunity to answer, performance support, or consultation may strengthen the evidence that the asserted ground existed and justified dismissal, while its absence may weaken that evidence. Due inquiry has the direct statutory force which s 14(1) gives it in the circumstances of that provision. JGP, however, held that the safeguard also applies where the employer relies upon misconduct despite dismissing with notice or salary in lieu (at [82]–[90] and Annex A at [A.134]). Since that remains a first-instance ECT holding, its extension beyond the express words of s 14(1) is contestable, and no reported appellate decision had resolved the point by 6 August 2026, I would confine the secure proposition to the evidential significance just described. Whether a procedural defect can independently make dismissal unjust under s 14(2) remains unresolved.

The remedies that may be granted if the protection is engaged

The remedial framework was examined at [172]⁠–⁠[222] and reconstructed at [472]⁠–⁠[474] above. Where the ECT is satisfied that the employee was dismissed without just cause or excuse, s 14(3) permits it to grant the remedy claimed: reinstatement in her former employment or compensation. The power is discretionary and is exercised notwithstanding any rule of law or agreement to the contrary. An employee should accordingly identify the remedy sought when commencing the claim, because the statutory form does not confer a general power to grant any order which appears vindicatory or fair.

Reinstatement restores the employment relationship and permits the ECT to direct payment of the wages which the employee would have earned had she not been dismissed. Though the legislation prescribes no detailed criteria for choosing that remedy, its practicality will ordinarily be affected by whether the position and work remain available, the parties’ ability to resume the relationship, the employee’s conduct, the operational consequences, the time elapsed, and whether restoration answers the established wrong. The common law’s reluctance to order specific performance may illuminate a practical difficulty, but it cannot render the statutory remedy unavailable in principle.

As to compensation, it is calculated in accordance with reg 17A and the Second Schedule to the EC Regs, subject to the applicable claim limit. It comprises compensation for loss of income and compensation for harm, with the former capped at three months of the claimant’s gross rate of pay at dismissal. The claimant must identify the income allegedly lost, establish the necessary causal connection with the wrongful dismissal, provide a principled basis for the period and amount, and account for mitigation and sums which would otherwise duplicate the same loss.

Compensation for harm, in turn, begins with a base amount not exceeding two months of the claimant’s gross rate of pay, which is then increased or reduced by up to 50% in accordance with the prescribed aggravating and mitigating considerations. The ECT must select the base and apply those adjustments by reference to the harm for which this component compensates, avoiding any double use of a matter already reflected in income loss or another adjustment. The regulations do not, however, answer every question concerning proof, causation, and the protected interest, and the uncertainties identified earlier remain.

Upon a contested s 14(2) claim, the ECT cannot compel an apology, agreed reference, correction of records, declaration, or other non-statutory form of vindication merely because the claimant seeks it. Though those terms may be achieved by agreement in mediation or settlement, the tribunal’s wrongful dismissal jurisdiction after adjudication is confined to reinstatement or compensation, together with the ancillary and procedural orders authorised by the ECA.

The protection against dismissal under section 84(2)

Section 84 supplies a pregnancy-specific protection with two connected consequences. Section 84(1) preserves maternity-related payments where a notice of dismissal given during the protected period was given without sufficient cause, while s 84(2) permits the qualifying female employee to seek reinstatement or compensation through the ECT where she considers that the notice was not given for sufficient cause. Though both consequences depend upon the same merits inquiry, the payment protection and the remedial claim remain legally distinct.

The institution to which a claim under section 84(2) is bound

The same exclusive route follows from [484]⁠–⁠[522] above. Like a claim under s 14(2), a claim for reinstatement or compensation under s 84(2) is bound to the mediation and ECT route created by the ECA. The ordinary civil courts do not possess concurrent original jurisdiction over that claim merely because a court may need to consider the sufficiency of cause when determining the distinct payment entitlement under s 84(1). Transfer and appeal, in turn, operate only in the manner authorised by the ECA.

The personal scope of protection

The relationship between the two EA-based dismissal protections, s 14(2) and s 84(2), was considered at [540]⁠–⁠[544] above. The personal scope of pregnancy-related dismissal claims requires closer explanation because the Third Schedule to the ECA contains two entries which both lead to s 84(2) of the EA. This results from Singapore’s division of the maternity-leave framework between two statutes. The EA contains the basic maternity-leave entitlement provided by law generally, while the CDCSA contains the enhanced Government-Paid Maternity Leave scheme. Since the protection accorded to a pregnant employee is enacted in s 84, the ECA must account for employees falling within either scheme. Item 2 of the Third Schedule therefore concerns employees whose maternity entitlement arises under the EA and who may bring a claim directly under s 84(2). Item 3 concerns employees whose maternity entitlement arises under the CDCSA but who may nevertheless invoke s 84(2) by virtue of s 12(1) of the CDCSA.

In greater detail, the two leave schemes operate as follows. Under the basic EA scheme, a female employee is broadly entitled to be away from work for 12 weeks around her confinement (s 76(1) of the EA). The paid portion is narrower. Ordinarily, she is paid for the first eight weeks if she satisfies the statutory conditions, including service of at least three months before confinement (ss 76(1A) and 76(2B) of the EA). The CDCSA provides a more generous scheme where the child is, or soon becomes, a Singapore citizen. In the ordinary case, where the child is a Singapore citizen at birth and the employee has served her employer for at least three months before confinement, she is entitled to 16 weeks’ paid maternity leave under the CDCSA (ss 9(1) and 9A(1) of the CDCSA). Similar provision is made where the child acquires Singapore citizenship within 12 months after birth (ss 9(1A), 9(1B) and 9A(1A) of the CDCSA).

The dismissal-related protection is, however, substantively the same in both cases. Once the employee falls within either the EA or the CDCSA maternity-leave scheme, she may, if dismissed, bring a claim under s 84(2) for reinstatement or compensation where she considers that notice was not given for “sufficient cause”. In the ordinary modern case—that is, one in which notice was given on or after 1 May 2013 (cf ss 84(1)(a) and (c), which deal with older transitional situations)—three basic requirements must be satisfied. The notice must have been given during her pregnancy; the pregnancy must have been certified by a medical practitioner before notice was given; and she must have served the employer for at least three months immediately before notice. To file the claim in the ECT, she must also satisfy the requirements in reg 14(3) of the EC Regs, which are similar to those applicable to a s 14(2) claim under reg 14(5) (see [646] above), save that the employee need not satisfy the “relevant employee” requirement in s 14(2A).

