Benjamin Ong's notes

Notes about Singapore public law, etc. for my students (and possibly others)

Author: admin

  • Happy Valentine’s Day

    In previous years, I sent students this message on or around 14 February:

    Today, the 14th of February, is the anniversary of the date on which Ronald Dworkin, philosopher of law, passed away in 2013. I cannot pretend to recall off the top of my head everything he said, let alone readily to understand it/its implications. Nonetheless, if I may, I would like to share a couple ideas of his (or ideas which I understand to be ideas of his). 

    First, there is the idea that law consists not only of rules, but also of principles. Dworkin articulated this idea in ‘The Model of Rules’ (1967) 35 University of Chicago Law Review 14. By his definitions, rules (such as “A will is invalid unless signed by three witnesses”) apply in an “all-or-nothing” manner (at 25), whereas principles (such as “no man may profit from his own wrong”: at 25) can be assigned a different amounts of weight in each particular case, and may pull in different directions (for examples, see 25-26). Moreover, principles can masquerade as rules: for example, there may be a ‘rule’ saying that “every contract in restraint of trade shall be void”, but the definition of “restraint of trade” contains the word “unreasonable”, which is a matter of principles and not rules (at 28). Dworkin says that we must not forget that principles, too, are part of the law.

    Whatever is the significance of this point for his theory on a grander scale, what I immediately drew from it is that we cannot expect law to be a simple process of applying rules to facts to come up with a conclusion, as though all we need to do to succeed is to know what the rules are. To the contrary, working with principles — including arguing about what principles are/aren’t relevant in a particular case, and what weight to assign to each principle — is part and parcel of a lawyer’s work.

    Second, there is the idea of law as interpretation (‘Law as Interpretation’ (1982) 9 Critical Inquiry 179) and the metaphor of the chain novel. Dworkin is talking about ‘interpretation’, not of a word or phrase, but rather of the entire body of law generally. He says this: “an interpretation of a piece of literature attempts to show which way or reading (or speaking or directing or acting) the text reveals it as the best work of art” (at 183). He goes on to describe a process in which a group of writers writes a novel chapter-by-chapter, each writer writing a chapter that builds on all of the previous chapters (at 192-193); in this process, each writer must interpret the previous chapters in order to write the next so as to participate in the enterprise of writing one novel. Dworkin then goes on to say that this is what judging consists of (at 193):

    “The similarity is most evident when judges consider and decide common-law cases; that is, when no statute figures centrally in the legal issue, and the argument turns on which rules or principles of law ‘underlie’ the related decisions of other judges in the past. Each judge is then like a novelist in the chain. He or she must read through what other judges in the past have written not simply to discover what these judges have said, or their state of mind when they said it, but to reach an opinion about what these judges have collectively done, in the way that each of our novelists formed an opinion about the collective novel so far written. Any judge forced to decide a lawsuit will find, if he looks in the appropriate books, records of many arguably similar cases decided over decades or even centuries past by many other judges of different styles and judicial and political philosophies, in periods of different orthodoxies of procedure and judicial convention. Each judge must regard himself, in deciding the new case before him, as a partner in a complex chain enterprise of which these innumerable decisions, structures, conventions, and practices are the history; it is his job to continue that history into the future through what he does on the day.”

    Sometimes, a lawyer’s job is to aid a judge in this process. At other times, it is to predict what a judge will say (and this applies no less to lawyers doing transactional work than to litigation lawyers). Either way, we will have to do more than recall what past cases have said (after all, judges have LawNet too); we will have to put forth arguments explaining which features of the existing body of law are relevant (or irrelevant), what sense to make of them, and, more generally, what way forward will cast the law in its “best possible light” (a phrase Dworkin used in his book Law’s Empire). (Come to think of it, was all this not what the Court of Appeal of England and Wales was explicitly trying to do in E?) If you are interested, read ‘Law as Interpretation’ for an account of the kind of issues that arise when one seeks to interpret the law in the manner just described.

    The 14th of February is also, according to some accounts, Valentine’s Day. Putting aside the dubious origins of the modern connotations of this ‘event’, I suppose it is worth remembering a couple of things. First, we must not confuse lust with love, and lust for the law with love for the law: to love is to grapple persistently with another’s flaws and weaknesses. Second, no matter how much we love the law the law cannot love us back; so is love for the law true love? Even as we pursue the law, let us not neglect our human loved ones, including ourselves.

  • Is English common law from before 12 November 1993 binding on any Singapore courts?

    Some of you asked: Does section 3(1) of the Application of English Law Act (AELA) mean that Singaporean courts other than the Court of Appeal are bound by rules of English common law that were in place before 12 November 1993?

    In my view, the answer must be no. No English decision is binding on any Singapore court. According to section 3(2), English common law before 12 November 1993 applies in Singapore only “so far as it is applicable to the circumstances of Singapore and its inhabitants and subject to such modifications as those circumstances may require”. Section 3(2), read plainly, does not place a limit on which court may make such “modifications”. It does not say, for example, that only the Court of Appeal can make such “modifications”. Neither does it elevate any English case law to the precedential status of decisions of any particular court in Singapore, such as the Court of Appeal. Therefore, the effect of s 3 is that any court – not just the Court of Appeal, but other courts, including, say, a District Court or Magistrate’s Court – is not bound by any English decision. 

    There is no contradiction between this proposition and the High Court’s statement in CHS CPO GmbH v Vikas Goel [2005] 3 SLR(R) 202 at [87] that Singapore law should not “depar[t] from the received English law merely for its own sake”. That latter statement means merely that Singapore law should not discard a useful legal rule merely because it is English in origin. That does not mean that Singapore law is obliged to keep that rule. In this regard, see also the observation of the Court of Appeal in Man Financial (S) Pte Ltd v Wong Bark Chuan David [2008] 1 SLR(R) 663 at [133] that “local courts must… recognise that the days of a uniform common law [across jurisdictions] are no longer a given. It is true that in the commercial context, there is more likelihood of (and desirability for) uniformity. However, even in the commercial setting, uniformity should not be taken too far.”

    One might next attempt to argue that, though any Singaporean court may decline to follow a pre-12 November 1993 English decision, there are limited reasons for which it may validly do so. One might, for example, point to the statement in s 3(2) that a court may decline to follow a pre-12 November 1993 English decision for reasons relating to the “circumstances of Singapore and its inhabitants”, and attempt to argue that only a difference of “local culture and mores” (to use a phrase from TQ v TR [2009] 2 SLR(R) 961 (CA) [52]) can justify treating a pre-12 November 1993 English case as non-binding. 

    In my view, this cannot be correct. The “circumstances of Singapore” includes not only socio-cultural circumstances, but also the fact that Singapore is a sovereign independent state with its own legal system. From the fact of Singapore’s independence, and the independence of Singapore’s legal system, it must follow that a Singapore judge is entitled to develop the common law, on the basis not only of social or cultural differences between Singapore and England, but also of his or her views as to “the cogency as well as logic of legal arguments and doctrines” as part of his or her “search for principle” (see Andrew Phang’s remarks in Andrew Phang, Goh Yihan, and Jerrold Soh, ‘The Development of Singapore Law: A Bicentennial Retrospective’ (2020) 32 SAcLJ 804, [136]).

    I am fortified in these views by the words of Andrew B L Phang and Goh Yihan, Contract Law in Singapore, 2nd ed (Walters Kluwer, 2021) at [50] (emphasis added): 

    “the local courts are free to either accept or reject the common law [of England] and equity. This may be done on the basis that it was not part of the corpus of Singapore law in the first place or that even if it were so part of the law of Singapore, section 3(2) furnished the necessary flexibility in any event.”

    If statements from case law were needed to illustrate this point, I would point to two decisions of our courts. The first is Tang Kin Hwa v Traditional Chinese Medicine Practitioners Board [2005] 4 SLR(R) 604. At [27]-[28], the High Court stated (emphasis added): 

    “English law… has now to be cultivated with an acute awareness of the soil in which it has been transplanted. It must also be closely scrutinised for appropriateness on a more general level – that of general persuasiveness in so far as logic and reasoning are concerned. This is the essence of the ideal of developing an autochthonous or indigenous legal system sensitive to the needs and mores of the society of which it is a part. […] It is therefore to be welcomed that English law is no longer accepted blindly. This is not to state that it has not served jurisdictions such as Singapore, even outstandingly well. But there ought to be departures where either local conditions and/or reason and logic dictate otherwise.” 

    The second is ACB v Thomson Medical Pte Ltd [2017] 1 SLR 918. There, the Court of Appeal declined to follow the decision of the House of Lords in Rookes v Barnard [1964] AC 1129, which formed part of the common law of England in place before 12 November 1993. The Court of Appeal described the rule laid down in the latter case as “no longer applicable to the circumstances of Singapore” (at [175]). In saying that, the Court of Appeal was referring to matters of legal principle (see [174] and the discussion leading up to it), and not to the culture of Singapore or the nature of Singaporean society.

    But, appeals to authority aside, there is a broader point of principle at stake. To believe that differences of culture aloneare a valid basis for declining to follow pre-12 November 1993 English law would simply not do justice to Singapore’s independence as a legal system. (After all, even English colonial authorities recognised that English law, as applied in the Straits Settlements, had to be “adapted to the condition and wants of the inhabitants” such as differences in religious beliefs and customs: see Choa Choon Neoh v Spottiswoode (1869) 12 Ky 216, 221.) But if one stopped there, and held that there Singapore law is otherwise manacled to English law, that would be no less than an act of neo-colonialism – and, worse, self-imposed neo-colonialism – which I cannot imagine was the intention of the Legislature when it passed section 3 of the AELA. I cannot imagine that any Singaporean judge would be willing to knowingly reach a decision whose reasoning he or she believes to be unprincipled, merely because similar reasoning has been applied in ‘mother England’ of which Singapore is but “some corner of a foreign field” (Rupert Brooke, ‘The Soldier’ (1914)). As we have seen, Phang JC (as he then was), for one, certainly refused to do so in Tang Kin Hwa. If I may respectfully say so, his approach appears to have been informed, not by picking at the wording of s 3 of the AELA, but rather simply by a recognition that he was a judge in a Singaporean court who owed a duty to none other than the Republic of Singapore and its law

  • Bias, apparent bias, and pre-judgment: some notes

    The ‘rule against bias’ is sometimes said to be embodied in the Latin phrase ‘nemo iudex in causa sua’, which, translated, means ‘nobody shall be a judge in his own cause’ — in other words, nobody with decision-making power shall act in a self-interested manner. However, the modern law on bias encompasses not only situations in which the decision-maker is self-interested, but also situations where, more generally, the decision-maker is unable to decide impartially. To put it another way, a decision-maker may be biased not only because of partiality toward one party, but also for reasons such as being biased in favour of decisions he has already made, etc. In a previous post, I made reference to the International Code of Ethics for Professional Accountants; the more I think about it, the more I see an overlap between those ethical rules and the law on procedural fairness.

    The definition of ‘bias’ that I have found most helpful is that provided by Peter Cane, Administrative Law, 5th ed (OUP, 2011) at 70-71: “A decision-maker is actually biased ‘if motivated by a desire to favour one side or disfavour the other’ [R v Gough [1993] AC 646, 659] ‘for reasons unconnected with the merits of the issue’ [R v Inner West London Coroner, ex p Dallaglio [1994] 4 All ER 139, 151], such as ‘prejudice, predilection or personal interest’ [R v Inner West London Coroner, ex p Dallaglio [1994] 4 All ER 139, 162].”

    This explains why pre-judgment can amount to actual, not just apparent, bias (Sim Yong Teng v SSC [2016] 2 SLR 489 (CA) at [50] and [52]). The problem with prejudgment is that it is, in substance, a “surrender by a decision-making body of its judgment” (R v SSE, ex parte Kirkstall Valley Campaign Ltd [1996] 3 All ER 304 at 317, cited in Sim Yong Teng at [50]). So the problem with prejudgment is not only that the decision-maker behaves in a manner that would give the impression that he has already made up his mind; rather, it is that the decision-maker has actually made up his mind, and is therefore incapable of making the decision on its merits. 

    Similarly, the problem with excessive interference with adversarial proceedings is not only that it may give the impression that the decision-maker is not impartial; rather, it is that the decision-maker will actually not be impartial. See Mohammed Ali bin Johari v PP [2008] 4 SLR(R) 1058 at [175(b)], cited in BOI v BOJ at [111]: ‘descending into the arena’ has the effect of “clouding [the decision-maker’s] vision and compromising his or her impartiality”. As the Court of Appeal put it in BOI v BOJ at [112], the problem with ‘descending into the arena’ is that the decision-maker may end up “slipping into the perils of self-persuasion”, hence “impair[ing] its ability to evaluate and weigh the case presented by each side”.

    In both of these situations, the actual bias is inferred from facts, rather than based on an outright admission by the decision-maker that it is biased/unable to make the decision impartially. Apparent bias, too, involves a process of inference from facts. The difference is that, in the case of actual bias, the law aims at rooting out an actual inability to decide impartially, and not merely the appearance of this; consequently, tests for actual bias do not make reference to public perception.

    What about the cases on automatic disqualification, which some textbooks refer to as ‘imputed bias’? Some take the view that the cases on automatic disqualification should be assimilated into the law on apparent bias. Harry Woolf, Jeffrey Jowell, Catherine Donnelly, and Ivan Hare, De Smith’s Judicial Review, 8th ed (Thomson Reuters, 2018) states at [10-019]: “it has been queried whether the doctrines of automatic disqualification and apparent bias are actually separate doctrines”. 

    The difference between the doctrines is that the cases on automatic disqualification (e.g. Dimes and Pinochet), when describing situations in which a decision-maker is disqualified from making a decision, make no reference to public perception. Of course, you are free to take the view that these cases are about what a fair-minded observer would think, and that they just lay down the rule that a fair-minded observer must think that a decision-maker is biased if he has a pecuniary or (sometimes) non-pecuniary interest in the outcome of the decision. Nonetheless, whatever you think is the theoretical foundation of these cases, the fact remains that, when applying cases such as Dimes and Pinochet, it is not necessary to make reference to a fair-minded observer etc.


    Are actual bias and apparent bias merely two points on a scale?

    It is not true that, where there is actual bias, there must necessarily be apparent bias. It is also not true that apparent bias and actual bias exist on one ‘scale’ on which apparent bias is ‘lower’ than actual bias. Rather, the two concepts are separate. This is because they aim at upholding different ideals. The e law on apparent bias aims to uphold the ideal that justice is seen to be done. (This remains so even though we may criticise the court’s views as to the type of ‘eyes’ through which decision-making processes are seen.) By contrast, the law on actual bias aims to uphold the ideal that justice is actually done. 

    It is true that, in some cases, justice is not seen to be done, which leads to a suspicion that justice has not been actually done. This kind of case is what would lead us to think that the difference between actual bias and apparent bias is merely one of evidence. In this kind of case, apparent bias is merely evidence that there may be actual bias. 

    But that is not the only possible kind of case. There may be cases where justice is not seen to be done even though it is actually done: in such a case, there would be no actual bias, but there would be apparent bias. Conversely, there can be a case in which just is not actually done even though a fair-minded observer would think that it is done: in such a case, there would be no apparent bias, but there would be actual bias.

  • Deference and illegality: a worked example

    It is true that, in Singapore law, the courts have the final say on the legality of the respondent’s action. What does the doctrine of deference have to do with this?

    Take, as an example, section 15(2)(a) of the Infectious Diseases Act, which provides that the Director of Medical Services “may order any person who is, or is suspected or continues to be suspected to be, a case or carrier or contact of an infectious disease, or who has recently recovered from or been treated for such disease, to remain and to be isolated and (if necessary) be treated, in his own dwelling place — for such period of time as may be necessary for the protection of the public”. In theory, that means that the Director has no power to order someone to be isolated for longer than is “necessary for the protection of the public”. If the Director did that, the Director would be acting ultra vires his powers.  (Note that the statute doesn’t say that the Director has the power to order isolation for, say, ‘a period of time which he thinks necessary for the protection of the public’.) 

    Now, suppose someone who had been ordered to be isolated for a certain period of time sought judicial review of the decision to isolate him, on the ground that the period of isolation was longer than “necessary for the protection of the public”. This would be review for illegality. In Craig’s model, “Y” would be “to isolate for a period of time that is necessary…” and the court would simply be performing its role of determining whether what the Director did falls within “Y” or not.So the court would have to decide whether the period of isolation was indeed longer than “necessary for the protection of the public”. That would be the court’s job. But how would the court know? 

    The answer would probably be that, though the court’s constitutional role extends to determining whether the period of isolation ordered was longer than “necessary for the protection of the public”, the limits of its institutional competence would warrant deferring to the Director’s view as to whether the period of isolation was “necessary for the protection of the public”. This is similar to how, in other areas of law, while the court has the power to make a finding of fact, the court may see fit to make that finding using expert evidence — see any medical negligence case for an example.

    That does not mean that the court would simply accept the Director’s view. The court would still retain the power to reject the Director’s view: for example, it would be open to the court to hold that the Director’s methodology was flawed, or that the Director was plain wrong in his assessment — but it would be difficult to prove this in practice.

    Now, suppose the statute said that the Director had the power to order isolation for “a period of time which he thinks necessary for the protection of the public”. In this case, to use the language of Craig’s model, “Y” would be “to isolate for a period of time which he thinks necessary…”, not “to isolate for a period of time that is necessary…” So it wouldn’t be possible for the court to review the decision to isolate on the ground that it isn’t “necessary…” However, the Director’s decision to isolate could still be reviewable on other grounds.