The material scope of protection

The material scope must be read with the analysis at [540]⁠–⁠[544] and [573]⁠–⁠[587] above. The personal-scope analysis identifies the pregnancy, certification, service, statutory-scheme, and prescribed connection requirements which open the protection. The separate material question concerns the event which s 84 treats as dismissal and the quality of the cause upon which the employer relies. Though the older circumstances in ss 84(1)(a) and (c) remain for their transitional operation, the ordinary current case arises under s 84(1)(b).

The merits question is whether the notice was “not given for sufficient cause”. Though this differs from the words “without just cause or excuse” in s 14(2), and the connection of s 84 with notice and maternity payments must be respected, I do not think that the two provisions prescribe materially different intensities of justification once their respective gateways are satisfied. Both formulations require the ECT to attend to the reason’s factual existence, operative role, weight, and justice. The reason must, as such, be established, operative, sufficiently weighty, and just in the circumstances.

For the reasons given in the analysis of the definition, I do not think that the older expression “a notice of dismissal given” should be treated as deliberately excluding the concept of forced resignation which was introduced into the EA’s general definition many years later. Thus, where the employee proves that her resignation during the protected period was involuntary and forced by the employer’s conduct or omission, s 84(2) should, in principle, be capable of applying. The linguistic seam nevertheless remains and would benefit from correction.

The procedure needed to invoke section 84(2)

The ECA route described at ⁠[651]–[653] above applies, albeit with a different statutory time limit. The employee must submit the dispute to mediation no later than two months after confinement, a period which should not be confused with the one-month period running from dismissal under s 14(2). If mediation does not resolve the matter, she must obtain a claim referral certificate and lodge the ECT claim within four weeks after it is issued, while complying with the applicable requirements of form, fee, service, and procedure.

The claim seeks either reinstatement or compensation and falls within the same wrongful dismissal claim limit of $20,000 or, where the prescribed condition is met, $30,000. However, eligibility under both ss 14(2) and 84(2) does not permit the employee to recover twice for the same loss or harm. Each cause and its burden must therefore be identified, and the resulting orders coordinated to prevent duplication.

The content of the protection and the proof required to engage it

The difference in burden noted at [540]⁠–⁠[544] above is express. Unlike the residual position which may arise under s 14(2), the ECA places the merits burden upon the employer in a pregnancy claim. Section 27(2)(c) applies to a claim arising directly under s 84(2), while para (d) applies where the CDCSA route engages that provision. In either case, the employer must prove the sufficient cause upon which it relies for giving notice.

The Guidelines and ordinary evidential principles assist in the same subordinate manner described for s 14(2). Though process may affect the reliability of the asserted cause, no general procedural code should be inserted into s 84 merely by analogy with s 14(1). The employer’s express legal burden also removes the difficulty of proving a negative which may otherwise confront an employee who was given no reason.

The remedies that may be granted if the protection is engaged

The remedial analysis at [172]⁠–⁠[222] and [472]⁠–⁠[474] applies subject to the terms of s 84. If the ECT is satisfied that notice was not given for sufficient cause, s 84(3) permits it to order the remedy claimed: reinstatement in the former employment with the wages which the employee would have earned but for dismissal, or compensation. Though the order is made notwithstanding any rule of law or agreement to the contrary, the distinct payment protection in s 84(1) remains separately enforceable, through the route assigned to it, where its conditions are satisfied.

Compensation, in turn, is calculated under reg 17A and the Second Schedule in the same manner as compensation for a s 14(2) wrongful dismissal claim, and remains subject to the applicable $20,000 or $30,000 limit. Loss of income and harm must accordingly be kept separate; causation and mitigation must be addressed; and duplication with maternity payments, notice remuneration, or another statutory award must be prevented.

The ECT has no general power under s 84(3) to grant an apology, declaration, reference, or other non-statutory order. Reinstatement or compensation is the adjudicative remedy, though wider terms may remain available by agreement in mediation or settlement.

The protection against dismissal under section 35(3)

As explained at [309]⁠–⁠[327] and [537]⁠–⁠[539] above, s 35(3) preserves the older route by which an eligible employee may invoke Ministerial intervention through her trade union within the industrial-relations framework of the IRA. It applies where she considers that she has been dismissed without just cause or excuse in circumstances falling outside a contravention of s 82.

The institution to which a claim under section 35(3) is bound

I traced the administrative function at [406]⁠–⁠[414] above. The decision-maker is the Minister, at whose written request the Commissioner may inquire and report. If the Minister decides to deal with the representations himself, s 35(5) requires him, before deciding, to give the employer an opportunity to make written representations explaining why it considered the dismissal justified. Though the function is administrative and informed by the industrial-relations setting, the Minister remains bound by the enacted criterion, procedural fairness, and the applicable public-law limits.

The ECT, the ordinary civil courts, and the IAC possess no original jurisdiction to determine representations under s 35(3). The IRA declares the Minister’s decision final and conclusive, subject to the constitutional supervisory jurisdiction which such language cannot wholly remove. A Ministerial direction also brings into operation the statutory bar upon a damages action in respect of the wrongful dismissal. As at 7 August 2026, however, the enacted Workplace Fairness amendment which would enlarge that bar has not commenced and forms no part of the present law.

The personal scope of protection

The union route follows from the settlement described at [316]⁠–⁠[327] and the boundary examined at [530]⁠–⁠[536] above. The employee must fall within the IRA’s definition of “employee” and proceed through her registered trade union. There is no personal filing route which bypasses the union, and the provision does not compel a union to transmit representations which it is unwilling or not entitled to make. A recognised trade union may, however, represent an executive employee individually for this purpose within the limited-representation provisions in s 30A, subject to the statutory treatment of an employer’s objection.

Section 35(3) contains neither a six-month service requirement equivalent to s 14(2A), nor an ECT monetary or status gateway. Its operation is, however, confined to circumstances outside a contravention of s 82. Where the complaint is that one of the protected trade-union circumstances caused the dismissal, the distinct route in ss 35(2) and 82, together with any separately available route under s 14(2), must be considered instead.

The material scope of protection

For the reasons given at [588]⁠–⁠[598] above, “dismissed” in s 35(3) includes an employer-initiated termination, whether with or without notice and whether purportedly for cause or otherwise. It also includes a resignation which the employee proves was involuntary and forced by the employer’s conduct, omission, or course of conduct. That conclusion derives from s 35(3) itself, though the definition in s 2(1) of the EA confirms and usefully expresses it.

The substantive question is, in turn, the same general question posed by s 14(2): whether the dismissal lacked just cause or excuse. The identical words, their common history, and the ECA’s treatment of the provisions as alternative routes strongly favour a coherent construction. Upon that construction, the employer’s actual ground must have existed, borne upon the decision, and possessed sufficient justice and weight. Contractual notice does not answer that inquiry, while a specifically prohibited or pretextual reason points towards the absence of justification.

However, a coherent substantive meaning does not make every incident of the Ministerial and ECT routes identical. The Guidelines, the burdens in s 27, the EC Regs, the claim limits, and the ECT compensation formula do not govern the Minister, while the administrative sequence brings the employer’s justification before the decision-maker in a different manner. Present law provides no answer as to how a published judicial construction and an unpublished Ministerial practice should be reconciled if they diverge.

The procedure needed to invoke section 35(3)

The procedure preserves the sequence described at [316]⁠–⁠[327] and [406]⁠–⁠[414] above. Within one month of dismissal, the employee must make written representations to the Minister through her trade union seeking reinstatement in her former employment. Since the provision states no power to extend that period, the union requirement, written form, relief sought, and time limit should be treated as statutory gateways whose operation does not depend upon ECT procedure.

The Minister may then request the Commissioner in writing to inquire whether the dismissal was without just cause or excuse and to report. However, if the Minister decides to deal with the representations himself, he must first give the employer an opportunity to make written representations explaining why it considered the dismissal justified. The Minister thereafter considers the union’s representations, any representations by the employer, and any report by the Commissioner before deciding whether the statutory criterion has been satisfied.

The route is mutually exclusive with an ECT wrongful dismissal claim in the manner now prescribed by ss 16(5) and (6) of the ECA. Thus, making representations under s 35(3) discontinues the corresponding existing ECT claim, while prior representations which have not been withdrawn or have been determined prevent the ECT claim. Sections 16(3) and (4) separately coordinate the specified claims for salary in lieu and dismissal payments arising from the same dismissal. Those provisions must be applied according to their precise terms and should not be treated as generally extinguishing every independent contractual right.

The content of the protection and the proof required to engage it

The contrast with the ECT burden rules at [119]⁠–⁠[171] and the institutional difference considered at [530]⁠–⁠[536] above matter here. The Minister must be satisfied that the employee was dismissed without just cause or excuse. The special burdens in the ECA do not apply, and the IRA states no general legal burden in the language of s 27. Its procedure nevertheless ensures that the employer’s asserted justification is elicited through the Commissioner’s inquiry or the written opportunity in s 35(5), after which the Minister evaluates that justification, the employee’s answer, and the surrounding industrial facts by reference to the statutory criterion.

The protection, it bears emphasis, extends beyond contractual breach, specifically prohibited motives, and the categories identified in the Guidelines. Misconduct, performance, redundancy, incapacity, or operational requirements may justify dismissal if established and sufficient, while victimisation falling within s 82 belongs to the separate route described below. The Minister may draw upon lawful industrial-relations policy and specialist experience, but cannot substitute an unexpressed policy for the words which Parliament enacted.

There is no published corpus from which one may discern how the Minister has allocated proof or treated marginal categories in practice. The safest present restatement must therefore be confined to the statutory sequence: the employee and her union invoke the protection; the employer’s justification is brought into view; and the Minister must be satisfied upon the material obtained that the dismissal lacked just cause or excuse. The more particular ECT law of burdens should not be transplanted in the absence of legislation.

The remedies that may be granted if the protection is engaged

The remedy continues the Ministerial structure described at [316]⁠–⁠[327] and differs from the ECT arrangement examined at [400]⁠–⁠[402] above. If satisfied, the Minister may direct the employer to reinstate the employee in her former employment and to pay the wages which she would have earned had she not been dismissed, or direct payment of such amount of wages as compensation as the Minister determines. The direction operates notwithstanding any rule of law or agreement to the contrary. Accordingly, although the representations seek reinstatement, the statutory power permits the Minister to select either restorative or monetary relief.

The IRA’s own enforcement and finality provisions apply without the ECT claim cap, three-month income ceiling, two-month harm base, or Second Schedule adjustment. The employer must comply, and a direction brings the current bar upon a damages action into operation. Under s 35(10), an employer which fails to comply with the Minister’s direction commits an offence punishable upon conviction by a fine not exceeding $10,000 or imprisonment for a term not exceeding 12 months or both. Where a directed sum remains unpaid after conviction, s 35(11) further permits the District Court to recover it as if it were a fine and requires the recovered amount to be paid to the employee. The uncommenced Workplace Fairness amendments do not alter that statement of the present law.

The protection against dismissal under section 35(2)

The narrower structure of ss 35(2) and 82, together with its relationship to the general protections, was examined at [537]⁠–⁠[539] above against the historical IAC setting at [261]⁠–⁠[308]. In circumstances arising from a contravention of s 82, s 35(2) removes the ordinary prohibition upon the IAC considering a trade dispute relating to dismissal or making an award concerning reinstatement. Section 82 supplies the substantive protection, while s 35(2) preserves the industrial institution which may adjudicate the resulting dispute.

The institution to which a claim under section 35(2) is bound

The institutional route follows from [261]⁠–⁠[308] and [537]⁠–⁠[539] above. When exercising the relevant powers, the IAC is constituted by its President alone. It must, however, first acquire cognizance of a trade dispute through one of the routes in s 31, which, in an ordinary victimisation dispute, will commonly be a joint written reference or a Ministerial direction published in the Gazette. Section 35(2) accordingly creates no individual claim form by which an employee may proceed directly to the IAC apart from the industrial-dispute machinery.

Section 82 also creates a distinct criminal route in the District Court. A contravention is an offence punishable upon conviction by a fine not exceeding $5,000 or imprisonment for a term not exceeding 12 months or both; a prosecution requires the Public Prosecutor’s prior written consent; and an authorised public officer may exercise the investigation powers conferred by s 84. Section 82(6), however, permits the IAC to inquire into dismissal or reinstatement before District Court proceedings commence. The routes must therefore remain separate. The IAC does not conduct a criminal prosecution, and the District Court does not determine an industrial award.