  • Some notes on review for error of fact

    1. It is uncontroversial that, if a statute stipulates a precedent fact requirement that has to be met before a person has the power to take some action, and that precedent fact requirement is not met, then any purported action taken under that statute is unlawful, and may be quashed by the court.
    2. It is controversial whether a court ought to have the power to quash executive action on the ground that the actor has, in deciding to take that action, committed an error of fact (other than an error of precedent fact). This is because:
      • the court may not be institutionally competent to judge whether the respondent has in fact committed an error of fact; and
      • the court may not be constitutionally competent to answer the question of fact, given that the statute often provides that it is the respondent and not the court which is to have the power to answer the question of fact.
    3. When, then, can a court intervene to correct an (alleged) error of fact?
    4. First, the court can find that the question of fact is a question of precedent fact, and therefore a question for the court to answer. Whether or not it is is a matter of statutory interpretation. So we may criticise the courts for applying what may appear to be a strained interpretation of the statute.
      • For example, there are cases such as A, which say that a question of fact must be a question of precedent fact because it is the sort of question which a court is capable of answering, viz. a question with a yes/no answer which can be answered using evidence of the sort that a court is institutionaly competent to assess. Fong Thin Choo (first analysis) seems to say the same thing: see [26], [29], [33]. But then we can ask: just because the court can answer the question, does that necessarily mean that the legislative intention was to let the court be the one to answer the question? 
    5. Second, if some other ground of review is made out that has caused the (alleged) error of fact, then the court can intervene on that ground. For example, if the respondent has allegedly ended up committing an error of fact because it has acted in a procedurally improper manner (e.g. failing to give the applicant a fair hearing), then the court can quash the respondent’s decision on the ground of procedural impropriety. We wouldn’t call this ‘judicial review on the ground of error of fact’; we could just call this ‘judicial review on the ground of procedural impropriety’.
      • We see this in Coleen Properties and in the ratio of Tameside.
      • This is one of the ways in which the court in Fong Thin Choo sought to justify its decision on the second analysis. The court said that the respondent had:
        • Committed an error of law, by misunderstanding the word “evidence” in the statute: [54]. (It is not clear to me, however, exactly what the respondent’s misunderstanding was.)
        • Committed procedural impropriety, by failing to “give a fair hearing to the applicants”: [54].
      • Sometimes, the court claims that it is merely engaging in judicial review on the ground that the respondent has “failed to take into account relevant considerations”, e.g. Fong Thin Choo at [57(c)]. The court says that a certain piece of evidence is a “consideration” which the court ought to have taken into account. However:
        • I think this approach is not correct. To my mind, in the law of relevant/irrelevant considerations, a “consideration” refers to a variable (e.g. wrongly taking into account the “consideration” of public anger, as in Venables; failing to take into account the “consideration” of potential clashes with other areas of law, as in Jessie Tan), not a piece of evidence.
        • Even if we do say that a piece of evidence is a ‘consideration’, the prevailing orthodoxy is that the court is not entitled to review the weight which the decision-maker gave to various considerations: Tesco Stores Ltd v Secretary of State for the Environment [1995] 1 WLR 759 at 780. Fong does not follow this orthodoxy: at [53], it states that the customs officer “placed undue weight” on certain pieces of evidence.
    6. Third, the court can hold that there is a separate ground of judicial review known as error of fact.
      • This must be what is happening in cases where the court gets to see evidence that wasn’t available to the decision-maker at the time. For example, E concerned evidence that became available after the respondent made its decision. So did Fong Thin Choo: at [50]. 
      • Tameside contains dicta that this ground exists. But it is not clear to me whether this ground formed part of the ratio of that case, or whether the court was in substance proceeding on the basis of another ground of review (para 5 above). I am fortified in this view by Timothy H Jones, ‘Mistake of fact in administrative law’ [1990] Public Law 507 at 513-514, which states that Tameside and similar cases do not make clear “whether… ‘an incorrect basis of fact’ constitutes an independent head of challenge to an administrative decision”.
      • The courts nonetheless see fit to hold that there are limits to this ground, e.g. the criteria in the E case; the criteria implicit in the Fong case. Our question is whether such limits adequately address the concerns of institutional/constitutional competence.
      • As for the concern of institutional competence, it is sometimes said in response that, while the Legislature intended that (a) it be the respondent and not the court who makes the finding of fact, the Legislature also intended that (b) the respondent not make an incorrect finding of fact. However, this view may be criticised as being somewhat artificial. While (a) is readily apparent from the face of the statute, it is unlikely that the Legislature ever even directed its mind to (b). As we will see later, this relates to a broader debate over whether the grounds of judicial review exist because the Legislature impliedly intends them to exist or for some other reason. 
    7. I think that the decision in Fong Thin Choo may be criticised for coming close to, or outrightly, crossing the line from reviewing the legality of the respondent’s decision into examining its merits. My personal view (with which you may or may not agree) is that there is nothing wrong with the court reviewing the merits of the respondent’s decision — this view of mine goes against orthodoxy, but I still hold it.
      • I am not alone in this view: see Swati Jhaveri, ‘Revisiting Taxonomies and Truisms in Administrative Law in Singapore’ [2019] 351 (her analysis of Fong Thin Choo is on page 365).
      • Even if we criticise Fong Thin Choo on the ground that it impermissibly intrudes into examining the merits of  the respondent’s decision, I would like to suggest that we ought to consider the extent of such intrusion. 
    8. E does not say that judicial review on the ground of error of fact is available in the event of ‘unfairness’. Rather, “unfairness… is the consequence of finding that the requirements [in E] are met” — in other words, ‘unfairness’ refers to the conclusion that the ground that is made out, not something that you need to prove in order to make out the ground: Chalfont St Peter Parish Council v Chiltern District Council [2014] EWCA Civ 1393, [100], cited in Christopher Forsyth, ‘Error of Fact Revisited: Waiting for the “Anisminic Moment”‘ (2018) 23 Judicial Review 1 at [13].
    9. If you are more interested in reading about E, try Paul Craig, ‘Judicial Review, Appeal and Factual Error’ [2004] Public Law 788.

  • The ‘would have made no difference’ argument

    There are cases in which courts have stated that, even if there was something unlawful about the respondent’s decision-making process, the court will refuse to quash the decision if the unlawfulness made no difference to the decision itself. For example:

    • In Lines International, the respondents had been influenced by an irrelevant consideration; yet the High Court refused to quash the respondents’ guidelines on this ground, because that irrelevant consideration had not been the “substantial reason for the implementation of the guidelines” (at [89]). 
    • In Shell (at 951g-j), the court had to address the issue of what happens if the respondent has exercised its powers “for more than one reason or purpose”. The Court of Appeal, citing R v Broadcasting Complaints Commission, ex parte Owen [1985] QB 1153 (EWCA) 1177, stated, inter alia, that: “where the court is quite satisfied that even though one reason may be bad in law, nevertheless the statutory body would have reached precisely the same decision on the other valid reasons, then this court will not interfere by way of judicial review.”
    • And in MCST Plan No 301 v Lee Tat Development Pte Ltd [2011] 1 SLR 998 (CA), a case which we will be studying as part of Topic 4, the Court of Appeal said at [59] that, even if a person is entitled to a fair hearing on certain matters but has not been given one, “if a hearing on those matters will not change the ultimate outcome of the case… then the hearing would be in vain and an exercise in futility. In such a situation, the court would be entitled to exercise its discretion not to grant the aggrieved party a hearing on the matters on which he was not heard…”

    I would like to point out that, in these cases, it is not necessarily easy for the respondent to persuade the court that the unlawfulness in the decision-making process made no difference to the ultimate decision. In fact, the courts have suggested that the tactical (even if not the legal) burden of proof is on the respondent to prove that the unlawfulness made no difference. I draw this inference (which I stated in Tuesday’s class) from the following:

    • In Lines International, what persuaded the High Court that the irrelevant consideration had not been the “substantial reason” for the decision as to the content of the guidelines was the respondents’ evidence (see [89]). 
    • In Owen (cited in Shell), the Court of Appeal stated that it will not quash the decision if it has been “satisfied” that “precisely the same decision” would have been reached (at 1177). 
    • In MCST 301, the Court of Appeal stated that the applicant has a “prima facie right to be heard on the matters on which he has not been heard” (at [59]). 
    • And in John v Rees [1970] Ch 345 (EWHC) (cited in SSHD v AF (No 3) [2010] 2 AC 269 (UKHL) at [60]), the High Court said:

    “It may be that there are some who would decry the importance which the courts attach to the observance of the rules of natural justice. ‘When something is obvious,’ they may say, ‘why force everybody to go through the tiresome waste of time involved in framing charges and giving an opportunity to be heard? The result is obvious from the start.’ Those who take this view do not, I think, do themselves justice. As everybody who has anything to do with the law well knows, the path of the law is strewn with examples of open and shut cases which, somehow, were not; of unanswerable charges which, in the event, were completely answered; of inexplicable conduct which was fully explained; of fixed and unalterable determinations that, by discussion, suffered a change…”

    This suggests that the ‘made no difference’ principle exists does not necessarily place the applicant at a disadvantage, because the respondent is the one who has to prove that the unlawfulness ‘made no difference’.

    Question: Does the ‘makes no difference’ principle fly in the face of the dignitarian rationale for fair hearings?

    I would like to begin by pointing out that, when courts deploy the ‘make no difference’ argument, they do so in the exercise of their discretion to refuse a remedy; it is not that there is a rigid rule forcing the courts not to grant an application for judicial review once it is shown that a hearing would make no difference.

    Academic literature aside, the idea that a fair hearing serves a dignitarian purpose is a relatively new idea in the (UK) case law, and has not been explicitly explored in the Singapore case law. I wonder whether, as dignitarianism is increasingly recognised as a rationale for procedural fairness, the ways in which the courts treat the ‘make no difference’ argument will change over time. (A cursory search on Lexis throws up no subsequent UK cases that explicitly cited AF (No 3) or Osborn for the proposition that upholding the dignity of the subject is a reason for upholding standards of procedural fairness.)

    Finally, I leave you with this memorable passage in this regard is from Lord Hoffmann’s judgment in AF (No 3), a case concerning “closed material procedures” following which certain persons suspected of terrorism could be placed under “control orders” on the basis of evidence which was not disclosed to them. Lord Hoffmann grudgingly agreed that the particular “closed material procedures” in question violated, inter alia, the human right to a fair hearing (set out in Article 6 of the ECHR) because they provided insufficient information about the case against potential controlees to give them a fair hearing. But he could not help but remark at [73]-[74] that:

     “It is true that a case which appears, on the basis of one side’s evidence, to be incapable of rebuttal can sometimes be destroyed. The remarks of Megarry J in John v Rees [1970] Ch 345, 402 about ‘unanswerable charges which, in the event, were completely answered’ is often cited. Most lawyers will have heard or read of or even experienced such cases but most will also know how rare they are. Usually, if evidence appears to an experienced tribunal to be irrefutable, it is not refuted.

     There are practical limits to the extent to which one can devise a procedure which carries no risk of a wrong decision. It is sometimes said that it is better for ten guilty men to be acquitted than for one innocent man to be convicted. Sometimes it is a hundred guilty men. The figures matter. A system of justice which allowed a thousand guilty men to go free for fear of convicting one innocent man might not adequately protect the public. Likewise, the fact in theory there is always some chance that the applicant might have been able to contradict closed evidence is not in my opinion a sufficient reason for saying, in effect, that control orders can never be made against dangerous people if the case against them is based ‘to a decisive degree’ upon material which cannot in the public interest be disclosed. This, however, is what we are now obliged to declare to be the law.”

  • Irrelevant considerations and improper purposes: why bother?

    Helen Fenwick (gen ed), Supperstone, Goudie & Walker: Judicial Review, 6th ed (2017), at [7.25.1], says of the law on relevant/irrelevant considerations and improper purposes: “It is difficult to deal with these aspects of the ambit of judicial review of discretion at a level which extends beyond platitudes.” What, then, is the point of studying the case law, if it all boils down to ‘it depends on the particular statute in question’?

    There are many areas of law which really boil down to ‘it depends’: the standard of care owed by one person to another; the interpretation of a particular phrase in a contract; the assessment of the credibility of a witness. Yet one does not merely learn the general principles — the “platitudes”; one studies particular examples. 

    The reason for this is that the examples show us how to frame our arguments before the courts. For example, Fong Thin Choo shows us how we can attempt to mount what is in substance an application for judicial on the ground of error of fact by casting the respondent’s conduct as an instance of having failed to take into account a relevant consideration; Jessie Tan shows us that ‘clashes with other areas of law’ can be a relevant consideration; and CBB demonstrates how we may infer from the statutory scheme what considerations are/aren’t relevant. This is similar to how, for example, NTUC Foodfare v SIA Engineering demonstrates to us how to deal with arguments that a claim for purely economic losses, if allowed, would open the ‘floodgates’.

    Another reason to study the cases is that their facts can provide useful analogies. If, for example, you have a situation in which you wish to allege that the decision-maker has failed to investigate some evidence which has newly emerged since the time he made his decision, Fong Thin Choo would be a possible go-to authority to demonstrate to a court that it can intervene on that ground. Similarly, Venables can be a useful authority to cite when we are dealing with a situation in which the respondent, in making a decision, has been influenced by public opinion. This is similar to how the Brinkibon case is a useful case to cite when dealing with a situation involving emails sent outside office hours.

    Yet another point is that the cases allow us to draw more general evaluative conclusions about the law. For example: Tesco Stores suggests that the law on irrelevant/mandatory relevant considerations, by leaving the question of the weight accorded to various considerations to the decision-maker, ensures that the court does not venture into territory that ought to be traversed by the political branches. But one could very well argue that this is not so, for the court’s very choice of what counts as a ‘consideration’ (or, in the law of improper purposes, a ‘purpose’) can itself be a smoke-screen for the court to become what is, in substance, a ‘mini-legislature’. Consider the notorious case of Roberts v Hopwood [1925] AC 578 (UKHL), in which a local authority had instituted a policy of paying all its workers of a certain grade the same wage. One might think that this is uncontroversial; but the House of Lords held that it was unlawful on the grounds that, inter alia, as Lord Atkinson put it, the local authority had “allowed themselves to be guided in preference by some eccentric principles of socialistic philanthropy, or by a feminist ambition to secure the equality of the sexes in the matter of wages in the world of labour” (at 594). Does Roberts v Hopwood illustrate that there is something unsatisfactory about the very idea of judicial review on the grounds of ‘irrelevant considerations’/improper purposes? Maybe (but what would replace it?) Maybe not (but does that mean that the outcome in that case is satisfactory?) Consider also the Westminster Corporation case, which demonstrates the difficulty of ascertaining the purpose for which the respondent has acted.

  • Miscellaneous notes on irrationality and proportionality

    The facts in Wednesbury

    My apologies: I had slightly mis-stated the background to Wednesbury in class. In particular, the condition imposed on the cinema was that children under 15 would not be admitted at all – even if their parents were with them. 

    For a more complete account of the background to the case, see Stephen Sedley, Lions under the Throne: Essays on the History of English Public Law (2015) at 23-24, which cites Michael Taggart, ‘Reinventing Administrative Law’, in Nicholas Bamforth and Peter Leyland (eds), Public Law in a Multi-Layered Constitution (2003) at 313ff.  Sedley writes that there was originally a ban on entertainment on Sundays, but:

    “by the 1930s cinema had become a mass entertainment medium. To enable picture houses to open on a Sunday, the one day when most working people were able to go, Parliament in 1932 passed the Sunday Entertainments Act, allowing local authorities to lift the ban on Sunday entertainments on such conditions as each authority thought fit to impose”.

    In the town of Wednesbury:

    “a Sabbatarian group secured a majority of seats on the local council. One of their aims was to halt the growing disregard of Sunday as a day of rest and prayer and the use of the Sabbath for secular entertainment. So, when the chain which owned the local Gaumont cinema applied to the council for permission to open on Sundays, the councilors granted the application (probably because they had been advised that they could not adopt a blanket policy of refusal), but made it a condition that the permission was not to include the admission of children under fifteen. For the sabbatarians this was a very satisfactory condition because it meant, in the days before universal television, that parents would not be able to go to the cinema on Sundays either, unless they were prepared to leave their children unsupervised in the house. Effectively the councilors had succeeded in keeping the Gaumont closed on Sundays.”


    Over the past couple of weeks [some time in 2019], we had some interesting discussions about irrationality (and proportionality) as a ground of review. I write to summarise some interesting points that emerged from the discussion.

    1. For short, we can use the phrase ‘Wednesbury principles’ (or, as Paul Craig puts it, Wednesbury in the ‘umbrella sense’) to refer to all the grounds of judicial review listed in Wednesbury itself, which include irrelevant considerations etc.; and we use the phrase ‘Wednesbury unreasonableness’ (which Paul Craig calls Wednesbury in the ‘substantive sense’) to refer to the specific concept of irrationality.
    2. One might think that purported examples of Wednesbury unreasonableness, such as the ‘red-haired teacher’ example, can be explained as examples of illegality (e.g. irrelevant considerations; an irresistible inference of improper purposes). This would reduce ‘irrationality’ to nothing more than a synonym for illegality, or a label applied to decisions which have been found to be illegal. But the courts have rejected this view: see the ‘one sense’ referred to in Wednesbury, and the pronouncement in Tan Seet Eng that irrationality is conceptually different from illegality.
    3. The difference between illegality and irrationality is summed up well in Daniel Tan’s article and Tan Seet Eng.
    4. Irrationality is a species of substantive review, which is, by definition, review of a decision for its substance. Does this go against the idea that judicial review is not supposed to be about the “merits” of a decision/about the “decision itself”? Or is this idea not correct — or, as Garreth Wong and Paul Craig argue, must it be seen with more nuance?
    5. Jowell and Lester argue that ‘unreasonable’ is not, in and of itself, a useful description: exactly what about a decision makes it ‘unreasonable’? Some of you thought that this is not a problem: it is right that Wednesbury exists as a ‘second safety net’, a ‘secret weapon’, or a ‘reserve’ that the court can tap onto to strike down executive actions. This sounds useful, but one could easily retort: exactly when is the court supposed to be able to pull out this secret weapon? etc.
    6. One argument against the proportionality test is that the court might end up substituting its decision for that of the executive. But, as Garreth Wong points out, this is not necessarily true. For example, the court could say ‘X is disproportionate’ rather than ‘X is disproportionate, so you should do Y instead’. The latter is substitution; the former is not.
    7. The proportionality test as applied in ECHR law (and summarised in Bank Mellat) requires that state action, among other things, (a) go no further than is necessary to achieve its aim; and (b) not outweigh the rights of those affected.
      1. This is different from what was applied in Hook, where the Court of Appeal seemed to pay little attention to the aim of the ban.
      2. What if the state action seeks to strike a balance between achieving multiple aims? This was the problem in Keyu. One might think that, in such cases, proportionality analysis should be done with reference to the fundamental rights of the persons affected. The problem in Keyu was that no such right was at stake.
      3. This does not mean that the courts thereby arrogate to themselves the power to replace executive/legislative decisions for any reason whatsoever. To the contrary, the courts apply a measure of deference to the elected branches of the state when applying the proportionality test. See Kavanagh’s article, as well as International Transport Roth GmbH v Secretary of State for the Home Department [2003] QB 728 (EWCA) [80], citing R v Director of Public Prosecutions, ex parte Kebiline [2000] 2 AC 326 (UKHL) 380-381: “In some circumstances it will be appropriate for the courts to recognize that there is an area of judgment within which the judiciary will defer, on democratic grounds, to the considered opinion of the elected body or person whose act or decision is said to be incompatible with the Convention… It will be easier for such an area of judgment to be recognised where the Convention itself requires a balance to be struck, much less so where the right is stated in terms which are unqualified.”
      4. Some would say that proportionality review and irrationality review are, in substance, similar/the same thing: Paul Craig; Lord Sumption and Lord Carnwath in Pham; Lord Kerr in Keyu.
      5. For all these reasons, it would be wrong to think of proportionality analysis as a ‘silver bullet’ that would greatly improve administrative law in Singapore. Maybe we already have it in practice. Maybe there are good reasons not to have it. Maybe its advantages are greatly exaggerated. etc.
    8. Thio Li-ann argues that the Singapore courts have not conclusively ruled out the concept of proportionality analysis in administrative law altogether, and that it is open to the Singapore courts to develop a proportionality test. I say “a” and not “the” because there are multiple possible flavours of proportionality – look, for example, at the difference between the EU law and ECHR tests. 
    9. David Mead critiques UK case law on substantive review on human rights grounds (i.e. examining the substance of executive action for compliance with human rights), and suggests that human rights be protected through review on grounds such as irrelevant considerations (i.e. asking whether the executive took human rights into account in the course of deciding to take such action). Consider the case of Vijaya Kumar in this light.
  • The President and ceremonies

    The President not only exercises duties set out in the Constitution, but also duties which involve the President acting as a symbol. I do not mean that the President’s symbolic role is in reality unimportant, nor that her role ought to be only that of a number-checking, bill-scrutinising administrator. To the contrary, presenting a good impression, on the domestic and international stages, by performing symbolic, ceremonial duties, and by interacting meaningfully with members of the local and international community, is very important. True, there is no way to quantify how much more has been raised through the President’s Challenge, or how much better regard other states have held us in, as a result of the President’s efforts. (Consider: why does the President’s Challenge exist, and why is it successful at raising funds even though many other charitable drives exist?  The answer, to my mind, is that the President is truly capable of operating as a unifying image, and that this image is truly of utility in serving the common good.) These duties are not even set out in the Constitution. But there is much that is important that is not set out in the Constitution, and there is much that is important that is unquantifiable; yet it is still important.