The personal scope of protection

The limited personal scope follows from the statutory boundary examined at [537]⁠–⁠[539] above. The protected person must be an “employee” within the IRA and must have suffered one of the forms of employer prejudice by reason of one of the circumstances enumerated in s 82(1). Since those circumstances extend to actual and proposed trade-union membership or office, the protection is not confined to a long-standing union member. Access to the IAC nevertheless depends upon a trade dispute entering the statutory machinery, and an individual cannot compel the IAC to assume cognizance merely by asserting victimisation.

The limited-representation provisions permit a recognised trade union to represent an executive employee individually in proceedings concerning dismissal or reinstatement arising from s 82, subject to the IRA. For other employees, the union and recognition setting of the trade dispute will ordinarily determine how the matter is advanced. No minimum-service or ECT status requirement applies.

The material scope of protection

Section 82(1) prohibits an employer from dismissing or threatening to dismiss an employee, injuring or threatening to injure her in employment, or altering or threatening to alter her position to her prejudice, by reason of any of seven circumstances. Those circumstances concern: actual or proposed office or membership in a trade union or applicant association; entitlement to a collective agreement or award; appearance or proposed appearance as a witness, or giving or proposed giving of evidence, in proceedings under the IRA; dissatisfaction with working conditions as a union member seeking improvement; membership in a union involved in a s 18 notice, statutory negotiations, or a notified trade dispute; specified absence for union duties where leave was unreasonably deferred or withheld; and specified absence to perform IAC panel duties after notice to the employer.

The words “by reason of” require a causal connection between the protected circumstance and the prejudicial act. The protected circumstance must therefore have operated in the employer’s decision; the mere temporal priority of union membership, dissatisfaction, absence, or participation is insufficient. A close sequence of recognition activity, selection concentrated upon organisers or union members, inconsistent explanations, and an implausible reorganisation may, however, support an inference of victimisation even in the absence of any direct admission.

Section 35(2) becomes relevant to dismissal and reinstatement only if the circumstances arise from the contravention. A dismissal which is generally harsh, mistaken, disproportionate, or unjustified does not enter the IAC through this exception unless the protected causal reason is present. Though such a dismissal may engage s 14(2), s 35(3), contract, or some other law, union membership alone does not engage s 35(2).

The procedure needed to invoke section 35(2)

The union or the parties must bring forward a trade dispute and secure the IAC’s cognizance under s 31. The statutory routes include a joint written reference, the specified unilateral references, a Ministerial direction published in the Gazette, and a Presidential proclamation in the circumstances which govern each route. Historically, victimisation disputes were commonly directed by the Minister. There is no personal application under s 35(2), express one-month filing period, or monetary claim limit.

Once seised, the IAC may inquire under its industrial powers and determine whether the dismissal arose from a contravention of s 82, while the President may use the Court’s statutory powers to obtain evidence and endeavour to settle the dispute. If criminal proceedings in the District Court have already commenced, however, s 82(6) no longer preserves a later IAC inquiry into the same alleged dismissal or reinstatement.

The content of the protection and the proof required to engage it

The party alleging victimisation must place material before the IAC capable of supporting the protected causal inference. Though direct evidence will often be unavailable, it is unnecessary. The timing of the union activity, the identity and pattern of employees selected, the employer’s knowledge, the treatment of comparable employees, the objective reality of an asserted redundancy or misconduct, and any change in explanation may together establish a prima facie case, which the employer may answer by proving a genuine and independent reason.

No provision imposes upon the employer in the IAC proceeding the reverse burden found in s 84(1) of the IRA. That subsection applies expressly to proceedings for an offence under s 82 or s 83 once the other necessary facts have been proved, and should not therefore be transplanted into an industrial reference. The older IAC awards treated the union alleging victimisation as bearing the burden of establishing it, while recognising that a close circumstantial pattern may suffice.

The historical awards illustrate both sides of that threshold. Dismissals concentrated upon organisers or members immediately after recognition activity, and unsupported by the employer’s asserted reorganisation, have justified an inference and reinstatement (see Singapore Industrial Labour Organisation v Kie Hock Shipping (1971) Pte Ltd (IAC Case D105/1977); Lea Hin Co v Singapore Industrial Labour Organisation [1976] SGIAC 4; Singapore Industrial Labour Organisation v National Aerated Water Co (IAC Case D90/1978)). Mere union status, the severity of the sanction, or suspicion has not, however, sufficed where misconduct, inefficiency, closure, or redundancy supplied a genuine explanation (see Food, Drinks and Allied Workers’ Union v Sorrento Pte Ltd (IAC Case D37/1970); Pioneer Industries Employees’ Union v Fiberglass Pte Ltd (IAC Case D80/1976)). The inquiry remains confined to the protected causation required by s 82, and general industrial fairness detached from that provision supplies no substitute.

The remedies that may be granted if the protection is engaged

Through the industrial route, the IAC exercises the award-making powers in ss 35 and 42 within the narrow jurisdiction preserved by s 35(2). Proven historical cases have resulted in reinstatement, restoration of continuity, and back wages, while the IAC’s settlement powers permit agreed terms. Though the present IRA prescribes no ECT-style claim cap or compensation formula, the award must remain within the trade dispute and the jurisdiction which s 35(2) preserves.

Under the criminal route, the District Court may, upon conviction, order payment of wages lost and direct reinstatement to the former or a similar position. If wages ordered under s 82(3) remain unpaid, s 82(4) makes them recoverable as if they were a fine and requires that they be paid to the employee. A failure to comply with a reinstatement direction is a further offence under s 82(5), punishable upon conviction by a fine not exceeding $5,000 or imprisonment for a term not exceeding 12 months or both. Those post-conviction powers should, however, be stated separately from the IAC’s award, of which they are not the source.

Enforcing both contractual and statutory protections

Where a dismissal engages several protections, the claimant must identify the source and cause of each, proceed in the institution assigned to it, observe the relevant elections and procedural bars, and coordinate relief so as to prevent duplicate recovery. As the pathways and restrictions examined at [487]⁠–⁠[539] above show, genuinely distinct causes may proceed in their proper settings consistently with finality and the rule against double recovery.

Overview of the institutional pathways and restrictions

Contractual notice and other civil rights belong to the ordinary courts except where the ECA schedules a salary-related claim within the ECT’s jurisdiction. Claims under ss 14(2) and 84(2) originate through TADM and the ECT; representations under s 35(3) proceed through the employee’s trade union to the Minister; and a victimisation dismissal within ss 35(2) and 82 reaches the IAC only through the trade-dispute machinery in the IRA, with a separate criminal route where a prosecution is instituted.

The restrictions operate at different levels. They begin with jurisdiction, which determines whether the institution can decide the cause at all. Sections 15–17 of the ECA then regulate excess amounts, duplicate proceedings, discontinuance, and transfer, while ss 16(5)–(6) require an election between the ECT and Ministerial routes for the same wrongful dismissal. Separately, estoppel, extended res judicata, abuse of process, and the rule against double recovery prevent a legitimate institutional division from becoming repeated or duplicative litigation.

Enforcing both protections only in the ECT

An eligible employee may bring both a statutory wrongful dismissal claim and a scheduled salary-related claim in the ECT, the ordinary example being a claim under s 14(2) together with one for contractual or statutory salary in lieu of notice. They are separate claims, with separate elements and separate $20,000 or $30,000 baskets. Thus, payment of notice may satisfy the salary claim while leaving the statutory merits claim open. Conversely, success under s 14(2) does not itself establish that notice pay remains due.

The tribunal must therefore identify which order answers which right. Notice remuneration responds to the contractual or statutory payment obligation, while reinstatement or compensation answers a dismissal without just cause or excuse under the wrongful dismissal scheme. Where both sums would replace the same income over the same period, the governing remedial provisions and the rule against double recovery must be applied, since the existence of distinct causes does not make every award cumulative.

Enforcing both protections by representations to the Minister

That route enforces the statutory protection in s 35(3), and does not encompass the whole contract. The employee’s written representations must seek reinstatement, must be made through her trade union, and must reach the Minister within one month. If the Minister is satisfied, the IRA provides for reinstatement with lost wages or wages compensation. A contractual claim for notice, an incorporated procedure, or another civil right remains juridically distinct.

The ECA nevertheless coordinates the routes more closely than that distinction alone might suggest. Sections 16(3)–(6) prevent or discontinue the specified ECT wrongful dismissal and dismissal-payment claims when representations under s 35(3) are made or determined in the prescribed circumstances. An employee who elects the Ministerial route must therefore examine those provisions before attempting to maintain an ECT salary or wrongful dismissal claim. Whether an independently framed contractual right survives is, in turn, subject to s 35(9) of the IRA, which bars an action for damages “in respect of the wrongful dismissal” after a determination under s 35(6), the precise bars in the ECA, and the ordinary doctrines of finality. Section 35(9) does not make clear whether “wrongful dismissal” refers only to a dismissal without just cause or excuse under s 35(3), or also encompasses the common-law action. The effect of a Ministerial order upon an independently framed contractual claim therefore remains uncertain, and I need not decide that question here.

Enforcing both protections in the civil courts is not permissible

An employee may place the whole of her properly justiciable contractual case before a civil court, together with any independently actionable tort or statutory civil cause within the court’s jurisdiction. She may not, however, detach the reinstatement-or-compensation claim under s 14(2) or s 84(2) from the ECA route and commence it as an ordinary civil action. The Legislature created those rights together with their institution, procedure, remedies, and monetary limits. Accordingly, where a civil court decides an incidental question or receives a transferred ECT proceeding, the original statutory allocation of the cause remains intact (see [517]⁠–⁠[522] above).

Enforcing both protections in the IAC is not permissible

I examined the confined IAC jurisdiction at [537]⁠–⁠[539] above. The IAC has no general jurisdiction to decide that an employee was dismissed without just cause or excuse, or to order reinstatement upon that basis. Indeed, s 35(2) provides that the IAC must not consider a dispute relating to dismissal or make an award relating to reinstatement except in circumstances arising from a contravention of s 82. The Ministerial protection in s 35(3), and the ECT protections in ss 14(2) and 84(2), cannot therefore be carried into the IAC merely by characterising the dismissal as a trade dispute.

The IAC may determine a contractual matter within a trade dispute where the IRA confers cognizance, including through the specified limited representation of executive employees, though that does not enlarge its dismissal jurisdiction beyond s 35(2). A unionised employee must accordingly distinguish protected victimisation under s 82 from a general complaint about justification and from a private contractual breach.

Enforcing the separate protections in separate institutions

The legitimacy of separating distinct causes follows from the allocations at [484]⁠–⁠[539], subject to the transfer mechanism examined at [517]⁠–⁠[522] above. A wider contractual claim outside the ECT schedules may, for example, remain for the civil courts even though a s 14(2) claim or scheduled notice-pay claim has been brought in the ECT. Similarly, a dismissal for protected trade-union activity may engage both the general statutory protection and the industrial protection in s 82, subject to the applicable bars and route. Such institutional separation reflects the Legislature’s allocation, and factual overlap does not erase it.

That separation remains subject to the ordinary controls. A claimant cannot divide an indivisible cause by relabelling it, withhold an issue for tactical relitigation, or recover the same loss twice. A final determination may create cause-of-action or issue estoppel, while extended res judicata and abuse of process address matters which ought to have been raised earlier. Where intertwined ECT and civil proceedings should be decided together, s 17 supplies the authorised transfer mechanism; upon the construction adopted above, however, it does not enlarge the transferred cause. Precise pleading, respect for jurisdiction, coordinated remedies, and the ordinary doctrines of finality are therefore the safeguards which make concurrent protection workable (see Goh Hui En Rebecca at [67]–[79]).

Some practical guidance for disputants

The preceding restatement has necessarily been technical because the protections engaged by dismissal are distributed across different sources and institutions. Disputants will ordinarily require a simpler account of what they should do when an employment relationship ends. I therefore offer the guidance which follows, without suggesting that it replaces either the law stated above or the facts of any particular case. Even so, I think that these elementary distinctions should prevent many claims and defences from beginning upon the wrong juridical footing.

For employees intending to initiate dismissal claims

An employee should begin by identifying the complaint which she wishes to make. If it is that the employer failed to provide the notice, salary in lieu, wages, or contractual benefits which were due, the complaint is contractual or concerns a salary-related entitlement. If it is that the employer dismissed her without just cause or excuse, the complaint is statutory and arises under s 14(2). A pregnant employee may also possess the distinct protection in s 84, while an eligible unionised employee may have the alternative route of representations to the Minister under s 35(3) of the IRA. Though these protections may arise from the same dismissal, they are not interchangeable. The employee should therefore identify each protection relied upon, the institution in which it may be enforced, and the applicable time and monetary limits before commencing her claim.