    If I may put it another, more colloquial, way: in my view, the President’s job may be described as involving a show, but not a mere show. To the extent that the President’s job involves a show that is meant to be seen – and it does – that show is an important show. It is as important as the show that consists of the judges wearing their ceremonial red robes at the Opening of the Legal Year ceremony or the Serjeant-at-Arms marching into Parliament with the mace. These shows serve a purpose, because they prompt people to respond in a certain manner. They prompt people to share in an air of solemnity surrounding proceedings that affect the lives of millions; and to pause and think about how those lives are being affected. These ceremonies involve officers dressing and taking part in processions in a manner unlike that of ordinary people because the roles they play are quite unlike those of ordinary people. 

    I would also venture to suggest that ceremonies are directed not only at the observing public, but also at the officials themselves.

    I suppose that some of this is outside the domain of the lawyer, and perhaps within the domain of the sociologist or anthropologist. Consider, for example, the body of scholarship about visual images of judges (http://blogs.bbk.ac.uk/research/2014/09/30/judicial-images/; https://www.jstor.org/stable/43862420). 

  • How a Bill becomes law

    For a quick overview of how a Bill becomes law, see this note on Parliament’s website: https://www.parliament.gov.sg/parliamentary-business/glossary/Details/passage-of-a-bill/Passage%20of%20a%20Bill. 

    For more details, including examples, note the following:

    Second Reading stage

    A Bill is debated at the Second Reading stage. The purpose of Second Reading debates is to discuss the “general merits and principle of the Bill” (Standing Order 70(2)). 

    Committee stage

    After that, the Bill goes through the Committee stage. The purpose of the Committee stage is “not [to] discuss the Principle of the Bill but only its details” (Standing Order 72(1)). 

    Usually, at the Committee stage, Parliament resolves itself into a Committee (or, in full, a “Committee of the whole Parliament”) on the Bill. The Bill is then scrutinised clause by clause and MPs can propose amendments to specific clauses. For an example of this, see the debates on the Miscellaneous Offences (Public Order and Nuisance) (Amendment) Bill in February 1996: https://sprs.parl.gov.sg/search/topic?reportid=011_19960227_S0003_T0007. At column 704 onwards, we see that during the Committee stage, Prof Walter Woon — then an NMP — proposed several amendments. Eventually Parliament did not vote in favour of any of these proposed amendments (“Amendment negatived.”), but the proposed amendments were debated.

    Instead of a Committee of the whole Parliament, Parliament may resolve to commit the Bill to a Select Committee on the Bill. This consists of a small number of MPs, which examines the Bill, may receive representations from the public, and eventually produces a report recommending changes to specific clauses, if any. The report is then placed before Parliament.  To access Select Committee reports, see here: https://sprs.parl.gov.sg/selectcommittee/home. The most recent Select Committee on a Bill to have issued its report appears to be the Select Committee on the Cross-Border Railways Bill. For an interesting read on the jury system, see the Report of the Select Committee on the Criminal Procedure Code (Amendment) Bill presented to Parliament on 9 December 1969.

    An aside: Select Committees that don’t pertain to Bills

    (You will notice that not all Select Committees pertain to Bills. Indeed, a Select Committee is simply, as the Parliament website puts it, “[a] group of Members selected by Parliament to inquire into and report on a particular subject” (https://www.parliament.gov.sg/parliamentary-business/glossary/Details/select-committees/Select%20Committees) — there are Select Committees on Bills, but also on other topics (e.g. the Standing Select Committee on Nominations for Appointment as Nominated Members of Parliament; the Select Committee on Deliberate Online Falsehoods).)

    Third Reading stage

    After the Committee stage comes the Third Reading stage, at which MPs may only propose “[a]mendments for the correction of errors or oversights”, and not “amendments of a material character” (Standing Order 81(2)). 

    PCMR

    For PCMR reports, go to the National Archives website (https://www.nas.gov.sg/archivesonline/search) and search for “Presidential Council for Minority Rights”, then click “Government Records”. 

  • Some notes on the Hinds case and judicial independence

    In Hinds, the Privy Council held that the legislation creating the Full Court Division of the Gun Court was unconstitutional. This is because that legislation would have required that certain cases, which fell within the exclusive jurisdiction of the Supreme Court, could only be heard by the Full Court Division and not the Supreme Court. This point was premised on a division of the Jamaican judiciary into the “higher judiciary” and the “lower judiciary” (218B). According to the Privy Council:

    • The Supreme Court was part of the “higher judiciary”.
    • Serious criminal cases must be heard only by the “higher judiciary” (221C).
    • Why? Because:
      • The “higher judiciary” had a “greater degree of security of tenure” than the “lower judiciary” (218F).
      • The more serious the matters at stake, the more strongly protected must be the independence of the judges who deal with these matters. For example, the reason why the “higher judiciary” had a higher degree of security of tenure was that the “importance of the jurisdiction that they exercise” was higher than that of the “lower judiciary” (219C). 

    One may ask: How does all this apply to Singapore, where, unlike in Jamaica, the Supreme Court does contain judicial officers who do not have security of tenure until age 65? This difference is indeed important, and I commend all those of you who have spotted it and raised questions about it.

    First, though, I would like to say a word about the phrase “security of tenure”. Sometimes, that phrase refers to the idea that, during a certain period of time, one cannot be removed from office (save in certain exceptional circumstances). Sometimes, that phrase refers to the idea that the aforesaid period of time must be very long. Hence, it is no contradiction in terms to say that the security of a Judicial Commissioner’s or Senior Judge’s tenure is protected (in that it is not easy for a JC/SJ to be removed from office: see Art 98(3)-(5)) although that tenure lasts only for a few years. So the fact remains that the security of a JC’s/SJ’s tenure is more strongly protected than that of a Magistrate’s or District Judge’s tenure in both senses: (a) it is harder to remove a JC/SJ during the period of his/her tenure; (b) a JC/SJ does have a fixed tenure in the first place. On hindsight, I should have been more precise in highlighting these distinctions in class; I apologise.

    That having been said, to my mind, the following points from Hinds are of enduring relevance to Singapore.

    First, the Privy Council appears to have assumed that one defining characteristic of a court is that there are in place measures to safeguard its independence. Sometimes the measures are watertight than in other cases. But that does not change the importance of independence. (See 213B.)

    Second, the Privy Council also stated the point, followed in Faizal,that the power to pass sentence according to law in an individual case is a judicial function. The Privy Council added, in a passage that wasn’t on the reading list (perhaps it should have been?), that such a function can only be exercised by one who has certain qualifications (226E, and see Faizal at [17]). Now, it is true that the importance of having certain qualifications was characteristic of judicial power in Jamaica because the Jamaican Constitution said so. But this did not stop Chan CJ in Faizal from drawing from Hinds the broader proposition that having legal qualifications is one of the “characteristics of a court” (at [17]).

    Third, there is the broad principle that some matters should be heard by judges whose independence is more strongly protected than others.  According to Hinds, there are certain cases which can only be heard by the Supreme Court and not by a subordinate court because Supreme Court judges have greater security of tenure. While, in Singapore, Supreme Court Judges have more strongly protected security of tenure than JCs/SJs, JCs/SJs do still have more strongly protected security of tenure than DJs/Magistrates. Note also that there are restrictions on discussions in Parliament of the conduct of JCs/SJs, whereas I can find no similar provision that applies to DJs/Magistrates.

    Fourth, there is the idea that, even if the Supreme Court does not have original jurisdiction in certain matters, it must still have supervisory jurisdiction over the courts which do have original jurisdiction. (If it did not, then there would be nothing from stopping the lower court from exceeding its jurisdiction!) In this regard, two commentators have, following Hinds,argued that decisions of the Military Court of Appeal must be amenable to judicial review by the High Court because, inter alia, the Military Court of Appeal’s independence is less strongly protected than the High Court’s is (Victor Leong Wai Meng and Roland Samosir, ‘Forever Immune?’ (1986) 28 Malaya Law Review 303, 311ff). (They argued that Abdul Wahab bin Sulaiman v Commandant, Tanglin Detention Barracks [1985-1986] SLR(R) 7 (HC) was wrongly decided. For reasons which I very briefly alluded to in class today, and which I will seek to justify during our later seminars on ouster clauses, I respectfully agree.)

    And, finally, there is Lord Diplock’s explanation that, because our constitutional order is “evolutionary not revolutionary”, we can interpret what it says about certain institutions by reference to the institutions from which they had ‘evolved’, and we can “tak[e] for granted that the basic principle of separation of powers will apply to the exercise of their respective functions by [the legislature, executive and judiciary]” (212C-E). In the next few weeks, we will see how the courts put the separation of powers in action, illustrating our points by reference to English jurisprudence. The aforementioned insight from Hinds explains why that English jurisprudence is relevant.

    I would also like to point out, on an unrelated note, that the Chief Justice has the final say on appointments of Magistrates and District Judges, and the Executive doesn’t have a say — see the State Courts Act, ss 9(1) and 10(1).

    A final thought: Are safeguarding judicial independence, and public confidence of judicial independence, matters to be pursued at all costs? For example, what would be the costs of requiring all judges in subordinate courts to have security of tenure for life? Would there be enough people willing to take up the job? Would they need to be paid more to be convinced to take up the job?

  • Some notes on the Liyanage case and judicial independence

    [I sent this note to students in a previous run of this course. It was a response to a student who asked: why, in Liyanage, did the Privy Council focus on the fact that, under the statute in question, the executive would have the power to nominate the judges who would preside over certain cases?]

    I certainly see the force in your (new) views: why would the specially-chosen judges be less independent than any judge?

    The Privy Council appears simply to have assumed that the power to nominate the judges who hear a particular case is part of the judicial power, without explaining why. Indeed, they didn’t need to, because, after the Minister’s power to nominate had been successfully challenged in the Supreme Court of Ceylon, the Minister’s nomination had been nullified by statute and replaced with a nomination by the Chief Justice (see [1967] AC 259 at 282, discussing the “second Act”).

    So let’s look at the judgment of the Supreme Court of Ceylon, R v Liyanage (1965) 67 New Law Reports 193 <https://www.lawnet.gov.lk/wp-content/uploads/2016/11/060-NLR-NLR-V-64-THE-QUEEN-v.-D.-J.-F.-D.-LIYANAGE-and-others.pdf>. Perhaps, on hindsight, we should have put the Supreme Court’s judgment on the reading list. (Perhaps my omission to discuss it indicates a sense of internalised neo-colonialism on my part, in the form of an over-eagerness to privilege the voice over the Privy Council over the voices of local judiciaries. Or perhaps I was merely misguided in my attempt to keep the readings brief.)

    The crux of the Supreme Court’s decision is at page 352 onwards. At first glance, it appears to rest on the slender point that the power to nominate judges to hear a case “appears to us to reside in the Court” (at 352) for historical reasons (last paragraph on 358). This is the same style of reasoning we see in 1: that ‘judicial power’ simply means ‘the powers which have traditionally been classified as being part of judicial power’.

    But I suspect that, although that’s as far as one needs to go when arguing a case in court, you won’t really be satisfied with that answer; and neither, personally, am I. So we can go on to look at the further points which the Supreme Court made at page 358:

    • Judicial power belongs to judges (see 357: “The right of a judge to exercise judicial power…”). Therefore, whoever has the power to nominate judges has the power to “prevent a particular judge or judges from exercising any part of the strictly judicial power vested in them by the Constitution”.
    • The Chief Justice is the “constitutionally appointed Head” of the court. A power of nomination would therefore create the possibility that the Chief Justice is prevented from acting as the head.

    Personally, I don’t agree with the first point: can’t it be said that the Chief Justice (who is normally in charge of nominating the judges) also has the power to never nominate Justice X, and therefore prevent Justice X from ever exercising judicial power? 

    To my mind, the strongest point is on page 359: “justice should not only be done, but should manifestly and undoubtedly be seen to be done” (quoting R v Sussex Justices, ex parte McCarthy). Specifically, said the Supreme Court:

    • The alleged offences consisted of a conspiracy to overthrow the Government. Therefore, a power vested in a member of that very Government to nominate the judges could raise the suspicion that “the ordinary or reasonable man would feel that this court itself may be biassed” (at 360). 
    • Also, people may think: “Why should the Minister, who must be deemed to be interested in the result of the case, be given the power to select the Judges whereas the other party to the cause has no say whatever in a selection?”
    • It may be that Parliament (which discusses the executive’s conduct) may end up discussing “the merits and demerits of respective Judges” in the course of discussing the minister’s nomination (eventually circumventing restrictions on discussion of judicial conduct in Parliament — for an example of such restrictions, see Art 99 of our Constitution). 

    To my mind, these points — the risk of a loss of public trust in the impartiality of the judiciary — may well be enough to justify holding unconstitutional the Minister’s power of nomination.  We can reach this conclusion without asking about whether the fact of being nominated creates even more fear than usual in the judge. 

  • Some notes on judicial independence

    How a Judge of the Supreme Court (etc.) may be removed

    It is not Parliament that has the power to initiate the process by which a Judge of the Supreme Court (etc.) may be removed from office. It is the Prime Minister or the Chief Justice who has the power to initiate the process. Please see Art 98(3)-(5) of the Constitution for details.

    What Parliament can do is pass a motion which will allow it to discuss the conduct of a Judge of the Supreme Court (etc.): see Art 99. While Parliament does not have the power to set in motion the process that can allow a judge to be removed from office, Parliament can still play a role in checking judicial power, in that Parliamentary discussion of judicial conduct, like any Parliamentary discussions, attract the glare of publicity. On the other hand, Art 99 itself functions as a check on Parliament’s power to discuss judicial conduct.

    Case in which a judge hearing a case had previously allegedly given related advice as Attorney-General

    In Yong Vui Kong v AG [2011] 2 SLR 1189 (CA) (which we will discuss later in the course), one issue was whether, in considering whether to exercise the power to grant clemency to a convicted offender (under Art 22P of the Constitution), the President could act in his sole discretion (as opposed to being obliged to act in accordance with the advice of Cabinet). 

    Yong’s counsel argued that Chan CJ ought to recuse himself from hearing the case, because, when he was the AG, he had allegedly given advice to the President on that very issue that the President had had no personal discretion in the matter, and therefore Chan CJ would not be able to decide the issue fairly because he might be unwilling to reach the conclusion that his previous advice had been wrong. 

    Chan CJ refused to recuse himself, but took pains to explain why. His reasons were that:

    1. Chan CJ had never, in fact, given such advice to the President. [18]
    2. Even if Chan CJ had given such advice to the President, it was correct. [19] Note that Chan CJ did not stop at asserting that he was correct (although he was); he took pains to point out that the other two judges in that case, as well as various other authorities, agreed that he was correct [19]-[20].

    The Tan Cheng Bock case

    In Tan Cheng Bock, the Supreme Court had asked the parties whether they objected to Menon CJ being one of the panel of judges hearing the appeal, given that Menon CJ had chaired the Constitutional Commission which recommended the legislative provisions and constitutional amendments whose interpretation was in question before the court. Evidently, the perceived potential threat to independence was the possibility that Menon had committed himself to a certain position, and therefore would be reluctant to decide to the contrary even if the law required that he decide to the contrary. The Supreme Court’s letter to the parties underscores that the Supreme Court was aware of the possibility of this perception. (See this Straits Times report: https://www.straitstimes.com/singapore/five-judges-to-hear-dr-tan-cheng-bocks-case-in-court-of-appeal, and the Facebook post by Tan linked to therein.) In the event, neither the AG nor Tan objected to Menon CJ’s hearing the appeal. 

    The Kho Jabing case

    In Kho Jabing v PP [2016] 3 SLR 1259 (CA), Kho’s counsel alleged that Phang JA ought not to have heard the Prosecution’s appeal against Kho’s sentence in 2015 because Phang JA had been one of the judges who had heard and dismissed Kho’s appeal against his conviction in 2011. The argument was that, in 2011, Phang JA had made up his mind, and would therefore now be hearing an appeal against his 2011 decision, which he would not be able to do fairly because he would be biased in favour of thinking that his 2011 decision was correct.

    The Court of Appeal unanimously rejected this argument for the following reason. In 2011, the decision involving Phang JA was a decision relating to Kho’s conviction. The present issue was that of Kho’s sentence. In 2011, Phang JA had decided nothing about Kho’s sentence; indeed, he could not have, because, at the time, the law only allowed for one sentence (viz. death) for murder (which was Kho’s offence) [5], and the courts simply had no choice as to what sentence to pass. So it could not be said that Phang JA was, in the 2015 hearing regarding sentencing, hearing an appeal against a decision which he himself had made in 2011. Rather, the decision in 2011 was about a totally different matter than the decision before the court in 2015.

    A possible way of thinking of threats to independence in a systematic manner

    I would like to suggest that the accounting profession has a potentially useful framework for thinking of threats to independence, which is set out in various parts of the International Code of Ethics for Professional Accountants promulgated by the International Ethics Standards Board for Accountants (IESBA) <https://www.ethicsboard.org/international-code-ethics-professional-accountants>. This might be useful because that profession considers “objectivity” to be one of its “fundamental principles”, and defines “objectivity” as not “compromis[ing] [one’s] professional or business judgment because of bias, conflict of interest or undue influence of others” (at [R112.1]). The analogy with judicial independence should be clear. At [120.6 A3], the Code sets out five types of “[t]hreats to compliance with the fundamental principles”, namely:

    “(a) Self-interest threat – the threat that a financial or other interest will inappropriately influence a professional accountant’s judgment or behavior;

    (b) Self-review threat – the threat that a professional accountant will not appropriately evaluate the results of a previous judgment made; or an activity performed by the accountant, or by another individual within the accountant’s firm or employing organization, on which the accountant will rely when forming a judgment as part of performing a current activity;

    (c) Advocacy threat – the threat that a professional accountant will promote a client’s or employing organization’s position to the point that the accountant’s objectivity is compromised;

    (d) Familiarity threat – the threat that due to a long or close relationship with a client, or employing organization, a professional accountant will be too sympathetic to their interests or too accepting of their work; and

    (e) Intimidation threat – the threat that a professional accountant will be deterred from acting objectively because of actual or perceived pressures, including attempts to exercise undue influence over the accountant.”

    It seems to me that, in Yong Vui Kong, Tan Cheng Bock, and Kho Jabing, the alleged/perceived potential threats to Chan CJ’s, Menon CJ’s, and Phang JA’s independence respectively fell into something analogous to category (b): the threat that, because they had previously considered a question and come up with an answer, they may be unduly unwilling to depart from that answer now. Perhaps we can think of some of the other various safeguards of judicial independence as being designed to combat some of the other types of threat to independence.

  • Note on the tort of harassment as a case study in the limits to judicial law-making

    In class, I mentioned the issue of whether a tort of harassment exists as a case study in the nature of (and limits to) judicial law-making. AXA Insurance Singapore Pte Ltd v Chandran s/o Natesan [2013] 4 SLR 545 (HC) sums up the issues at stake.