Because the applicable periods are short, the employee should act promptly. An ordinary dispute under s 14(2) must be submitted for mediation within one month after dismissal, whereas a dispute under s 84(2) must be submitted within two months after confinement. If mediation does not resolve either dispute, the ECT claim must be lodged within four weeks after the claim referral certificate is issued. The wrongful dismissal claim limit is ordinarily $20,000, increasing to $30,000 where the prescribed union or tripartite-mediation condition is met. Representations under s 35(3) must instead be made in writing to the Minister, through the employee’s trade union, within one month after dismissal, and must seek reinstatement. Section 35(2), by contrast, creates no personal claim form, because the IAC must first acquire cognizance of the trade dispute through the machinery of the IRA. These are only the principal limits and, as such, the employee should promptly confirm every gateway which applies to her claim.

For a claim under s 14(2), the employee should state as clearly as she can what happened, who brought the employment relationship to an end, what reason, if any, was communicated, and why she says that the dismissal lacked just cause or excuse. If she relies upon a positive wrongful reason, such as discrimination, retaliation, or an attempt to deprive her of a benefit, she should identify the facts from which that reason is said to follow. She should not, however, assume that a failure to prove that particular reason necessarily ends the general statutory inquiry. Depending upon the circumstances, confirmation, appraisals, remuneration decisions, the work entrusted to her, the absence of recorded concerns, the timing of events, and inconsistencies in the employer’s account may together bear upon whether any just cause or excuse existed.

The employee should also retain, by lawful means, the documents bearing upon those matters and provide a reasonably complete account when the dispute is mediated and, if necessary, adjudicated. She should identify the remedy which she actually seeks. Reinstatement and compensation are the remedies available for a claim under s 14(2), and compensation must be addressed separately as loss of income and harm, with evidence directed to each component. An apology, correction of records, reference, or agreed statement may form part of a consensual settlement, but the employee should not assume that the ECT can compel every form of vindication which she regards as desirable.

For employers facing dismissal claims

An employer should likewise keep the contractual and statutory questions distinct. A termination which complies with an express notice clause or s 11(1) may be contractually effective and leave no notice pay outstanding. Where the employee brings a statutory complaint, however, the employer must still identify the actual basis of its decision and answer that claim. Compliance with the notice obligation goes only to the separate contractual question.

There is no general rule which requires an employer to state a reason whenever it terminates employment with notice, but a reasons-based statutory inquiry remains necessary when a claim is brought. If the employer later relies upon misconduct, poor performance, redundancy, operational requirements, or suitability, it should be prepared to place before the mediator or the ECT the contemporaneous material and witnesses capable of showing what the concern was, whether it actually bore upon the decision, and why it justified dismissal. High-level assertions, accounts supplied only after the event, or evidence from persons who did not directly observe the material events may carry less weight than contemporaneous records and the testimony of those who made or informed the decision. Above all, both parties should identify the protection invoked before relying upon the familiar label “wrongful dismissal”.

My decision in this case

Having stated the juridical source, content, proof, and remedies of the protection, I return to the four issues identified at [31]⁠–⁠[32] above. They concern, in sequence, the legal burden applicable to this particular form of dismissal; Mr C’s positive allegation of mental-health discrimination or retaliation; the separate question whether the evidence nevertheless established the absence of just cause or excuse; and the relief which followed. I address them in that order because the second and third issues call for different findings, though much of the evidence bears upon both.

On whom did the burden of proof lie?

RPL brought Mr C’s employment to an immediate end by relying upon the Termination Clause and paying salary in lieu. RPL did not purport to exercise the disciplinary power in s 14(1), and its cessation letter communicated no disciplinary or performance decision. Paragraph (a) of s 27(2) was therefore inapplicable because Mr C was not dismissed under s 14(1), while para (b) was inapplicable because he was not dismissed with notice. The residual legal burden of proving that his dismissal was without just cause or excuse accordingly remained upon Mr C.

Mr C could discharge that burden by relying upon the objective course of his employment and the inferences which the evidence, considered cumulatively, properly supported. He was not confined to direct evidence of RPL’s internal deliberations or to the positively wrongful reasons identified in the Guidelines. If his evidence became sufficient to call for an answer, an evidential burden could fall upon RPL to meet it, though the ultimate risk of non-persuasion remained with Mr C. Thus, if the existence and absence of sufficient justification remained equally probable at the close of the case, his claim had to fail.

I applied the substantive inquiry in the form stated at [464]⁠–⁠[471] above. Since it was common ground that RPL dismissed Mr C, I had to identify the basis said to constitute cause or excuse, decide whether the facts comprising that basis existed and bore materially upon the decision, and determine whether the proved and operative basis was sufficiently weighty to justify dismissal in the circumstances. RPL described its basis in general terms as “business, suitability, and operational considerations”, while maintaining that it had terminated the contract “not for any particular reason”. Its compliance with the Termination Clause explained the manner in which the employment ended. For the reasons given at [90]⁠–⁠[117] above, however, the remaining statutory questions required their own answers.

Did the claimant prove his positive case of discrimination?

Mr C’s positive allegation rested principally upon the sequence summarised at [11]⁠–⁠[12] above. He disclosed his diagnosis in June 2025, was moved from the SB Project to HQ, continued to assist with Project G until the tender was submitted, and was dismissed on the following day without being given a reason. When considered with his confirmation, bonus, increment, and the absence of earlier formal performance action, that sequence was capable of raising a suspicion that RPL had responded adversely to his condition or its disclosure.

I accepted, however, RPL’s evidence that, after learning of the diagnosis, it sought medical information, arranged an engagement with Mr C, encouraged him to consider its counselling programme, honoured his medical leave, and permitted him to remain away until he was fit. When the second memorandum recommended shorter hours, RPL assigned him to work at HQ during regular hours. Those steps supplied a coherent explanation for the change in assignment, consistent with both the medical advice and RPL’s operational needs. In that setting, the evidence concerning the move to HQ did not make the discriminatory explanation more probable.

The other matters upon which Mr C relied did not carry the positive allegation further. Records upon the company laptop, if they corroborated the long hours which he said precipitated his condition, would have borne principally upon how that condition arose. Preparations for his replacement, if proved, would have shown that the dismissal was anticipated. The reason for the dismissal would still have had to be inferred from other evidence, and neither matter provided a proper basis for a discriminatory inference. In any event, an adverse inference concerning missing evidence could operate only upon an existing case in the manner explained at [165] above.

Accordingly, though the proximity of the disclosure, reassignment, tender submission, and dismissal required careful examination, the evidence did not make it more likely than not that Mr C’s condition or its disclosure caused the dismissal. I therefore rejected his positive case of discrimination and retaliation. As I explained at [37]⁠–⁠[38] and [104]⁠–⁠[106] above, however, that conclusion removed one possible route to liability and left the general statutory inquiry to be decided separately.

Was the claimant dismissed “without just cause or excuse”?

I therefore turned to the objective record as a whole. RPL confirmed Mr C after probation on 8 October 2024, paid him a bonus of $10,011 in December although he had worked for less than half of that year, and increased his salary from $9,000 to $9,150 with effect from 1 January 2025. Those decisions concerned an earlier period and could not establish that his work remained adequate until August 2025. They were nevertheless objective acts by which RPL had treated him as suitable for continued employment and deserving of reward, and therefore bore some weight when it later asserted a materially more serious deficiency.

The inference was reinforced by the events after June 2025, during which RPL continued to assign Mr C work upon Project G—including reinforced-concrete planning, preparation of the master schedule, and assistance with methodology and tender compilation—throughout the period immediately preceding the submission. That continued assignment did not determine the quality of his outputs, which had to be evaluated from the evidence. It nevertheless sat uneasily with a suggestion that, before the work was undertaken, his incapacity or unsuitability was already so serious and settled that dismissal had become justified. The tender criticisms therefore required proof and evaluation in their own right.

There was also little contemporaneous material from which I could evaluate whether serious concerns about Mr C’s performance or conduct had been put to him. I accepted that an employee of his seniority need not invariably be placed upon a formal performance-improvement plan, and that s 14(2) does not require the ECT to impose its own view of desirable human-resources practice. The absence of a warning or formal process therefore supplied no independent basis for liability. As explained at [469] and [659] above, however, process and contemporaneous records bear evidentially upon whether the asserted ground existed, was conveyed while correction remained possible, persisted despite correction, and actually informed the dismissal. Their relative absence therefore affected the weight which could be given to the explanation advanced after the event.

Taken together, the confirmation and remuneration decisions, the continued assignment of work, and the absence of material by which the later criticisms could be evaluated formed a cumulative record which, if unanswered, made it more likely that RPL dismissed Mr C without a sufficiently established and operative cause. The silence of the cessation letter and Mr C’s assertion that he had done nothing wrong formed only parts of that wider record. Cumulatively, the evidence was sufficient to call for an answer in the sense explained at [161]⁠–⁠[164] above. Once RPL identified performance, suitability, and operational considerations, the issue was whether that particular answer displaced the inference, while the legal burden remained upon Mr C.

RPL adduced relevant evidence in answer through Mr RW, who addressed the SB Project, and Mr RR, who presented RPL’s account of the later tender work, as described at [22] and [24]⁠–⁠[26] above. I accepted Mr RW’s direct evidence that, during the lean-concrete incident, Mr C left the site without adequately ensuring that the supply shortage and the follow-up needed for completion had been addressed. This was a genuine, though confined, shortfall in the performance of his responsibilities. Upon RPL’s evidence as a whole, I also considered it more likely than not that this incident, together with some wider concern about performance or suitability, formed part of the considerations which contributed to the eventual dismissal.

The tender criticisms stood upon a materially weaker evidential footing. RPL attributed the assignment, correction, and assessment of that work principally to Mr PM and its Technical Director, but called neither of them. Their evidence would have enabled me to determine what Mr C had been instructed to produce, the clarity and number of corrections given, the work returned after those corrections, the degree of any departure, and the alleged occasions upon which he raised his voice or refused advice. I remained mindful of the ECT’s deliberately informal process and gave Mr RR’s general account the weight which its source and content could bear. Even so, the material before me did not permit an independent assessment of the extent, persistence, and seriousness of the alleged tender deficiencies. Since Mr C’s case already called for an answer, the evidential consequence concerned the weight and evaluability of that answer, consistently with [164] and [471] above.

I therefore accepted that some concerns probably existed and that the lean-concrete shortfall probably bore upon RPL’s decision, but I could not find that the wider tender allegations had been established with the seriousness and persistence asserted. Their operative role was also left uncertain by RPL’s own position that the dismissal was “not for any particular reason” and that the criticisms were advanced principally to rebut discrimination. Nor did the general language of “business, suitability, and operational considerations” identify a separate reorganisation, redundancy, or other particular operational event capable of being tested. Giving RPL the managerial latitude described at [468]⁠–⁠[471] above, the proved and operative material therefore consisted of a confined shortfall and some concerns of uncertain extent. Against the objective employment record and continued assignment, that material was insufficiently weighty to make dismissal a just response.

Looking at the evidence as a whole, and continuing to place the legal burden upon Mr C, I was satisfied on the balance of probabilities that the absence of a sufficiently established, operative, and just cause or excuse was more probable than its existence. RPL’s evidence showed that the employment was attended by some concern, but the proved content and connection of that concern fell short of the gravity required to justify dismissal. Mr C accordingly discharged his burden under s 14(2), though he did so without proving the discriminatory motive upon which he had principally relied.

How was the “wrongful dismissal” to be remedied?

Mr C confined his claim for relief to compensation of $18,000 and the two forms of non-monetary relief described at [13]⁠–⁠[15] above, without seeking reinstatement. For the reasons explained at [185]⁠–⁠[222] and consolidated at [472]⁠–⁠[481] above, compensation had to be assessed separately as loss of income and harm under reg 17A and the Second Schedule to the EC Regs. His decision to confine the overall monetary claim to $18,000 did not permit those components to be merged or their different requirements to be left unaddressed.