    In that case, Chandran repeatedly contacted AXA’s staff and used “vulgar and abusive language” against them [3]. AXA sued Chandran for the alleged “tort of harassment” [6]. Choo Han Teck J held that such a tort did not exist at common law [7], and refused to create such a tort. He said that, even if one thought that developing such a tort would be a “necessary step in the development of the law”, 

    “we are missing some steps which, even if I were to attempt to cover them, there would be too many, and beyond the scope of the court. I am of the view that a law against harassment must be delineated and legislated by Parliament… Having made reference to acts of harassment through legislation, it is up to the legislature to determine whether the law should be used to govern annoyance caused by means of letters, emails, and telephone messages, and whether the present public order law ought to be expanded to allow a claim for civil remedies. And it is for Parliament to determine what the nature and extent of such remedies should be.” [8]

    One reason for this was that law made by the Legislature has “gone through the legislative process of deliberation and debate by members of the legislature accountable to the public” [9]. By contrast, “[t]he court is not subject to that process of accountability and must therefore be restrained in the law-making process” [9]. If the court developed a tort of harassment in an incremental manner, that tort would not necessarily be fit for purpose — at [9] (hyperlink added): 

    “In this case, the defendant was described by counsel as harassing the plaintiff’s staff, but from the point of the defendant, he was pressing for his rights as a policy holder. We do not know who was right because that issue was not before the court. The plaintiff may be right, but there may be other employers unlike this plaintiff, who might rely on a loosely proclaimed law of harassment to oppress its weaker and poorer opponent. If the defendant in Malcomson did not send the baby rattle, and if the defendant in this case did not use abusive language, would their conduct still be regarded as harassment? Would a loosely recognised law of harassment be used to oppress others and avoid one’s legal obligations?” 

    Therefore (at [10]):

    “Expanding the current boundaries of torts against the person… should be done only where the rule can be formulated clearly, comprehensively, and concisely (such as the ‘neighbour principle’ in Donoghue v Stevenson [1932] AC 562). I doubt that a clear and comprehensive law on harassment as a civil cause of action can be effectively formulated in a judicial pronouncement, more so because there are, in modern times, calls for laws relating to privacy. Civil action in harassment and laws relating to privacy are complex and connected and must be considered together. Finally, by allowing litigants to sue when they feel harassed when there is no direct contact nor proof of damage, the court may be creating a blockbuster tort which will have unpredictable consequences, some of which may not be desirable. These are matters that need public debate to have the social, moral, and legal dimensions brought into the open. The forum for that is in the well of Parliament.”

    Eventually, the Legislature enacted the Protection from Harassment Act in 2014. Note that, in his second reading speech, the Minister for Law cited (at [59]-[60]) AXA v Chandran as one of the cases which highlighted the need for legislation to tackle the problem of harassment in society. Note also that, in 2019, the Legislature amended the Act following a process of consultation by the Ministry of Law (see MinLaw’s press release dated 1 April 2019). A court would not only have been unable to come up with such a comprehensive law; it would not have been able to take it upon itself to review how that law had operated and refine it, as the Legislature did.

    I note also the views of Paul Yowell, ‘Legislation, common law, and the virtue of clarity’, in Richard Ekins (ed), Modern Challenges to the Rule of Law (2011) 101, especially at part VI. This piece is challenging to read, but, in essence, Yowell argues that legislation can be clearer and more stable than case law because, inter alia:

    1. Legislation consists of “canonically formulated general rules of the kind produced by Western [sic] legislatures in codes and statute books” (121), whereas a common-law rule does not necessarily have a single authoritative formulation.
    2. The legislature has the power to “act on its own initiative and to change more than one law at a time”, which is useful given that “[m]any topics in the law are interconnected in complex, subtle ways, and desirable reform of the law often depends upon changing several laws simultaneously, sometimes in disparate areas… The legislature has the capacity for reforming the law in this way, but courts can introduce only incremental, piecemeal legal change, one law at a time. And because courts can act only in response to the initiative of others, their ability to change the law is dependent on the particular issues and subjects raised by the cases and fact patterns that happen to come before them.” (125)

    I get the sense that these issues are what Choo J had had in mind.

    This is not to say that the Judiciary’s proper role in law-making is uncontentious. For instance, Professor Goh Yihan, in ‘The Case for Legislating Harassment in Singapore‘ (2014) 26 SAcLJ 68 at [35], criticised Choo J’s decision on the ground that, “if one were to take [it] to its logical conclusion, then it would render the courts powerless in the face of changed social conditions”. With respect to Professor Goh, perhaps one could defend Choo J’s decision that it did not mean to say that the courts cannot take into account changes in social conditions, but rather that the courts cannot respond to such changes in a manner other than one which is incremental — and piecemeal — in nature. Indeed, as Professor Goh recognised, the case law, taken together, furnished a definition of harassment that was not “precise” [37] and was not clear as to what the remedies for harassment would be [40]. Moreover, if the law of harassment were left to case law, then it would not necessarily be clear to victims that they had a right to sue in tort and how to go about exercising this right [41]-[42]. These points explain why, despite Professor Goh’s criticisms of the court, he ultimately concluded that “the time has come to legislate against harassment in Singapore” [71].

  • How were NCMPs selected following the 2020 general election?

    [I wrote this note in response to someone who asked about who the NCMPs would be following the 2020 general election. I hope it makes some of the statutory provisions clearer.]

    The total number of NCMPs is 12 minus the total number of elected opposition MPs. So there will be 2 NCMPs. (section 52(1))

     How are they chosen? First, candidates who 

    1.  did not manage to get elected (section 52(2))
    2. are not from the political party that forms the Government (i.e. the PAP) (section 52(2))
    3. got at least 15% of the votes in the division (section 52(3A)(a))

    are ranked according to the percentage of votes they received (section 52(2)). In the case of a group candidates in a GRC, the ‘percentage of votes’ for each individual candidate is taken to be the percentage of votes received by the group (section 52(3D)).

    According to this system, the top 5 who qualify to be NCMP are:

    • KHOO POH TIONG JEFFREY (PSP): 48.31%
    • LEONG MUN WAI (PSP): 48.31%
    • N. LOGANATHAN (PSP): 48.31%
    • HAZEL POA (PSP): 48.31%
    • TAN CHENG BOCK (PSP): 48.31%

    They are tied. How is the tie to be broken and 2 chosen from the 5? Since they all come from the same group, the Returning Officer must tell the group, who then have 7 days to select 2 of themselves to be NCMPs. This is because of section 52(3B)(a). 

    It would be different if there were more than 2 NCMP seats up for grabs. In that case, we would have to consider the rule that not more than 2 NCMPs can be drawn from the same group of candidates in a GRC (section 52(3A)(c)). This does not apply.

    You mentioned section 52(3A)(c). In effect, that says that, in the case of a constituency “that is not a group representation constituency” (i.e. an SMC), no more than 1 of the candidates from that constituency can be NCMP. This does not apply to our present situation. It would apply in the following hypothetical situation: Imagine that, in an SMC, there are 3 candidates: one from PAP, one from Party X, and one from Party Y. The PAP candidate wins. According to the ranking system described above, the Party X candidate ranks first and the Party Y candidate ranks second. In this case, the Party X candidate would become an NCMP; and section 52(3A)(c) would prevent the Party Y candidate from becoming an NCMP. 

    My understanding is confirmed by the speech of Mr K Shanmugam, Minister for Law, in introducing in Parliament certain amendments to the NCMP scheme in 2010 (see https://sprs.parl.gov.sg/search/topic?reportid=019_20100427_S0004_T0012 At column 233). He said that the amendments would have the following effect: “There will also be a cap of two NCMPs coming from any one Group Representation Constituency (GRC), and a cap of one from any Single Member Constituency (SMC). This will spread out the NCMPs more evenly and make them more representative of voters who had voted for the Opposition nationwide in the General Election. In addition, the cap on NCMPs coming from a GRC will clearly distinguish between the winning and losing teams in a GRC.” (emphasis added)

  • ‘Deference’ and ‘non-justiciability’

    The concept of non-justiciability (as it has traditionally been understood in the UK) has apparently fallen into desuetude in Singapore, as Tan Seet Eng at [106] suggests. Recall also that, in Colin Chan, the Court of Appeal said that the proportionality of the decision was non-justiciable, but the rationality was not non-justiciable.

    I am fortified in this view by comments by Menon CJ at the hearing of Nagaenthran last week [in January 2019]. When counsel for the AG described a certain manner as “non-justiciable”, Menon CJ retorted with something like ‘All that means is that we can review less intensively’. This being so, I believe that the concept of non-justiciability has collapsed into the concept of deference, and that the language of non-justiciability has nearly completely fallen out of favour. The Court of Appeal’s judgment in Nagaenthran, when handed down, will confirm whether I am right in saying so.

    [I was. In Nagaenthran, the Court of Appeal said that the “merits of PP’s determination… is not one that is capable of being adjudicated upon by a court of law” (at [59]), then later it used the word “non-justiciable”. You may ask: does that mean that if something is not non-justiciable, the courts can consider its merits? In response, I would refer to Garreth Wong’s and Paul Craig’s distinction between the court considering the merits of a decision, and substituting its decision for the executive’s. I suppose that, when it comes to a “non-justiciable” matter, the court may be less competent to even consider the merits of a decision. This might mean, for example, that the standard of substantive review is lowered (i.e. easier for the executive), or even that substantive review is not possible at all.

    But even before that case, consider Tan Seet Eng at [98(a)], where the court appears to have been using “justiciability” in a loose sense to refer to the degree of deference to be shown to the executive’s better access to expertise etc. In the chapeau of para [98], the court appears to have been talking about the need to distinguish carefully between a question which is non-justiciable in the strict sense and one that not.]

    The traditional view of non-justiciability is, as you say, the court saying “we won’t touch this”.  But the courts came to ask: won’t touch what?  As Mark Elliott and Jason Varuhas put it in Administrative Law: Text and Materials (5th ed, 2017), at 123-124, there is a difference between a “subject-matter-based” and an “issue-based” conception of non-justiciability. An example of this may be seen in Colin Chan, where the Court of Appeal said that the proportionality of the ban on the publications was non-justiciable but the reasonableness was justiciable. To put it another way, in the words of Peter Cane, Leighton McDonald, and Kristen Rundle, in Principles of Administrative Law (3rd ed, 2018) at 65: the old “all-or-nothing approach to justiciability has largely broken down and… certain categories of decisions or issues may be held non-justiciable in relation to some grounds of review, but not others”. So non-justiciability has come to shade into deference.

    But there is more to deference, as we can see from how that term came to be developed. In the UK, the expansion of non-justiciability has (roughly) coincided with the development of jurisprudence concerning the Human Rights Act 1998 in the UK, in which the word ‘deference’ was used as follows. The Human Rights Act 1998 calls on the UK courts to scrutinise legislation and executive action for compliance with the European Convention on Human Rights. The courts are, in so doing, to apply a ground of review known as ‘proportionality’. On its face, proportionality allows the court to strike down executive action/declare legislation incompatible with the ECHR for reasons relating to the substantive merits of the executive action/legislation, namely, that the action/legislation does not pursue a legitimate aim and/or goes further than is necessary to achieve that aim. To mitigate this risk, the doctrine of deference is applied: even as the courts apply the potentially intrusive concept of proportionality, this potential is mitigated because proportionality is applied in a deferential manner.

    This explains how the concept of ‘deference’ is even more flexible than the already-flexible version of non-justiciability. As we have seen, the flexible version of non-justiciability says that some grounds of review may be justiciable but others not. The concept of deference would appear to add even more nuance by, instead of saying that certain grounds of review are not possible, saying that they are possible but the way in which they can be implied can be in a more/less deferential manner.

    In the UK, this applies in Human Rights Act 1998 cases when applying proportionality. How about Singapore? There is no case squarely addressing this issue yet. But an interesting potential example is discussed in Calvin Liang and Tham Lijing’s article, ‘Fair Play and Match Fixing’, in which the authors say that Tan Seet Eng has hinted at the possible expansion of error of fact as a ground of review in future, but also that it be applied in a manner deferential to the executive (footnotes omitted; emphasis in original in bold; emphasis added in underline):

    “LJ: … The Minister’s justification for detention under the CLTPA (and also, the ISA) has two broad stages. The first stage is where the Minister assesses the evidence against the detainee, and comes to the conclusion that, for example, the detainee is involved in activities X, Y and Z. The second stage is where the Minister then sets out his grounds for ordering detention, which will invariably be based on the Minister’s initial assessment of the evidence (eg You have done X, Y, Z). 

     CL: Perhaps another way of looking at it is to see stage 1 as focusing on the substantive merits, and stage 2 as focusing on the formal grounds for detention.

     LJ: Yes, quite. The ratio of Tan Seet Eng focused on the second stage. On the facts, the Court of Appeal found that the formal grounds set out by the Minister did not satisfy the requirements for detention under the CLTPA. However, what I find interesting are some comments in Tan Seet Eng that suggest the Courts may judicially review not just the second stage, but also the first stage. Now that would be “groundbreaking”. Because it suggests that the Courts might have the power to review the truth of the allegations of fact found in the grounds of the detention (ie in your terminology, the substantive merits of the detention).

     These comments appear at least four times in Tan Seet Eng. The first is at para 54:

     It was not suggested to us that the foregoing analysis in Chng Suan Tze, including its acceptance of the position that Parliament could delegate fact-finding powers to some other branch than the Judiciary, was incorrect and we therefore do not examine this issue here (emphasis added).

     The second is at para 64:

    Chng Suan Tze accepts that Parliament may decide to entrust the power and the responsibility to make a decision to a particular decision-maker. This may extend to decisions as to what the facts are, the application of those facts to the relevant rules or considerations, and any consequential exercise of discretion. As we have noted at [54] above, we have proceeded on the basis that this statement is correct since nothing to the contrary was put to us (emphasis added).

    Again, at para 74:

    The CLTPA vests a potentially draconian power in the Executive. “Public safety, peace and good order” can cover a wide spectrum of possible scenarios. As was pointed out in the course of the oral hearing, even relatively minor regulatory offences could in theory be said to go against public safety, peace and good order. It is true that Parliament has vested this power in the Minister and as we have already noted, we have proceeded on the basis that this is permissible; but as we have just observed, the exercise of the power can and will be scrutinised objectively based on the usual principles (emphasis added).

    And finally, almost as if just in case the reader had missed it, at para 129:

     First, as we have mentioned earlier, and subject to the fact that the contrary was not argued before us, as we have noted at [54], [64] and [74] above, the Court is generally not to review the truth of the allegations of fact found in the grounds of detention.

    So I wonder, what if the Appellant had argued that the Courts can and should, in certain circumstances, and subject to suitable “deference” to the executive, review the truth of allegations of fact found in the grounds of detention (ie stage 1)?

    […]

    LJ: … should the Court of Appeal find, in a later case, that the Courts can judicially review the truth of allegations of fact, it can continue to accord due deference to the executive by, among other things, tempering the intensity of review and the burden of proof. Perhaps the Court of Appeal may adopt an approach similar to that used in the review of the exercise of Prosecutorial discretion. There are many unanswered questions at this stage.”

    (By the way, this is an example of the sort of sharp observation that can only be made by reading cases carefully in full!)

    Aileen Kavanagh also discusses case studies of deference in her article ‘Defending Deference in Public Law’. One that stands out to me is A v Secretary of State for the Home Department [2005] 2 AC 68 (UKHL), also known as the “Belmarsh Prison case”. In that case, shortly after the 9/11 terror attacks, the UK legislature introduced section 23 of the Anti-terrorism, Crime and Security Act 2001, which provided that foreigners who had been suspected of terrorism could be detained indefinitely in the UK in cases where they could not be deported. Besides incompatibility with various provisions of the ECHR, various judges in the House of Lords described this is “irrational”. This was despite the national security context. See https://youtu.be/PZtYENfNa7k?t=2772 (beginning at 46:14) for background; Baroness Hale (beginning at 48:39) explains this view. Note the use of the word “irrational” – irrationality is a recognised ground of judicial review in Singapore too. (And if anybody thinks that the case was decided the way it was because of something particular to the ECHR, I would suggest that a similar conclusion may well have been reached under the equality provision in Article 12(1) of Singapore’s Constitution – compare our Article 12(1) with invite them to compare Article 14 of the ECHR with Article 12(1) of Singapore’s Constitution.)

    (By the way, on the topic of videos: you may be interested in having a look at the UK Supreme Court’s YouTube channel at https://www.youtube.com/channel/UCdkf93h71xVAl28v467Hk7w, which contains videos of judgments being handed down; and there are videos of hearings on the UK Supreme Court website: https://www.supremecourt.uk.)

  • Why study theories relating to the law, instead of just the law itself?

    What is the use of studying prescriptive theories? After all, if someone in power is set in his/her ways, there is no point trying to change their minds by reference to a theory of what he/she should do. In such a situation, we should instead focus our energies on what he/she does do, so that we can predict how he/she will act and plan our lives accordingly.

    The answer, I think, begins with the fact that that person appears to be set in his/her ways. Why would we think so? Take, for example, the case of SGB Starkstrom, in which the Court of Appeal expressed scepticism about the doctrine of substantive legitimate expectations because it might “require the courts to review the substantive merits of executive action as opposed to questions of process and of legality and jurisdiction” [61(a)] (emphasis in original). Now, the court was not saying that the doctrine of substantive legitimate expectations cannot be recognised. To the contrary, it said that it can be recognised if we overcome the merits/legality problem. Far from being set in its ways, one can say that the court was inviting proposals to embrace the doctrine of substantive legitimate expectations in a way that responds to the merits/legality problem.

    Given the decision in SGB Starkstrom, how should we act if our client wishes to claim that he/she has suffered the breach of a substantive legitimate expectation?

    1. We could tell our client to give up and abandon the claim. That is not in and of itself wrong: there are sometimes cases where the best advice a lawyer can give is to give up. But if giving up were the default option, the law would stagnate, and lawyers would risk becoming useless.
    2. We could attempt to reframe the claim in language that avoids that of legitimate expectations. This is slightly more useful advice. But there is always the risk that the court would accuse us of trying to mount a legitimate expectation-based claim by the back door. But this does not mean that we should give up. For, as we have seen, the courts are not opposed to the doctrine of substantive legitimate expectations; they are willing to embrace the doctrine if we surmount certain theoretical difficulties. Therefore, in response, we need to adopt one of the following two strategies.
    3. We could explain to the court why what we are asking it to do does not, in fact, risk taking the courts into ‘reviewing’ the merits of the executive’s decision. And this is where the theory comes in. We will need to explain to the court what exactly a legitimate expectation-based claim entails, and to show the court how we can reconcile our claim with the court’s concerns by presenting to the court a framework which can accommodate our claim while avoiding the problems that the court foresees.
    4. We could attempt to persuade the court why it is actually OK for the courts to ‘review’ the merits of the executive’s decision. Again, we will need to tap on theory to explain why.

    Prescriptive theories can change minds.  Consider, for example, the case of Khawaja [1984] 1 AC 74 (UKHL). In Zamir [1980] AC 930 (UKHL), the House of Lords held that it is for an immigration officer, acting in his/her discretion, to determine whether a person is an “illegal entrant”. In Khawaja, the House of Lords changed its mind, holding that whether a person is an “illegal entrant” is a question of precedent fact, on which the courts must have the last word. We are told at 125G that what changed the court’s mind was “the fuller argument [it had] heard in these appeals”. What argument? Not just practical arguments attempting to distinguish Zamir (at, e.g., 84D), but also theoretical arguments based on the importance of fundamental liberties (at 84H and 85A; see the court’s response at 97H, 101A, 105E, 111F, etc.).

    Often a person who is, at first glance, hard-hearted and unwilling to change his/her mind is not so, but rather troubled by what he/she perceives as one or two key obstacles. If we use theory to explain how those obstacles can be circumvented or removed, or why they are not truly obstacles, we can pave the way for the securing of a practical outcome.

    Of course, such arguments may fail. But one should at least try. If one genuinely thinks the status quo to be iniquitous, one should do what is in one’s power, lawful, and ethical to attempt to change it. The attempt must then (a) fail and expose the iniquities of the status quo for the world to see; (b) fail, but explain why the status quo is not so iniquitous after all; (c) succeed. None of these can be said to be a fruitless outcome. What would be fruitless is giving up easily.

    Of course, theory is not an end-point. Consider, again, Khawaja. Lord Scarman said at 111F: “If Parliament intends to exclude effective judicial review of the exercise of a power in restraint of liberty, it must make its meaning crystal clear.” In other words, Parliament could reverse the ratio of that case by legislation. But this would, in turn, begin the cycle anew. How would the courts respond – for instance, by holding that Parliament’s new legislation is not “crystal clear enough? by inventing some new ground of judicial review?  Again, theory would have a role to play in assisting the courts in deciding on that novel issue.


    What is the use of studying descriptive theories? After all, such theories attempt to model reality. But why model reality, when we can study reality itself?  Why not content ourselves with reading the cases to see what the courts do? Take, for example, the problem of constitutional interpretation. There are lots of theories on how to go about constitutional interpretation, and no one seems to be the dominant one.

    That argument would be correct if our job, in studying reality, were merely to describe it. But we are more than describers; we are critical thinkers and evaluators. If reality is out of step with what theory, there are two possibilities. First, the theory purports to be a descriptive theory, but is wrong. Second, the theory is a prescriptive theory, and what is done in reality is wrong. Theory is either a useful yardstick, or a useful guide to understanding.

    Or both, for, as we have seen, a theory is not necessarily purely descriptive or purely prescriptive. If a series of acts shows such consistency that it may be boiled down to a pattern, then we may be – and ought to be – suspicious when practice begins to deviate from that pattern; and thus what begins as descriptive may become prescriptive. If practice deviates from a prescriptive model and purports to provide a good reason why, we ought to wonder whether we ought to stop condemning what is done in practice for failing to accord with the logic of our theory and instead ask whether what is done in practice has its own internal logic – which may, in turn, form the basis of a new descriptive model which explains why our old prescriptive model is deficient.


    A student asked: But do courts paper over theory with a mere appeal to ‘deference’ to other branches of the state? My answer:

    The courts rightly do not treat ‘deference’ as an excuse for forsaking doctrinal coherence. For the courts to consider otherwise would be to abandon the very enterprise of law itself.

    It is true that courts are generally unwilling to engage in extensive discussion of issues which are not required to be resolved in order to address the concrete case at hand. However, the courts address the case at hand in accordance with law, which in turn calls for sound legal principle. Therefore, the courts often should, and do, go far in laying down a sound legal framework which reaches further than is required to dispose of the case before it. If the common law did not do this, it would be nothing more than thinly veiled palm-tree justice. To the extent that theorists have a role in assisting the court in determining whether a legal framework is sound, the courts do not shy away from taking into account the views of theorists, nor from appealing to arguments from theoretical coherence to develop the law.

    We see this in, for example, Ochroid, in which the court laid down a comprehensive set of rules on illegality in contract law including common-law illegality, despite the fact that the facts of the case involved only statutory illegality and not common-law illegality. This, in turn, called for a survey of various academic articles, culminating in the court’s accepting Birks’s ‘non-stultification’ theory (at [145]ff). The court was unwilling to stop at saying that it would defer to the Legislature’s power to make laws regulating contracts, even at the expense of freedom of contract. If the court had done so, it would have, in one sense, performed its duty of disposing of the case before it, but neglected its duty to see to it that the law is developed in a coherent manner.

    Tan Seet Eng itself is another useful case study. The court did not need to say anything about irrationality to dispose of the case; yet the court did, drawing on academic writers’ explanation of the difference between illegality and irrationality. The court did this in order to ensure that the law it laid down was sound. The court was certainly not willing to invoke an abstract notion of ‘deference’ to sweep away any need for nuance and doctrinal coherence in the development of the law. To the contrary, the very idea of ‘deference’ which the court espoused was informed by a discussion of why ‘deference’ takes place at all, and hence how it should: see [90]ff.

    One might, in response, point to instances such as the Court of Appeal’s discussion of the Hart-Devlin debate in Lim Meng Suang at [162]ff, which concluded at [169] that “there are no clear answers” to the debate and the debate raises “issues that lie quintessentially within the sphere of the Legislature“. But this, too, was itself based on a theory of the proper role of the legislature (at [77]ff).

    Of course, the details of the framework created by a court may not be fully formed at the time the framework is laid down. In Ochroid, the Court of Appeal explicitly left open the question of whether the ‘non-stultification’ principle ought to explain areas of law other than unjust enrichment (at [161]-[168]). In Tan Seet Eng, the Court of Appeal deliberately avoided answering various “difficult questions” relating to irrationality (at [149]). But what is clear is that the courts created frameworks into which the answers to such questions may be situated. The courts have never taken the unwillingness to rule on hypothetical issues to the extent of always saying nothing more than is necessary to dispose of the case before it. Moreover, the courts have never sought to justify such a view in the name of “deference”.

    This must be so, because the courts are fully aware that a common-law legal system (if not any legal system) must value consistency, which calls for principle, which, by definition, is something (a) of greater application than in any one particular case; (b) justifiable by reference to something greater than any one particular case. One might retort with the observation (and it can only be an observation, not a rule) that hard cases (sometimes) make bad law. But for the courts to shy away from the need for doctrinal consistency/theoretical defensibility is even worse: it would mean that hard cases make no law at all.

  • Cakes and compelled speech

    Lee v Ashers Baking Company Ltd [2018] UKSC 49 involved a situation where Ashers Bakery had refused to make a cake for Lee with the words “support gay marriage” on it. The UK Supreme Court held that this was sexual orientation discrimination by Ashers: “the objection was to the message and not to any particular person or persons” (at [34]), and so Ashers’ treatment of Lee was not “on the ground of” sexual orientation (as the relevant discrimination legislation required).  

    There was legislation (known as “FETO”) that prohibited a provider of goods from refusing to provide goods to someone due to, among other things, “discrimination on the ground of… political opinion”. According to the UKSC, FETO could not be interpreted as compelling the supplier of goods to speak, because that would violate:

    • (a)    the provider’s “freedom of thought, conscience and religion” by “obliging a person to manifest a belief which he does not hold” (at [50]);
    • (b)    the provider’s “freedom of expression”, by compelling speech.

    In class, [a student] said of the bakery: “You’re baking the cake, but it isn’t your cake.” The UK Supreme Court disagreed: “what matters is that by being required to produce the cake they were being required to express a message with which they deeply disagreed” (at [54]). The UKSC appears to have thought that compelled speech is impermissible if you would be compelled to speak, even if the message wasn’t ultimately associated with you or even if you would not be “thought by others to support the message” (at [54]). This is different from the Singapore Court of Appeal’s reading of the doctrine of compelled speech in The Online Citizen, according to which, even if you are made to say A, that isn’t “compelled speech” if you are at liberty to say not-A.

    With respect, the Singapore court’s reading of Lee v Ashers Bakery is not entirely correct. According to the SGCA: “if Ashers Bakery had been required to provide the cake decorated with the pro-gay marriage caption, it would have had to advance a message with which it profoundly disagreed, and it did not have the option of expressing its disagreement with that message in the same medium (namely, on the cake)” (at [76]). But the UKSC simply did not say that.

    So, in my view, the SGCA’s decision boils down to this: the SGCA distinguished the majority’s decision in Wooley v Maynard on the ground that it concerned a “right to freedom of belief”, which the Singapore Constitution does not mention (at [70]); and it distinguished Lee v Ashers on the grounds that it concerned the “freedom to hold opinions” (at [75], quoting Art 10 of the ECHR).

    An interesting thought: Would Lee v Ashers be decided differently if, instead of “support gay marriage”, the order was for a cake that said “QueerSpace supports gay marriage” or “Lee supports gay marriage”? Then, I suppose it could be argued that Ashers wouldn’t be asked to “express a message with which they deeply disagreed”; they would merely be asked to state a fact. If I am right, then it is similarly unobjectionable to compel someone to state a message saying, in effect, ‘The Singapore Government says this is false.’

    For further commentary, see this post on the UK Constitutional Law Blog by Jacob Rowbottom: https://ukconstitutionallaw.org/2018/10/16/jacob-rowbottom-cakes-gay-marriage-and-the-right-against-compelled-speech/

    By the way, it is not true that the European Court of Human Rights ruled in favour of the bakery. It did not consider the case on its merits; it summarily held that the case was inadmissible because Lee had failed to exhaust his domestic remedies by raising arguments about the ECHR before the UK courts (Lee v UK, app no 18860/2019, 6 January 2022).

  • Some thoughts on ‘filters’ (including the requirement to apply for leave to apply for judicial review)

    In class, we discussed the need to apply for (and obtain) leave of the court before one can apply for judicial review. The purpose of the leave stage is to serve as a filter mechanism to filter out certain types of cases on grounds relating to matters such as: the nature of the subject-matter of the claim; who the applicant is; the applicant’s motives; the strength of the applicant’s case. Let us call these the ‘filter criteria’.

    There are two ways in which the law can apply a filter.

    The first is to require the applicant to show that he meets the filter criteria before he can commence a legal proceeding. This is how judicial review works: the applicant must obtain leave before he can apply for a prerogative order. And the burden of proof is on the applicant to show that he meets the filter criteria.

    The second is to let the applicant commence legal proceedings first, but then the respondent may get the court to throw the claim out if the respondent can show that the applicant does not meet the filter criteria. This is how an ordinary suit works. If I sue you for breach of contract, the matter goes to trial unless you successfully ask the court to strike out my claim (under Order 18, rule 19 of the Rules of Court). You can ask the court to strike out my claim on grounds such as that it is clearly factually or legally unsustainable; “scandalous, frivolous or vexatious”; or “otherwise an abuse of the process of the Court”.

    Some of you asked whether having a filter makes it easier or harder to apply for judicial review, or neither easier nor harder. A number of you criticised the first type of filter as it is applied in Singapore: for example, the law on standing is unclear.  One of you went further and suggested that these criticisms may be addressed if we can move to the second type of filter. I write to respond to these views with some preliminary thoughts, with the caveat that they have not yet been fully formed and I have yet to articulate them in a precise manner.

    First, consider the case in which the filter requirements are a subset of what the applicant must show to succeed in the substantive application. In that case:

    1. The second type of filter is harder on the respondent than the first type of filter if the applicant’s case is unmeritorious: it is the respondent who must bear the burden of proof, not to mention the expense, of applying to strike out the claim. But that doesn’t make the applicant’s task any easier: the applicant still has to prove that his claim is made out, otherwise he will lose. In other words, the second type of filter can disadvantage the respondent (if the applicant’s claim is unmeritorious); it neither advantages nor disadvantages the applicant.
    2. The first type of filter makes the respondent’s task easier than if there were no filter, if the applicant’s case is unmeritorious. This is because the respondent has to put in less effort resisting an application for leave, compared to resisting the principal application. In other words, the first type of filter can advantage the respondent (if the applicant’s claim is unmeritorious); it neither advantages nor disadvantages the respondent.

    All that would be so if, as I have said, the filter requirements are a subset of what the applicant has to prove to succeed in the principal application. For example, imagine that the only filter requirement were that there was a prima facie case of reasonable suspicion that the applicant would succeed in the principal application. With the first type of filter, the arguments used by the applicant to obtain leave would be the exact same arguments used by the applicant at the substantive stage. With the second type of filter, the arguments used by the applicant to resist a striking-out application would be the exact same arguments used by the applicant at the substantive stage.

    But matters become trickier when what the applicant has to prove to obtain leave is something additional to, rather than a subset of, what the applicant has to prove at the substantive stage. For example, under Singapore law, the applicant must prove that he has standing. As one of you pointed out, it is, in theory, possible for one or more grounds of review to be made out, and yet the applicant fails for want of standing. So the leave stage makes things harder for the applicant by requiring the applicant to go through the effort of advancing not only the type of legal argument that he would have to advance at the substantive stage, but also another type of legal argument (viz. arguments on standing).

    Now, let us suppose that we have criticisms of the filter requirements. Of course, the law could address our criticisms by continuing to use the first type of filter, but by reforming the filter requirements. But could the law address our criticisms by moving from the first type of filter to the second? Not necessarily. It would, in principle, be possible for the law to change to using the second type of filter, but keep the same filter requirements. That would not address our criticisms of the filter requirements.

    In more concrete terms: Suppose that we say that the law on standing is problematic. One may say that, as a solution, instead of requiring the applicant to apply for leave before making the principal application for judicial review, the applicant can simply make the principal application but the respondent can apply to have it struck out. But this will not necessarily solve the problems of the law on standing. It depends on the grounds of striking out. For example, if the law on striking out is that the respondent may apply to have the claim struck out if the applicant has no standing, then the problematic law on standing will remain. Sure, a different person will bear the burden of proof. But the content of the law on standing would be no less open to criticism than is the present law on standing.

    Therefore, if changing from the first type of filter to the second is to simplify the law, what will really be responsible for the simplification is not the change in the type of filter, but rather the change in the filter requirements. It does not follow from a change in the type of filter that the filter requirements will therefore change.

    To put it more concretely: Whatever criticisms we have of the law on what one must show to obtain leave to apply for judicial review, those criticisms would not necessarily be addressed by abolishing the leave requirement and replacing it with the possibility of the court striking out the claim at the instance of the respondent. Such criticisms would be criticisms of what the filter requirements are. The type of filter is a completely separate matter. A change in the latter does not necessitate a change in the former. Therefore, it is possible, and may be desirable, to change the former without changing the latter.

  • Can A apply for judicial review, in A’s own name, of a decision affecting B only?

    Mr M Ravi once attempted to apply, in his own name, for an order that the execution of Prabagaran a/l Srivijayan be stayed. He was, at the time, not a practising lawyer. Does that mean that someone who is not a practising lawyer (A) can, in effect, represent someone else (B) in court by applying in A’s own name? If so, why bother becoming a practising lawyer at all?

    If you want to make an application in court, you can either get a lawyer to represent you, or represent yourself.  (By the way, to be a lawyer representing someone in court, one must have a valid practising certificate in force. It is not enough to be called to the Singapore Bar. For example, I have been called to the Singapore Bar, but do not have a practising certificate in force. To get a practising certificate, one must, inter alia, get insurance against the risk of one being liable for one’s own professional negligence.)

    To put it another way, if X is (lawfully) making an application in court, either:

    1. it is Y’s application, and X is representing Y by being Y’s lawyer; or
    2. it is X’s application.

    Ravi was X and Prabagaran was Y.

    Situation (a) was not possible, because Ravi had no valid practising certificate in force (and, even if he had, Prabagaran was not his client). So only situation (b) was possible. But, in that situation, the application was Ravi’s application, not Prabagaran’s. Therefore, if (for example) standing was an issue, then Ravi would have to show that Ravi (the applicant) had standing. It would not have been enough for him to show that Prabagaran had standing, because Prabagaran was not the applicant.

    None of this diminishes anybody’s right to be a litigant in person. But a litigant in person can only litigate in respect of claims by/against him. So if the law is such that A has no right of action, then A cannot succeed merely by saying that B has a right of action. If the law were otherwise, then just anybody would be able to sue anyone for any alleged wrong; there would be no such thing as standing, or privity of contract. There could, in theory, be such a legal system; but ours is not such a system.

    That is why there are debates on whether the law on standing should be changed. Typically, the situation is as follows. A’s right has (allegedly) been violated by the executive. But A is unable/unwilling to apply for judicial review. B, a stranger, is unhappy with what the executive has done. What can B do?

    1. If B happens to be a practising lawyer, B can represent A in court. That will mean that A is the applicant.
    2. B can try to persuade A to apply for judicial review, and/or give some money to A to hire a lawyer (which is not always lawful because of the law on champerty/maintenance). That will mean that A is the applicant.
    3. B could himself be the applicant.

    Options 1 and 2 are contingent on A’s consent. If A does not consent, that leaves option 3: B being the applicant.  But the law on standing does not allow B to be the applicant. So B would need to persuade the court to broaden the law on standing to allow B, and not just A, to be the applicant. (Often, this takes the form of an argument that B should have standing because any ‘concerned citizen’ should have standing.)

    A case study that illustrates all this is Madan Mohan Singh v AG [2015] 2 SLR 1085. That case involved a challenge to a policy that affected certain prisoners. The applicant, who was a volunteer counsellor, sought to challenge that policy. He failed because he had no standing: if any rights had been arguably violated, it was the prisoners’ rights, not the counsellor’s. Moreover, the counsellor could not make the application on behalf of the prisoners.

    To my mind, the easiest course of action would have been for him to persuade one of the prisoners to bring an application in his (the prisoner’s) own name. It may well have been that he did not manage to (see [17]).

    Does that mean that there was the potential that the prison’s policy was unlawful? In theory, perhaps, (although the High Court said, obiter, that the policy was not unlawful). But even if it were unlawful, as we have been told in Jeyaretnam, the court is there to redress disputes between the parties, and not just to go around trying to fix unlawfulness in the abstract. If the person who was the victim of the unlawfulness did not want to bring an application, why should a stranger be allowed to? If, for example, the victim of the unlawfulness wished to try to settle his dispute with the executive by means of negotiations, why should a stranger be allowed to scuttle those negotiations by going to court?

    Similarly, if Prabagaran wanted to have his application handled by Choo and nobody else, and (for whatever reasons, e.g. as part of a broader litigation strategy) mount legal arguments A and B and none else, why should a stranger be allowed to subvert his plans and mount legal arguments C and D instead?


    On a tangentially related note: Why did M Ravi make his application for a stay to the Court of Appeal directly, rather than the High Court?

    I haven’t seen a copy of the court papers, but I think the explanation is as follows. It seems that the Court of Appeal, having held that it has an inherent power to re-open a concluded appeal, also implicitly held that this power includes the power to stay the execution of a sentence which had been passed at the conclusion of that appeal. These powers are possessed by the Court of Appeal, not the High Court, and are an offshoot of the Court of Appeal’s “full power to determine any question necessary to be determined for the purpose of doing justice”: s 29A(4) of the SCJA, discussed in Koh Zhan Quan Tony v PP [2006] 2 SLR(R) 830 (CA) [18] and Koh Jabing v PP [2016] 3 SLR 135 (CA) [12] and [17]. One would therefore think that one can only ask the Court of Appeal, and not the High Court, to exercise these powers.

  • Did Vijaya Kumar s/o Rajendran v AG [2015] SGHC 244 broaden the standing test?

    I would prefer to say no, because saying yes would mean that the High Court had defied no fewer than three Court of Appeal cases. Rather, I think that the High Court’s decision is compatible with the standing test in Tan Eng Hong etc. What had been ‘broadened’, if anything, was the sense of what the Art 15 right consists of. In granting the 3rd applicant (the spectator/photographer) standing, the High Court effectively held that watching/photographing a religious procession is an act of “practice” of that religion.