I begin with compensation for harm. Upon the account stated at [478]⁠–⁠[479] above, this component recognises the non-income injury caused by the wrongful severance of the protected relationship, and the base must reflect the harm established or properly inferred before the prescribed adjustments are applied. I selected the maximum base of two months because the abrupt and unexplained severance, viewed against the course of employment described at [6]⁠–⁠[9] above, carried a substantial affront and disruption which could properly be inferred without proof of a recognised psychiatric injury. I did not treat either the rejected discrimination allegation or Mr C’s pre-existing diagnosis as an aggravating factor. Since his monthly salary was $9,150, the base was $18,300. The lean-concrete shortfall which I accepted was, however, a limited instance of poor performance which contributed to RPL’s decision. It engaged the mitigating consideration in para 7 of the Second Schedule to the EC Regs and justified a reduction of 0.75 month, leaving 1.25 months’ salary, or $11,437.50. I used the shortfall only for that adjustment and did not count it again when selecting the base.

I made no award for loss of income because, as explained at [476]⁠–⁠[477] above, this component required Mr C to establish the income which he would probably have earned if the statutory wrong had not occurred, the period over which it would have continued, and the causal connection between the wrong and the asserted shortfall, subject to mitigation and the avoidance of double recovery. His evidence that he made 47 applications, obtained two interviews, and remained unemployed for a period supported the fact that his job search was difficult. It did not establish the period for which his employment with RPL would probably have continued, or permit me to attribute any particular period of unemployment or reduced earnings to this dismissal rather than to the ordinary contingencies of obtaining another senior construction role. The neutral cessation letter was relevant because it contained no allegation shown to have impaired his prospects, though I did not treat stigma of that kind as a condition of recovery. Further, RPL had paid Mr C’s full salary for August, for which no overlapping income loss could be awarded. Since the evidence supplied no principled causal or quantitative basis for a period beyond that payment, Mr C did not discharge the burden upon this component.

I also declined to order the non-monetary relief because s 14(3) and the ECA confined the adjudicative remedies for this claim to reinstatement or compensation. An apology which stated that the dismissal was discriminatory would, moreover, have contradicted my finding, as stated at [738]. Though an apology, reference, or agreed statement could have formed part of a consensual resolution, I had no power after adjudication, absent RPL’s agreement, to compel either the apology or the proposed statement that the termination was without prejudice and that no negative reason would be disclosed.

Conclusion

These grounds have travelled a considerable distance because the parties’ cases brought together sources which perform different juridical work. For the reasons given at [60]⁠–⁠[118] and reconstructed at [451]⁠–⁠[481] above, s 14(2), read with the ECA, supplied the binding protection, while contract, the Guidelines, and the common law performed the separate and subordinate functions there described. A dismissal effected with notice or salary in lieu and without a stated reason must accordingly be decided by the approach set out at [464]⁠–⁠[471] and [654]⁠–⁠[659] above, including the applicable legal burden and the ordinary movement of an evidential burden.

Postscript: A necessary summary

I will nevertheless attempt a summary. The expression “wrongful dismissal” now performs more than one legal function in Singapore. At common law, it ordinarily describes a breach of the employment contract in connection with termination. Under the EA and ECA, it is also used for a claim arising from dismissal without just cause or excuse under s 14(2), and for the related pregnancy protection in s 84(2). Though contract and statute may apply to the same termination, the common label does not displace the different juridical sources, procedures, and remedies of those protections.

Because the protection in s 14(2) is statutory, the enacted words remain controlling. The ECT must have regard to the Guidelines, which may provide valuable illustrations and make earlier administrative experience accessible. Their legal force, however, is confined to that which the ECA gives them. It follows that their treatment of positive wrongful reasons and contractual notice must be reconciled with the enacted inquiry into just cause or excuse. Section 14(2) imposes an additional statutory restraint upon the employer’s exercise of the power to terminate, and a dismissal which is contractually effective may therefore still lack just cause or excuse. The inquiry concerns the employer’s actual ground: whether it existed, bore upon the decision, and was sufficient to justify dismissal in the circumstances. The familiar categories accordingly assist in applying the provision while leaving the statutory standard controlling.

The applicable legal burden depends upon the provisions governing the particular case. Where it rests upon an employee who was given no reason, she may rely upon circumstantial proof and the ordinary movement of an evidential burden. The ECT’s procedural powers assist that inquiry, but they do not reallocate the legal burden. Process bears upon the reliability and weight of the evidence, and acquires independent force where statute or contract gives it that force. The ECT must nevertheless decide whether the proved and operative reason provided substantive justification.

Once the statutory wrong is established, reinstatement or compensation must be addressed pursuant to the distinct remedial framework examined at [172]⁠–⁠[222] and reconstructed at [472]⁠–⁠[481] above. Salary in lieu belongs to the separate notice-pay position and must be credited where necessary to prevent double recovery. Loss of income concerns causation, proof, the appropriate counterfactual, and mitigation, while compensation for harm requires a principled base followed by the prescribed aggravating and mitigating considerations.

Historically, the words “without just cause or excuse” formed part of an industrial-relations settlement under which dismissal could be examined publicly upon its merits despite contractual notice, the employer’s reason could be elicited, and the resulting wrong could be answered through reinstatement or compensation. Singapore’s labour policy later came increasingly to pursue employment security across jobs through productivity, flexibility, skills, employability, redeployment, and re-employment, though the old words and their remedies remained in place. Thereafter, the common label allowed incidents of the notice-centred action and the later administrative conception to enter the statutory protection. The four connected failures examined at [440]⁠–⁠[450] above—the inherited settlement, the faded conception of industrial justice, the polyseme and unstable hybrid, and the incomplete institutional transfer—explain how the retained standard came into judicial adjudication without a settled account of its juridical source, proof, or remedy.

Although uncertainty remains at the boundaries between institutions, process, definitions, and concurrent enforcement, the reconstruction at [451]⁠–⁠[481] and the restatement at [605]⁠–⁠[723] show how the present law may be applied coherently. Some of that uncertainty can be managed judicially; the remainder would be placed upon a more secure footing by legislative attention.

I have attempted in these grounds to do no more than identify the law’s present condition, explain how it came to assume that form, and state the most coherent account which its enacted materials permit. Whether the remaining gaps should be filled, and whether the inherited words, institutions, and remedies should continue in their present arrangement, are ultimately matters for Parliament and the tripartite partners. Until they decide otherwise, however, the ECT must apply the law as it stands by preserving the distinction between statutory protection and contractual notice, confining industrial judgment within the enacted standard, and deciding each claim upon the evidence adduced and lawfully obtained.