    On the other hand, this is not a very great expansion. Consider the implication in Colin Chan that a mere ‘citizen’ who follows the Jehovah’s Witnesses religion would have had standing: “He does, however, have a constitutional right to profess, practise and propagate his religion, and it is alleged that the publications by IBSA are essential for him to carry out these activities.”  To the extent that this means that a ‘mere’ reader of these materials would have had standing, it is not that much of a stretch to say that, in Thaipusam, a ‘mere’ viewer would have had standing.

  • Defamation, freedom of expression, and “Parliament may by law impose…”

    What objections may one have to the idea that the common law of defamation is compatible with Art 14(1) of the Constitution because “Parliament [has] by law impose[d]” it by passing the Defamation Act / by enacting Art 162 of the Constitution?

    I.

    One argument is that it is quite artificial to say that Parliament has, by passing the Defamation Act, somehow given the legislative stamp of approval to the entire common law of defamation, past and future. Has Parliament really ever directed its mind to each and every rule of the common law of defamation?

    Moreover, if it is correct to say that the entire common law of defamation, as of the time the Defamation Act was passed / as of the commencement of the Constitution, is compliant with Art 14, then how can Review Publishing say that it is ever possible for Art 14 to necessitate the Reynolds privilege?  If the Defamation Act said that the common law of defamation sans the Reynolds privilege is constitutionally compliant, how could the courts ever say one day that the Constitution requires that Reynolds privilege be available? Indeed, if the courts were to say that (or even if they were merely to say that it is desirable that Reynolds privilege be available), wouldn’t they be going against the intention of the Legislature in passing the Defamation Act, which was to enact the common law of defamation as it was in 1965?

    II.

    Michael Hor, in his note on JJB v LKY [1992] 1 SLR(R) 791 (CA) (‘The Freedom of Speech and Defamation’ [1992] SJLS 542), addresses a different argument. The argument is premised on the idea that (in Hor’s words) the Parliament of Singapore, “upon separation from Malaysia in 1965… provided for the continuation of both Art 10 [of Malaysia’s Constitution – same as our present Art 14] and the existing rules of defamation” (Hor at 550), and hence intended that the common law of defamation not be incompatible with Art 10. Hor’s criticism of this argument is that it does not follow that Parliament thought that the common law of defamation was compliant with the Malaysian Art 10 (i.e. our present Art 14) – Parliament had simply not directed its mind to the question.

    As I understand it, the argument is that Parliament is presumed not to intend that one of its enactments contradict the other. Hor’s counter-argument is that this ignores the fact that one enactment is constitutional in nature and must therefore take precedence over the other, which is not; instead, the argument goes, the courts have in effect held that Parliament intended that the right in Art 10(1)(a) (now Art 14(1)(a)) be attenuated in order to accommodate the common law of defamation.

    III.

    One might next argue: isn’t Art 162 the means through which Parliament has by law enacted the law of defamation? The problem is that Art 162 provides that laws existing as of the commencement of the Constitution are to continue in force but are to be “construed… with such modifications, adaptations, qualifications and exceptions as may be necessary to bring them into conformity with [the] Constitution”. Therefore, one would think, it does not follow that the entire common law of defamation is constitutional in the absence of such modifications etc. Hor has criticised the courts’ response to this as circular:

    “The real issue is whether [the common law of defamation] need[ed] to undergo any change to comply with the requirements of the Constitution. All that the Court has to offer in response (in [paragraph 57 of JJB v LKY]) is a bald assertion that ‘the law of defamation is not inconsistent with the right of free speech.’ With respect, this begs the question and contributes nothing to the argument.”

    It appears to me that the Court of Appeal’s reasoning in Review Publishing contains the same circularity. The Court of Appeal said at [250] that the effect of Art 162 (which used to be Art 105(1) of the State Constitution) was to enact the common law of defamation as it was. Why not with modifications to make it constitutionally valid? Because it was constitutionally valid: “The imposition via Art 105(1) of this restriction on constitutional free speech was permitted by and also consistent with Art 10(2)(a) of the 1963 Federal Constitution (now Art 14(2)(a) of the Singapore Constitution).”

    So the Court of Appeal is saying that (a) the common law of defamation is compliant with Art 10(1)(a) because “Parliament [has] by law impose[d] it” via Art 162; (b) Art 162 provides that the common law of defamation was to be modified etc. to make it compliant with the Constitution; (c) there is no need for such modification etc. because the common law of defamation was already compliant with Art 10(1)(a), in that “Parliament [has] by law impose[d] it”.

  • Judicial review for error of fact: some difficult cases

    What did the High Court do in Fong Thin Choo? Did it, in substance, decide that courts have the power to review any executive decision for any sort of error of fact?

    Preliminary point: even if the answer is yes, that does not mean that the courts will be usurping the decision-maker’s role. Even if the courts had the final word on what the facts are, it would still be for the decision-maker to make use of the facts in his exercise of discretionary power.

    The High Court’s initial view is that it was a precedent fact requirement that the cigarettes had not been exported: “The DG must establish the fact of non-export before the exporter becomes liable for customs duty.” There is nothing noteworthy about this. What is noteworthy is that the High Court reached this conclusion based on a certain interpretation of “evidence” and “satisfaction” — evidence “sufficient… to prove that goods have indeed been exported”; satisfaction “not to be measured in terms of reasonableness but in terms of justification on the evidence”. One may wonder whether this is a strained reading of the statute, which used the phrase “not accounted for to the satisfaction of the proper officer of customs”. Taking the statute at face value, and following the reasoning in Chng Suan Tze, once the officer says ‘I am satisfied’ then the only grounds of review would be illegality, irrationality, and procedural impropriety. Yet the High Court said that it was a precedent fact requirement that the cigarettes were actually not exported, which meant that the High Court had the final word on whether or not they had been exported. True, this is error of law. But the High Court’s pronouncement on what the law said, in terms of the correct interpretation of “evidence” and “satisfaction”, led to the result that the High Court would be free to review the DG’s decision for the correctness of his conclusion.

    The High Court also referred to the Tameside case [27] – Chan J seems to have conflated this with the ‘precedent fact’ ground of review. In Tameside, Lord Wilberforce said that even if the precedent fact requirement is ‘that the Secretary of State “is satisfied”‘, that simply means that the Secretary of State is to “evaluat[e]… those facts”; but still “the court must inquire whether those facts exist, and have been taken into account, whether the judgment has been made upon a proper self-direction as to those facts, whether the judgment has not been made upon other facts which ought not to have been taken into account.”

    It gets better! The High Court then asked: what would the analysis be if it were not a precedent fact requirement that the cigarettes were actually exported? According to the High Court, even if “the DG has a discretionary power to decide whether the evidence produced by the applicants accounted for the export of the goods”, he had to have a “factual basis for… being dissatisfied”. One could retort that, even if there was some evidence tending to show that the cigarettes had been exported, there was also some evidence tending to show that they had not been exported; and the latter was the “factual basis”. But the High Court rejected this. How? By saying that the respondent had failed to take into account some other relevant evidence [52]; had failed to investigate certain additional evidence [51]; and had placed “undue weight” on some pieces of evidence but not others [53]. Is this review for error of fact in substance, thinly disguised as error of law?

    The reasoning is similar to that in Tameside. In that case, the local authority wanted to maintain the old education policy.  The Minister (different political party) directed the local authority to implement a new policy immediately. He pointed to a statute that said that the Minister may give directions to a local authority if he is “satisfied… that any local education authority… [has] acted or [is] proposing to act unreasonably with respect to the exercise of any power conferred or the performance of any duty imposed by or under this Act…” He said, ‘I am satisfied’.

    This was not a ‘precedent fact’ case. Yet, Lord Wilberforce said that the Minister “was operating under a misconception as to what would be reasonable for a newly elected council to do, and that he failed to take into account that it was entitled – indeed in a sense bound – to carry out the policy on which it was elected, and failed to give weight to the fact that [the old policy] was something which a reasonable authority might accept…”

    What did the UK Supreme Court decide in R (A) v Croydon LBC?

    The statute said: “Every local authority shall provide accommodation for any child in need within their area who appears to them to require accommodation as a result of…”

    The local authorities did not even consider whether the applicants “appear[ed] to them to require accommodation” because they thought that the applicants were not children.

    The High Court dismissed the application for judicial review, holding that it was not a precedent fact requirement that the applicants be children. The Court of Appeal agreed.

    We can break down the case into three questions.

    First: who decides whether the person seeking asylum is a “child”? Answer: the courts. Why? Because whether or not the person seeking accommodation was a “child” was a precedent fact requirement – if the person is a “child”, then the local authority must make some sort of decision. How did the court know that this was a precedent fact requirement? It is a matter of interpreting the statute, but the court was heavily influenced by the fact “whether a person is a ‘child’… is a right or wrong answer”; “The word ‘child’ is undoubtedly defined in wholly objective terms”. [32]

    On this basis, the local authorities had erred in law. Strictly speaking, the Supreme Court did not say that they had committed errors of precedent fact (i.e. thinking that they need not exercise their discretion because the precedent fact is not made out, when in truth they had to exercise their discretion because the precedent fact was true). This was because the evidence was not before the Supreme Court as this was an appeal on preliminary points of law only.

    Therefore, the Supreme Court remitted the case to (I presume) the High Court, for the High Court to decide whether the applicants were children.  [46]

    Presumably, if the High Court had found that the applicants were children, then it would have issued an order to quash the local authorities’ decisions not to grant them accommodation, and/or a mandatory order directing the local authorities to decide whether they were “in need” and whether it “appear[ed] to [the local authorities] [that the children] require[d] accommodation as a result of…”

    To reach this conclusion, the Supreme Court did not need to refer to Tameside or E, in which the errors of fact were not errors of precedent fact.

    Second: if the person is a “child”, is the person “in need”? Who decides? Answer: the local authority. Why? Because the statutory definition of “in need” “requires a number of different value judgments”; “[w]ithin the limits of fair process and ‘Wednesbury reasonableness’ there are no clear cut right or wrong answers”. [26]

    Third: if the person is a “child” “in need”, does it “appea[r] to [the local authority]” that the person “require[s] accommodation as a result of…”? The court said this is “clearly left to the local authority” [14]. (That is, subject for review on the GCHQ grounds.)

    If the court holds that a decision is unlawful because of an error of fact, what order will the court make?

    Generally, a quashing order, and, depending on the circumstances, perhaps a mandatory order to re-make the decision.

    Why was a prohibiting order made in Fong Thin Choo? Because that was what was asked for. Why was it asked for? Probably because of what the DG was threatening to do, namely, to deduct money from the banker’s guarantee furnished by the company (see para 4 of the letter reproduced at [11] of the judgment). In these circumstances, a prohibiting order – an order to not do something – would seem to be the most appropriate, given that the DG had threatened to do that thing.

  • Judicial independence: in whose hands?

    Consider section 3 of the UK’s Constitutional Reform Act 2005. Section 3(1) expressly confers on various executive office-holders the duty to “uphold the continued independence of the judiciary”.

    On its face, the content of the duty is that stated in s 3(5): “The Lord Chancellor and other Ministers of the Crown must not seek to influence particular judicial decisions through any special access to the judiciary.” (Note that the Lord Chancellor is also the Secretary of State for Justice, whose office is roughly equivalent to that of our Minister for Law.)

    But there is a case for saying that the duty extends to a duty to speak up against attacks on the judiciary’s independence. Graham Gee and Mark Elliott have written useful blog posts on this point.

    There is no similar express duty on Ministers in Singapore law, but one is nonetheless prompted to wonder: Who is responsible for safeguarding judicial independence, and how? Is judicial independence safeguarded only by a series of legal rules? I think not. As the UK’s experience shows, the conduct of politicians, and, more generally, certain persons outside the judiciary, can itself have a bearing on judicial independence as well as public perceptions thereof.

    That said, there is at least one situation in which an executive officer has a power to do something that has a direct bearing on the safeguarding of public confidence of the judiciary: the Attorney-General has always had the power to commence proceedings for contempt of court, and, from January 2021, has had the power to apply to court for eave to issue a non-publication direction under s 13 of the Administration of Justice (Protection) Act 2016.

  • Notes on Tan Gek Neo Jessie v Minister for Finance [1991] 1 SLR(R) 1

    To my mind, there are at least three possible ways to think of what happened in Tan Gek Neo Jessie:

    1. The Registrar of Businesses had failed to take into account the mandatory relevant consideration of whether her decision will conflict with the law on trade marks and passing off.
    2. The Registrar of Businesses had failed to take into account the following mandatory relevant considerations:
      • Whether Jessie Tan had used the name “JC Penney” as a trade mark;
      • Whether the other business (the American company) had used “JC Penney” as a trade mark in Singapore;
      • Whether the American company had any reputation in Singapore.
    3. The Registrar of Businesses had failed to take into account the mandatory relevant consideration of whether the public would actually be misled by Jessie Tan’s business’s name.

      Tan Gek Neo Jessie remains useful to us insofar as it demonstrates certain legal principles relevant to the study of administrative law. However, the correctness of certain aspects of the decision on the facts may be doubted, for reasons that you might find interesting if you are interested in the law of trade marks. 

      In Jessie Tan, the High Court said that there was no evidence to support the Registrar’s conclusion that Jessie Tan had the “ulterior motive of riding on the reputation” of the American company because the American company did not carry on any business in Singapore. But the High Court took a different approach in R H Macy & Co Inc v Trade Accents [1991] 2 SLR(R) 98 (HC) (cited in Robert G Howell with Khaw Lake Tee, ‘Parallel  Importation of Wares and Reputation Spillover: Examples of Transnationalization of Law’, ch 2 in Douglas M Johnston and Gerry Ferguson (eds), Asia-Pacific Legal Development (UBC Press, 1998), at 115-116). 

      The applicant in R H Macy was an American company which operated a chain of department stores named “Macy’s”. There was a Singaporean sole proprietor who had registered “Macy’s” as a trade mark in Singapore. The Trade Marks Act allowed someone in the applicant’s position to apply to the court to expunge the respondent’s trade mark from the register of trade marks on the ground that it is “likely to deceive or cause confusion”. (This is similar to s 22(1)(d) of the present Trade Marks Act 1998, which allows a person to apply for a trade mark to be revoked on the ground that it is “liable to mislead the public” because of how it is used.) The High Court allowed the American company’s application. This was despite the fact that the American company did not conduct any business in Singapore. 

      The High Court distinguished Jessie Tan, saying that the crux of that case was that “the Registrar acted on hearsay, exaggerated allegations and inferences not founded on facts” (at [39]). In my respectful view, I do not think that Jessie Tan can be so easily distinguished. That case stated quite clearly that the proprietor of the “JC Penney” and “Penneys” trade marks “could not maintain an action against [Tan] for ingringement of trade marks” (at [19]). 

      That said, that does not necessarily mean that the ultimate conclusion in Jessie Tan was wrong. After all, the statute in that case allowed a business to be told to change its name if the name was “calculated to mislead”; that is a different test from “likely to deceive or cause confusion”. That said, it appears that the High Court was not correct to say that “to direct the applicant to change her business name on the basis that it so nearly resembles the name of the proprietor of [the] two trade marks, in effect, gives to the proprietor more than what the laws on infringement of trade marks and passing off would provide” (at [19]).

      See also Jet Aviation (Singapore) Pte Ltd v Jet Maintenance Pte Ltd [1998] 1 SLR(R) 713 (HC). In that case, the plaintiff sued the defendant for the tort of passing off. To succeed, the plaintiff would have to show that it had goodwill in Singapore. It could not, so it failed. However, the High Court did remark that, in principle, one can acquire goodwill in Singapore even if one does not carry on business in Singapore (at [38]-[42]).

      (Sorry, I am not presently qualified to comment on the state of the law in this area other than as it is set out in the cases described above.)

      The phrase “calculated to mislead”

      In addition, I wish to clarify that the phrase “calculated to mislead” does not necessarily mean that there must have been an intention to mislead; I’m sorry if I suggested that it does. Here, “calculated to mislead” cannot simply refer to one’s intention. We can see this from the wording of the former Business Registration Act 1973, s 11 (which was the statutory provision applied in Tan Gek Neo Jessie). That provision allows the Registrar of Businesses to, in effect, order a business to change its name if the old name “so nearly resembles the name of any corporation or the name under which another person carries on business as to be calculated to mislead”. This suggests that “calculated to mislead” does not refer only to one’s intention, but is instead an objective test of similarity. (If it were only a matter of intention, the words “so nearly resembles…” would not be there, for one can intend to mislead through the name of a business even if the name does not resemble another’s.) We see the phrase “calculated to mislead” being used to refer to an objective test in Caterpillar Inc v Ong Eng Peng [2006] 2 SLR(R) 669 (HC) [17]. In other words, “calculated to mislead” appears to me to mean something more like “bearing features that tend to mislead” or “bearing features that would lead a reasonable person to infer that there was an intention to mislead”. 

      Anyway, the phrase “calculated to mislead” is not part of the Business Names Registration Act 2014 (which replaced the Business Registration Act 1973). Now, we have s 17(4)(d), which allows the Registrar to order a change of name if one’s business name “so nearly resembles the name of any corporation or limited partnership, or any registered business name, as to be likely to be mistaken for it”. If you ask me, the wording is much clearer now.

    1. Executive involvement in sentencing: life imprisonment with the possibility of parole

      [This post isn’t about Sentences for Enhanced Public Protection.]

      Consider the treatment of offenders aged below 18 who commit offences which would otherwise attract the death penalty. In the past, such offenders would be sentenced to detention at the President’s pleasure. Now, such offenders must be sentenced to life imprisonment: Criminal Procedure Code 2010, s 314. But the effect is not very different. This is because ss 50O-50P of the Prisons Act 1933 provide that after a person has served 20 years of a sentence of life imprisonment, the Minister must review his/her case and consider ordering that he/she be released (subject to the conditions set out in s 50S, a breach of which will attract the consequences set out in s 50T); if the Minister refuses, he must review the case again the following year. Is this a breach of the separation of powers?

      To answer that question, it would be useful to consider the decision of the European Court of Human Rights in V v United Kingdom (1999) 30 EHRR 121. In that case, Venables argued that the very system of detention at Her Majesty’s pleasure, to the extent that the duration of detention was determined by a minister, was a breach of Article 6(1) of the European Convention on Human Rights. Article 6(1) states, among other things, that in the “determination… of any criminal charge against him, everyone is entitled to a fair and public hearing within a reasonable time by an independent and impartial tribunal established by law”. The ECtHR held that the task of fixing the ‘tariff’ portion of a sentence of imprisonment during Her Majesty’s pleasure was a “sentencing exercise”, and is therefore part of the process of the “determination… of [a] criminal charge”. The minister was not an “independent an impartial tribunal” as he was “not independent of the executive”. Therefore, Article 6(1) was breached.

      R (Anderson) v Secretary of State for the Home Department [2003] 1 AC 837 (UKHL) concerned legislation similar to Singapore’s system of what one may call life imprisonment with the possibility of parole. (The term “parole” isn’t used in Singapore.) The difference was that, instead of a fixed 20-year tariff, it was up to the Minister to decide the tariff period. Again, the House of Lords held that Article 6(1) was breached, for the same reasons applied by the ECtHR in V.

      Then, in Director of Public Prosecutions of Jamaica v Mollison [2003] 2 AC 411, the Privy Council, following Hinds, held that legislation similar to that in Venables was unconstitutional because it violated the separation of powers: it was “incompatib[le] with the constitutional principle that judicial functions (such as sentencing) must be exercised by the judiciary and not by the executive” (at [13]).

      To the best of my knowledge, nobody in Singapore has sought to challenge the constitutionality of ss 50O-50P of the Prisons Act, or any other legislation providing for similar executive involvement in the process of determining the length of someone’s sentence of imprisonment. (The legislation challenged in Mohammad Faizal bin Sabtu is quite different.)

    2. Hearings by licensing authorities, etc.: inherently biased?

      (In what follows, references to various textbooks are references to the latest editions available in the Course Reserves section of the SMU law library as of March 2022.)

      There are many situations in which you asks a public body for something — say, a licence. The public body has a duty to hear you. But, some of you asked, wouldn’t that make the public body a judge in its own cause? Shouldn’t there be three persons — the applicant, someone to present arguments against the granting of the licence, and a judge to hear both and decide? Why are there so many situations in which there are only two persons involved — the applicant and the public body? I will call these two possible models of decision-making system, for short, the ‘three-body’ and ‘two-body’ model respectively.

      I have borrowed this language from Timothy Endicott (at 175). He notes: “Many public authorities, unlike courts and tribunals, do not hear a dispute between parties, but hear responses to their own (tentative, or even concluded) decisions or policies or plans. So they are already involved in the issues, rather than serving as a neutral third party resolving a dispute between two sides.” 

      The easiest answer is: even if the public body is a judge in its own cause, the necessity principle demands that that be tolerated. But we can go deeper.

      Endicott, like other authors of textbooks, does not take the view that the ‘two-body’ model is intrinsically unfair. After all, the role of the public body is not to oppose the applicant, but rather to decide impartially in the public interest. Therefore, for example, a licensing authority should be thought of more like a judge than like state counsel. (The analogy is not perfect, but the main point still stands: the public body should not be thought of as playing an adversarial role vis-a-vis the applicant”.) For more on this, please see B Johnson & Co (Builders) Ltd v Minister of Health [1947] 2 All ER 395 at 399 (cited in Alconbury at [47]).

      One may draw a further analogy with a judge’s power to summarily punish for contempt committed in the face of the court. In Wilkinson v Lord Chancellor’s Department [2003] EWCA Civ 95 (cited by Endicott at 183), Wilkinson used “foul and abusive language” and committed acts of physical violence in court; the judge assigned counsel to him, then tried him for contempt in the face of the court and sentenced him to imprisonment. (In Singapore, the court can do this under section 25 of the Administration of Justice (Protection) Act 2016.) On appeal, Wilkinson argued that she should have been tried for the contempt by another judge, not the same judge. The Court of Appeal pointed out that “the power to impose summary punishment for actions which one has oneself witnessed and is oneself the victim does appear to place the judge in the position of being witness, prosecutor and judge” (at [24]). However, the Court of Appeal disagreed that the matter had to have been heard by another judge: “strictly speaking the procedure does not offend against the rule of natural justice, nemo iudex in sua causa (no-one shall be judge in his own case), since the prosecution is not aimed at protecting the judge personally but protecting the administration of justice” (at [25], citing Borrie and Lowe’s The Law of Contempt, 3rd ed (1996) at 522). Therefore, the mere fact that the same judge conducted the summary trial for contempt did not mean that there was apparent bias. (Nor did it render the judge other than “impartial” within the meaning of Art 6(1) of the ECHR.)

      There is another case that agrees, though it makes the point slightly differently. In R (Alconbury Ltd) v Secretary of State for the Environment, Transport and the Regions [2003] 2 AC 295 (UKHL), various businesses applied to certain local authorities for planning permission to build certain facilities. The local authorities dismissed the applications. The law was that they had the right to appeal to the Secretary of State for the Environment, Transport and the Regions. The businesses argued that this system was incompatible with Article 6(1) of the ECHR, which provides: “In the determination of his civil rights and obligations or of any criminal charge against him, everyone is entitled to a fair and public hearing within a reasonable time by an independent and impartial tribunal established by law.” They argued that the Secretary of State was not “independent and impartial”, because he “lays down policy and directs what he or the department considers to be the most efficient and effective use of land in what he sees to be the public interest”, and would make decisions, not impartially, but instead in line with these views (at [24]).

      The House of Lords unanimously rejected this argument. For present purposes, I would call your attention to [48], where Lord Slynn said: “it is not right to say that a policy maker cannot be a decision maker or that the final decision maker cannot be a democratically elected person or body”. Lord Hoffmann agreed for reasons that I will not set out here because they repay reading in full: see [69]-[76], and, if you’re interested, the rest of his judgment. 

      The conclusion was that there was no violation of Article 6(1) because, to the extent that the Minister was not “independent and impartial”, the system as a whole was because one could apply for judicial review of the Minister’s decision. 

      That is not to say that, under the ‘two-body’ model, the rules of procedural fairness cease to exist, even as regards first-instance decision-makers. Cooper v Wandsworth has clearly put paid to that idea; and in Wilkinson, the court did apply the test for apparent bias. Rather, the point is this: the principles from the cases on procedural fairness — most of which arise in the context of a ‘three-party’ decision-making system, such as a court — must be applied with certain modifications to suit them to the ‘two-party’ model. What a fair hearing requires under one model may be different from what a fair hearing requires under the other. 

      Besides Endicott, others make this point. Peter Cane, for example, says (at 75) of the task of fact-finding: “the rules of procedure… will be different according to whether an adversarial or a non-adversarial mode of discovering the truth is adopted”; for example, “under an inquisitorial system, the decision-maker’s input is much greater”. The relevance to administrative law is this: “in many administrative contexts, there are only two parties: an applicant (for a benefit or a licence, for instance) and a decision-maker. In that context, certain procedural steps (such as cross-examination of witnesses) might simply be inapplicable. More generally, investigation by the decision-maker may seem more appropriate and less problematic in two-party than in three-party situations…” 

      Paul Craig agrees (at 382), and adds: “We should…. be prepared to think more broadly about what procedural protection connotes. Procedures developed against a backdrop of adjudication may well be the most appropriate in certain areas. They may not, however, be equally suited to all the institutions that comprise administrative law.”

      Sorry, I don’t have a conclusion, just as the authors of the textbooks appear not to have one.

    3. An interesting case involving burning an entire row of houses

      I might have mentioned the following case in class: Beaulieu v Finglam (1401) YB Pas 2 Hen IV, fo 18, pl 6, translated and reported in John Baker, Baker and Milsom: Sources of English Legal History (2nd ed, OUP, 2010) at 610-611.

      In that case, which was heard by the Court of Common Pleas, it appears that an employee or a guest of Roger Finglam negligently caused a fire to spread to an entire row of houses.

      Justice Markham stated the law as follows: “A man is bound to answer for his servant’s act, as for his lodger’s act, in such a case. For if my servant or lodger puts a candle on a wall and the candle falls into the straw and burns the whole house, and also my neighbour’s house, in this case I shall answer to my neighbour for the damage which he has suffered.”

      Counsel then submitted: “This defendant will be undone and impoverished for ever if this action is maintainable against him, for then 20 other suits will be brought against him for the same matter.”

      Chief Justice Thirning retorted: “What is that to us? It is better that he should be utterly undone than that the law should be changed for him.”

      I can’t remember what point I was trying to make, but I have never forgotten this case since I first read it.

    4. Asking judges to recuse themselves

      Some of you asked why an application to disqualify a judge from hearing a case is heard by that very judge. I did a bit of digging around, and found a paper by the Australian Law Reform Commission that sums up various arguments for and against this (at [34]-[47]). See also this speech by Justice John Sackar, Judge of the Supreme Court of New South Wales, at page 35 onwards. That speech cites the El Faragy case; see para 32 of that case, and also see this blog post by Mr Jason Hadden for criticism. For a more radical view, see this article by Professor Dmitry Bam which argues that the practice in the USA of having judges hear applications to recuse themselves violates the parties’ constitutional right to “due process of law” under the Fourteenth Amendment to the US Constitution.

    5. On the legitimacy of state institutions

      Some institutions derive their legitimacy from their having been popularly elected. Examples include Parliament and the Cabinet. However, other institutions derive their legitimacy through other means. NMPs, for example, are representative of various communities, though not elected. The same can be said of certain members of the House of Lords in the UK. The Judiciary derives its legitimacy partly through its expertise and the soundness of its logical reasoning. The same may be said of the public servants who assist the Cabinet. 

      Moreover, some institutions derive their legitimacy from their not having been popularly elected. An example is the Judiciary. If the Judiciary were popularly elected, there would be a risk that it would make decisions motivated by populism instead of showing strict fidelity to the law, rather than providing a counter-majoritarian check on the elected branches of the state.

      Can it then be said that the Judiciary, and judicial review, are non-democratic? No. The classic counter-argument is that Judiciary applies the law, which is made by the democratically-elected Legislature. Of course, that is not entirely true. Some of the law is common law made by the Judiciary. But the democratically-elected Legislature could overrule that law if it wished to. Moreover, in choosing what areas in which to make law and how extensively to do so, the Judiciary exercises a sense of responsibility and sensitivity to the limitations of its own role, such as its lack of expertise in certain specialised areas and its lack of democratic mandate (which is especially significant when it comes to potentially contentious issues of public policy).

      Could we next argue that, because of the possibility of such a legislative override, the Judiciary’s efforts are ultimately futile? More generally, can we argue that all checks and balances are ultimately useless because they can be overridden?

      As a preliminary point, the mere fact that the Legislature has reversed the effect of a judicial decision does not mean that the separation of powers is being violated, or that the Judiciary is unable to serve as an effective check on the elected branches of the state. Take, for example, the case of Lam Leng Hung. The Court of Appeal held that directors of a company are not subject to a more onerous punishment than others who commit criminal breach of trust. The Government has suggested that it intends to get Parliament to change this. Does this render the Court of Appeal’s decision useless? Far from it: the judicial system is working as it should. The Court of Appeal has pointed to a lacuna in the law and recognised that the Legislature should do something about it, given that the law which the courts are currently bound to apply is deficient.

      But suppose that the Legislature passed a law to override a court’s decision for some improper motive, such as greed. It is, in theory, possible for the Legislature to do so while not falling foul of the fundamental liberties defined in the Constitution or breaching the separation of powers. In fact, the Legislature might theoretically even be able to, using the power in Article 5 of the Constitution, modify the Constitution so as to abolish the Supreme Court, transforming Singapore into a dictatorship!* But this does not mean that nothing stops the Legislature from doing so. Let us put aside the fact that the Legislature (and Executive) presently consist of persons who would not countenance doing such things, for a future Legislature might well not. The more important point is that, in doing these things, the Legislature would face the glare of the public eye, which watches the news on TV, the newspapers, and Hansard. The public would then make its ire known through the ballot-box. One may retort that the Legislature could abolish voting too. That would bring us from the realm of law to the realm of political science, for, while I sincerely hope and trust that nothing of the sort will ever happen in Singapore, historical examples have shown that an extreme case of legislative tyranny can lead to the public making its ire known in other ways as the legal system fails to fulfil its functions. To put the point another way, it is in the interests of the legislature and executive to comply with the checks and balances which our legal system imposes upon it; for if they do not, they may face threats far greater than those posed by any court.

      But nothing of the sort is at all likely in Singapore, for the Singaporean legislature and executive have proven not to come into open defiance of the Judiciary. To the contrary, consider Tan Seet Eng, in which the Court of Appeal declared that Tan’s detention had been unlawful. The Legislature and Executive did not respond in a combative manner. To the contrary, the Ministry of Home Affairs explicitly stated that it “respects and accepts the Court of Appeal’s judgment”.  While the Ministry issued a fresh detention order against Tan, it did so in accordance with the law as pronounced by the Court of Appeal; and it – on its own motion – freed three other detainees on the grounds that, according to the law as pronounced by the Court of Appeal, it would be unlawful to continue to detain them. (For more details and commentary, see Swati Jhaveri, ‘Localising Administrative Law in Singapore: Embracing Inter-branch Equality’ (2017) 29 SAcLJ 828, [16].)

      We have seen the theme of publicity as a check and balance from Professor Kevin Tan’s study chronicling the steps that President Nathan took to publicly justify his decision to allow the Government to draw down on past reserves in 2009.  We will return to the theme of publicity as a check and balance as we continue the course. We will, for example, be exploring the view that judicial review has the effect of calling upon the executive to justify its actions. Whether those actions are lawful or not, the executive’s actions will be cast in the light of publicity and shown for what they are: patently unlawful; patently lawful; ostensibly arguably questionable but, on closer inspection, patently desirable; or lawful but highly undesirable.

      We began the course by asking what ‘the state’ is. Some of you said that the people are part of ‘the state’. I stated that we typically think of the people as being subject to the power of ‘the state’, rather than themselves being ‘the state’. But when legal checks and balances fail, political checks and balances wielded by the people – the ballot box; the press – kick in; and ‘the state’, if found wanting, must pay the political price.

      * It is not a foregone conclusion that this would be legally possible, as we will see when we study the ‘basic structure’ doctrine. But if a Legislature were hell-bent on doing something drastic as abolishing the Judiciary, surely they would not consider that a judicial decision would stop them.

    6. Did E v Secretary of State for the Home Department [2004] EWCA Civ 49 change the law on judicial review for error of precedent fact?

      I do not think that the criteria in E are relevant to errors of jurisdictional fact (precedent fact).

      What E purports to do is to describe a situation in which a decision may be “set aside on grounds of mistake of fact” (at [64]) — but this will not always be possible; it will only be possible when certain criteria are met (at [66]). These criteria are meant to explain the law in a rational way that avoids two possible extreme views of the law. The first view is that the only errors of fact which the court can correct are errors of precedent fact (at [53]: “where the fact is a condition precedent to an exercise of jurisdiction”). The second view is the view in Tameside, which we have argued in class would, taken to its logical conclusion, give the court very wide powers to intervene on the grounds of all kinds of alleged errors of fact (at [54]). What the criteria in E aim to do is to provide an account of the law that explains why the courts’ powers to correct errors of fact are wider than the first view seems to suggest, but not as wide as the second view (taken at face value) seems to suggest.

      So E does not change the principle which is common to both extreme views, namely, that the courts can always intervene to correct an error of jurisdictional fact (precedent fact).

      Indeed, it has long been the law in England and Singapore — long before E was decided — that “where the exercise of executive power depends upon the precedent establishment of an objective fact, the courts will decide whether the requirement has been satisfied” (Khawaja at 110F). In other words, the court can always intervene to correct an error of jurisdictional fact (precedent fact) — it is not the case that the court can only do so in cases where the criteria in E are made out. What E does is to state a framework to determine when the court can intervene on the ground of an error of non-jurisdictional fact (error of fact other than error of precedent fact).

      Of course, this leaves the question of how to tell whether a question of fact is a question of jurisdictional fact (precedent fact) or not. This is a question of statutory interpretation; the case law, such as Khawaja (importance of fundamental liberties) and A (whether the question of fact is an “objective” one, as opposed to a matter of judgment) provides us with some guidance.

    7. Some notes on the rule of law, constitutional supremacy, and the content of the constitution

      • There are various possible definitions of the rule of law. This should prompt us to wonder whether the rule of law is (or risks becoming) a stand-in, label, or cloak for some underlying principles which deserve to be specified in more detail. (Thio, A Treatise on Singapore Constitutional Law, at [03.027]-[03.028], describes the rule of law as “both a fundamental legal concept and a political slogan”, whose “content remains elusive”. This gives rise to the case for adopting a ‘thin’ definition of the rule of law: the rule of law would then serve the useful purpose of providing at least some minimum standards to which the law is to adhere. (See Thio at [03.042].)
      • That said, there are degrees of ‘thickness’ which a definition of the rule of law may have. Consider, for instance, Bingham’s building on Raz’s definition of the rule of law; and Barber’s statement that, for the ‘thin’ definition to be of any use, there must be a minimum standard of living in society.
      • One might also argue that the rule of law requires not just that the law and institutions have certain features, but also that society has certain features, such as an acceptance of the authority of the law. Indeed, one may argue that law whose authority is not accepted by either legal subjects or power-holders is not law at all: see my ridiculous example [presented in class], adapted from Frederick Schauer, ‘Amending the Presuppositions of a Constitution’, ch. 7 in Sanford Levinson (ed), Responding to Imperfection: The Theory and Practice of Constitutional Amendment (Princeton University Press, 1995), at pp 152-154, of the [fictitious] ‘Constitution of the Republic of Singapore’ that declares me the “Supreme Leader”.
      • If we start from the idea that the ‘rule of law’ is a good thing that the courts should uphold, then one’s views on what the rule of law is (which, in turn, have an impact on one’s views on what law is) have an impact on the content of case law. Consider, for example, Ong Ah Chuan. The Privy Council was unwilling to say that ‘law’ simply means something passed by the Legislature. But the Privy Council also did not say that, to be truly ‘law’, an enactment must be (for example) rational or based on relevant considerations; it only singled out the rules of “natural justice”, and not any other rules, to be built into the definition of “law”.
      • Tan Eng Hong held that even laws which predate the commencement of the Constitution may be held void if they are inconsistent with the Constitution. The view I offered was that this derives from the first half of Art 4: “This Constitution is the supreme law of the Republic of Singapore”. Some of you disagreed with this reading, considering that such a reading would render the second half of Art 4 otiose. Such a view is fascinating, because it would mean that the Court of Appeal located its doctrine of constitutional supremacy in a source outside Art 4.
      • In considering the issue of implied constitutional rules, I would suggest that you consider what you have learned in other courses about implication generally. For example, does the Constitution contain what the Court of Appeal in Sembcorp called a ‘true gap’ with regard to the issue of voting?
      • The idea that the authority of the Constitution ultimately rests on the political fact of its acceptance would explain why our Constitution is considered to have the authority of a constitutional document even if part of it (the RSIA and Constitution (Amendment) Act 1965)) was passed by a body which had no legal power to pass it. (But cf. Chan’s views at (2017) 29 SAcLJ 619, [68], that the Legislative Assembly of Singapore did have the power to pass the RSIA.)
      • “All powers have legal limits.” What is a limit? Consider, for example, a hypothetical constitutional provision that says “X has supreme executive power without limitation, and his decisions shall not be called into question in any court.” This seems to say that there are no limits to X’s powers. But it does tell us, for example, that the executive powers (if any) wielded by people other than X are limited. True, this is an instance of what Thio calls a “[c]onstitutio[n] without constitutionalism” (at [01.100]ff); but even then, as Thio says, such a constitution would at least “guid[e] political action through established procedural channels” (at [01.101]), and thereby constrain X to make his acts (no matter how despotic they are) take certain forms.
    8. Interpretation and implication

      Interpretation and implication are concepts relevant to all documents, not just contracts/statutes/Constitutional provisions. That is why Menon CJ’s 2013 Singapore Law Review lecture was titled ‘The Interpretation of Documents…’ <https://www.supremecourt.gov.sg/docs/default-source/default-document-library/media-room/the-interpretation-of-documents—saying-what-they-mean-or-meaning-what-they-say.pdf>. He speaks at [4] of ascribing meaning to a “text”; and at [60]-[61] read with [58] he says that the Wentworth Securities test for implying words into a statute is similar to the test for implied terms in contract law.

      Justice Kirby, too, takes a similar view: the title for his 2002 lecture, ‘Towards a Grand Theory of Interpretation: The Case of Statutes and Contracts’ <http://www.hcourt.gov.au/assets/publications/speeches/former-justices/kirbyj/kirbyj_statute.htm>, says it all. So do several other writers. That would certainly explain why our Professor Goh Yihan [as he then was], an expert in contract law, wrote a chapter in a book about constitutional interpretation: Goh Yihan, ‘The interpretation of the Singapore constitution: towards a unified approach to interpreting legal documents’, ch 10 in Jaclyn L Neo, ed, Constitutional Interpretation in Singapore (Routledge, 2017).

    9. “Generous” interpretation of rights provisions in the Constitution

      Minister of Home Affairs v Fisher [1980] AC 319 (UKHL) 328 says that the courts should give the Constitution “a generous interpretation… suitable to give to individuals the full measure of the fundamental rights and freedoms referred to” therein.

      As a number of commentators have observed, the Singapore courts have not engaged in detail with the ‘full measure’ principle. See, for example, Victor Leong, ‘Tan Cheng Bock v AG: Is Constitutional Interpretation As Settled As It Seems?’, Singapore Law Gazette (January 2018) <https://lawgazette.com.sg/feature/tan-cheng-bock-v-ag-is-constitutional-interpretation-as-settled-as-it-seems/> — he says that “the principle of generous interpretation… appears to have all but disappeared in the recent jurisprudence”.

      Fisher concerned a provision in the Constitution of Bermuda which stated that “a person shall be deemed to belong to Bermuda” (and, hence, to have the right to live there) if, inter alia, that person was a “child” of the wife of a person who “possesses Bermudian status”. A woman had four illegitimate children, then married a Bermudian. Did the children “belong to Bermuda”? The state argued that they did not because the word “child” in that provision did not include an illegitimate child.

      There was a canon of statutory interpretation according to which “child” was presumed to mean a legitimate child (only). This presumption could be rebutted depending on the context of the statute.

      But the Privy Council said that that interpretive presumption did not apply at all. Even if it applied to statutes generally, it would not apply to the Constitution because:

      “A Constitution is a legal instrument giving rise, amongst other things, to individual rights capable of enforcement in a court of law. Respect must be paid to the language which has been used and to the traditions and usages which have given meaning to that language. It is quite consistent with this, and with the recognition that rules of interpretation may apply, to take as a point of departure for the process of interpretation a recognition of the (a) character and origin of the instrument, and to be guided by the principle of giving full recognition and effect to those fundamental rights and freedoms (b) with a statement of which the Constitution commences.” (emphasis added; indicative letters added in brackets)

      Note the two phrases I have bolded. (a) suggests that the method of constitutional interpretation is different from statutory interpretation – more “generous” – because of the very idea of constitutional rights. This is Thio’s argument. (b) suggests that the method of constitutional interpretation is different from statutory interpretation because of the following statement of principle in section 1 of the Constitution of Bermuda:

      “1. Whereas every person in Bermuda is entitled to fundamental rights and freedoms of the individual, that is to say, has the right, whatever his race, place of origin, political opinions, colour, creed or sex, but subject to respect for the rights and freedoms of others and for the public interest, to each and all of the following, namely: – (a) life, liberty, security of the person and the protection of the law: (b) freedom of conscience, of expression and of assembly and association; and (c) protection for the privacy of his home and other property and from deprivation of property without compensation, the subsequent provisions of this Chapter shall have effect for the purpose of affording protection to the aforesaid rights and freedoms subject to such limitations of that protection as are contained in those provisions, being limitations designed to ensure that the enjoyment of the said rights and freedoms by any individual does not prejudice the rights and freedoms of others or the public interest.”

      According to the Privy Council, the significance of this passage, which corresponds to point (b), was that:

      “Prima facie, the stated rights and freedoms are those of “every person in Bermuda.” This generality underlies the whole of Chapter I… When one is considering the permissible limitations upon those rights in the public interest. the right question to ask is whether there is any reason to suppose that in this context, exceptionally, matters of birth, in the particular society of which Bermuda consists, are regarded as relevant.”

      Of course, point (a) would apply in Singapore, but not (b).

      The Privy Council went on to say that “child” did not mean just “legitimate child” because of, inter alia, the wording of the constitutional provision in question. The constitution provision said that a person who “(a) possesses Bermudian status;… (c) is the wife of [such a person]; or (d) is under the age of 18 years and is the child, stepchild or child adopted in a manner recognised by law of a person to whom any of the foregoing paragraphs of this subsection applies” was a person who “belong[ed] to Bermuda”.  According to the Privy Council:

      “In their Lordships’ opinion, paragraph (d) in its context amounts to a clear recognition of the unity of the family as a group and acceptance of the principle that young children should not be separated from a group which as a whole belongs to Bermuda.”

      So the Privy Council’s conclusion was based on the preamble in the Constitution of Bermuda, as well as the particular wording of the constitutional provision being interpreted. The Privy Council went on to add that it would be possible for a Constitution to use the word “child” to refer only to a legitimate child, if the drafter had used “express words” to “spell it out”.

      Therefore, the foundation of the idea of “generous” interpretation appears to rest, at least in (large) part, on some particular words in Bermuda’s Constitution – which are absent from Singapore’s Constitution. Of course, the counter-argument is that point (a) still remains.  

      Now let’s put that aside and focus on exactly how the principle of ‘generosity’ applied.

      In Ong Ah Chuan, ‘generosity’ meant interpreting the word “law” to mean something more than ‘something that has been passed by the Legislature’. As I mentioned in class, I am more sceptical of the notion that the right to silence is necessarily not one of the ‘fundamental rules of natural justice’, and I am therefore sceptical about whether Mazlan was correctly decided.

      On the other hand, I see considerable force in the restrictive reading of Art 9, and do not necessarily see this as conflicting with the case law on ‘generosity’ and ‘full measure’.

      In Fisher, ‘generosity’ meant not applying the normal rules on how to interpret the word “child” in statutes, instead applying the plain ordinary meaning of the word.  (This is similar to what happened in Edwards.) In other words, the idea was to give the word “child” the meaning, out of all possible meanings that would cohere with the context of the provision, that would give the fullest possible measure to the applicants’ rights. This principle would not help one who sought to argue that “life” in Article 9 includes the freedom to make a contract, etc., because (goes the argument) that does not fall within the ordinary meaning of the word ‘life’.

      As for “liberty”, the problem lies in Articles 9(2) to 9(4) of the Constitution, which provide for safeguards against unlawful detention (see Lim Meng Suang v AG [2015] 1 SLR 26 (CA) [46]ff). Fisher doesn’t say that words are to be given the broadest possible meaning they can bear in isolation; instead, it emphasis interpretation in context. That is why the Privy Council was convinced that, because the various limbs of the Bermudan constitutional provision in question, taken together, gave citizenship to husbands, wives, and children, the purpose was to uphold the “unity of the family as a group”. By contrast, the various sub-provisions of Art 9, taken together, suggest that the point of Art 9 is to protect our rights to be alive and not to be locked up, and not to protect our freedom to act as we please generally.

      A court could, in theory, take the approach seen in India’s law, where “life” has been interpreted to include everything from education to health to privacy. But then where would it end?  The justification for taking this Indian approach may well be the “unique social, political and legal circumstances” of India: Lim Meng Suang (CA) at [48]. But I don’t think the justification can be the ‘full measure’ principle. ‘Full measure of a right’ assumes that we have a clear definition of what the ‘right’ is, and then we are protecting the fullest possible measure of that right. But, by this logic, the ‘full measure’ principle would not justify the courts, in interpreting right A, seeking to give full measure to right B, which is another right.

      As for the ‘purposive interpretation’ argument, the simplest possible counter-argument is that the pronouncement in Ong Ah Chuan/Fisher about interpreting the Constitution with “principles of interpretation of its own” is a common-law rule, which can therefore, one would think, be abrogated by statute. But that is not necessarily a bad thing, because I do not think that purposive interpretation is necessarily at odds with ‘generous’ interpretation. This is because one could always say that the purpose of constitutional rights provisions is, at least on one level, to be ‘generous’. I also note that the cases which have focused on the idea of a ‘purposive’ interpretation of the Constitution – such as Const Ref 1/1995, Tan Cheng Bock – involved the interpretation of provisions which were not rights provisions. I have set out some thoughts (somewhat) related to this issue in an old piece, ‘Developments in the law on constitutional and statutory interpretation’, Singapore Law Watch Commentary (Issue 1/Sep 2013). My conclusion then was not entirely clear, but one point I sought to make was that there has been no clear reconciliation between the Fisher approach and the approach in s 9A of the Interpretation Act. On Victor Leong’s account, five years later, the interaction between the two has still not been settled.


      When applying a ‘purposive’ interpretation of an enactment, note the potential issue that an enactment may (a) pursue multiple purposes; (b) pursue purposes at different levels of generality. See Tan Cheng Bock  v AG for a view of how the Singapore courts address this issue. Bear this in mind when addressing the idea of a ‘generous’ interpretation.

      The view in Ong Ah Chuan, which Thio explains in A Treatise on Singapore Constitutional Law, is that a ‘generous’ interpretation is warranted because the purpose of all rights provisions generally – Thio backs this argument up by reference to the 1957 Reid Commission report.

      I would further invite you to consider if we can come up with a more nuanced picture of what exactly the ‘purpose’ of all our constitutional rights provisions is. In general, why might a legal system single out certain rights for protection? To my mind, there can be several possibilities:

      1. For the benefit of individuals: the designers of that system consider that every individual should have and enjoy those rights because they are valuable to individuals.
      2. For the benefit of society: the designers of that system consider that it would be good for society for every individual to have those rights. As Thio points out in ‘Protecting rights’ at 199, the Reid Commission thought that the fundamental liberties included in the Federal Constitution of Malaysia were “essential conditions for a free and democratic way of life”. Note also that the High Court of Australia, in Australian Capital Television (1992) 177 CLR 106 (HCA), justified the implied freedom of political communication on the ground that democracy requires it.
      3. To limit the powers of the state: the designers of that system, motivated by a general wish to prevent tyranny by the state, have noted that those fields are those in which the state’s powers are most open to abuse, and accordingly have singled out those areas for specific protection of persons against the state.
      4. To limit individual rights: as one of you pointed out, perhaps the rights singled out for protection are those which could potentially lead to chaos if everybody enjoyed them to the fullest extent; the rights provisions seek to manage those rights by limiting them. Consider, for example, that Malaysia’s Peaceful Assembly Act 2012 claims, in section 2, to aim to “ensure… so far as it is appropriate to do so, that all citizens have the right to organise assemblies or to participate in assemblies, peaceably and without arms; and… that the exercise of the right to organise assemblies or to participate in assemblies, peaceably and without arms, is subject only to restrictions deemed necessary or expedient in a democratic society in the interest of the security of the Federation or any part thereof or public order, including the protection of the rights and freedoms of other persons.” (On the constitutionality of this Act, see Public Prosecutor v Yuneswaran a/l Ramaraj [2015] 6 MLJ 47 (CA)).

      Which of these models (if any) describes rights protection in Singapore?

    10. Some notes on originalism and dynamism in constitutional interpretation

      We had an interesting discussion on originalism and dynamism in constitutional interpretation [in class one day]. For example, we considered whether the words “religion” in the Constitution ought to be interpreted to refer to:

      • (a) only those religions which existed at the time the Constitution was drafted;
      • (b) only religions sharing the characteristics of the religions which existed at the time the Constitution was drafted; or
      • (c) anything which falls within the modern understanding of the word “religion”.

      (I will use “(a)”, “(b)”, and “(c)” to refer to similar arguments in general terms, not necessarily specifically in the context of religion.)

      In this post, I wish to summarise our discussion on these three points; highlight a recent case which adds to this debate; and then, in response to some of your queries, end by distinguishing between two types of question of constitutional interpretation, which might call for different types of approach.

      I.

      The argument for saying that (c) is wrong is Goldsworthy’s point that, otherwise, the words in the Constitution will be reduced to “almost empty husks” – vessels into which the courts may pour any meaning they like, without giving sufficient weight to the democratic mandate of those who first drafted the Constitution.

      Another argument against (c) is Scalia’s argument that the point of the Constitution is to “embed certain rights in such a manner that future generations cannot readily take them away”, hence the meaning of the Constitution must not be allowed to change (Antonin Scalia, A Matter of Interpretation: Federal Courts and the Law (1997) at 40, cited in Michael Kirby, ‘Constitutional interpretation and original intent: a form of ancestor worship?’ (2000) 24 Melbourne University Law Review 1, 2). But this counter-argument, in my view, is not complete, because it says nothing about interpretations that would expand rights.

      The argument for saying that (a) is wrong is the fact that the drafters of the Constitution chose to use the general word “religion”. They could have listed specific religions singled out for protection (as in the Constitution of Ireland before the Fifth Amendment in 1972, which provided that “The State recognises the special position of the Holy Catholic Apostolic and Roman Church as the guardian of the Faith professed by the great majority of the citizens.” and “The State also recognises the Church of Ireland, the Presbyterian Church in Ireland, the Methodist Church in Ireland, the Religious Society of Friends in Ireland, as well as the Jewish Congregations and the other religious denominations existing in Ireland at the date of the coming into operation of this Constitution.”); but they did not.

      In Re Wakim; ex parte McNally, the High Court of Australia chose option (b) – the middle option. In essence, it said (at [42]), that the court can expand the denotation of words (their application) but not their connotation (their meaning). We can see this in action in Jumbunna Coal Mine.

      II.

      Now, in Public Prosecutor v ASR [2019] SGCA 16, the Court of Appeal has expressed a similar view. (Although this case involved the interpretation of a statute and not the Constitution, the Court of Appeal said that it was applying the framework set out in Tan Cheng Bock and Lam Leng Hung (at [75]ff) – a framework which applies to the interpretation of the Constitution too.) ASR dealt with the interpretation of section 83 of the Penal Code, which provides: “Nothing is an offence which is done by a child above 7 years of age and under 12, who has not attained sufficient maturity of understanding to judge of the nature and consequence of his conduct on that occasion.” The accused was chronologically 14 years old, but had had a mental age of between 8 and 10 years. Did section 83 apply to him? Did the word “age” include “mental age”?

      The Court of Appeal, at first glance, followed option (a): “to determine the ordinary meaning of the word ‘age’ in s 83, the court must begin by considering what that word was understood to mean at the time the Penal Code was adopted, that is, in 1872” (at [79]). But the Court of Appeal then noted the problem with this argument:

      “since the concept of mental age did not exist at the time the Penal Code was adopted, it is arguably also not possible to say that s 83 was not intended to refer to that concept, given that such an intention could not have been formed at that time in the first place. By way of analogy, if the question in Lam Leng Hung had been whether a Bitcoin broker was an ‘agent’ within the meaning of s 409, it would not have sufficed to say that because Bitcoin brokers did not exist at the time the Penal Code was adopted, the Legislature could not have intended s 409 to apply to them…” (at [80])

      Therefore, the Court of Appeal chose option (b):

      “Instead, the court would have had to consider whether the ordinary meaning and concept of the word ‘agent’ in s 409 at the time the Penal Code was enacted could logically extend to Bitcoin brokers. And the court might have concluded, for example, that because a Bitcoin broker is a fundamentally a kind of professional agent, he would be caught by s 409.

      In short, where the court is considering new phenomena which did not exist at the time the provision in question was adopted, it must consider whether the ordinary meaning of the provision at the time can logically extend to the new phenomena. Here, we considered that the ordinary meaning of the word “age” in s 83 did not logically extend to the subsequently developed concept of mental age. The former is fundamentally a measure of time, specifically, the length of time that a person has lived since birth, whereas the latter is not. It is instead a measure of a person’s cognitive ability with reference to a specific mathematical formula… The two concepts are therefore qualitatively different, and cannot be reduced to a common denominator in the way, for example, that a Bitcoin broker and a factor might both be regarded as professional agents for the purposes of s 409. In our judgment, therefore, the word “age” in s 83 does not include mental age.” (at [80]-[81])

      III.

      Can it nonetheless be argued that option (c) ought to be applied when interpreting the Constitution, given that the Constitution, unlike ordinary legislation, is designed to endure over time, etc.? In Review Publishing, the courts commented on such an option, but in the context of a different kind of question of constitutional interpretation. The question was not one of the meaning and scope of a particular word or phrase, but rather one of the effect of an entire provision, namely, “[t]he balance between constitutional free speech and protection of reputation”. The Court of Appeal chose (c), with the gloss that it demanded evidence that the modern understanding of the “balance… between constitutional free speech and protection of reputation” had changed from when the Constitution was drafted, failing which (c) would essentially collapse into (a). All this was obiter. But what is interesting is that the Court of Appeal conceded that the effect of the Constitution could, in theory, change in the light of a “change in our political, social and cultural values”, notwithstanding that the text of the Constitution did not change.

    11. Equality in sentencing

      A student asked: Suppose the Public Prosecutor argues that A and B should get life imprisonment. However, A is older than B, so the sentence of life imprisonment would have a greater impact on A. How does the concept of equality fit into all this?

      My response:

      What is the role of the PP?

      In this situation, the PP’s ‘decision’ to make various submissions about the sentences has no legal effect.

      That is because the arguments that the PP chooses to make in court don’t have a direct effect on the punishment imposed by the offender. The decision as to the sentence is the court’s alone; the court is not bound to impose a certain sentence (or a sentence from a certain range) merely because the Prosecution submits that such a sentence would be appropriate. See Janardana Jayasankarr v PP [2016] SGHC 161 at [10]-[12]; AG v Wham Kwok Han Jolovan [2019] SGHC 111 at [88] (not disturbed on appeal: [2020] SGCA 16).

      (Similarly, the court is not bound to impose the sentence recommended by, say, a probation suitability report: Praveen s/o Krishnan v PP [2017] SGHC 324 at [65]-[66].)

      Because of this, to the extent that the chain of causation has been started by the PP, it is broken by the court’s independent sentencing decision. In other words, if anyone can be accused of violating Art 12(1), it is not the PP, but the court (or possibly the Minister for Home Affairs – see below).

      By the way, while the Prosecution of course has discretion as to what submissions on sentencing to make, we typically don’t call this “prosecutorial discretion”; we use that term to refer only to the prosecution’s discretion as to what charges, if any, to tender against someone.

      What does the ideal of equality before the law require of the court?

       Same offence – same sentence for both offenders…

      The court gives effect to Art 12(1) in sentencing through what is known as the “parity principle” (see e.g. Chong Han Rui v PP [2016] SGHC 25, [1] – refers to parity between two co-offenders in the same criminal enterprise) and the general principle of “consistency in sentencing” (see the source cited in Lim Bee Ngan Karen v PP [2015] SGHC 183 at [30]). (That said, the courts tend not to mention Art 12(1) by name when discussing these principles, but that doesn’t change the analysis.)

      That said, in applying these principles, the courts do take into account that the two offenders might not be equally situated: for example, there may be a “relevant difference in [the two offenders’] responsibility or their personal circumstances”: see Muhamad Azmi bin Kamil v PP [2022] SGCA 68, [25]. In short, the law is that offenders who are equally situated should receive the same sentence; but offenders may not be equally situated, for various reasons.

      … or, rather, same impact on both offenders?

      The courts do also consider that a sentence of a certain length may have a disparate impact on people with different ages or different health conditions. This is known as the “principle of equal impact”: see M Raveendran v PP [2021] SGHC 254 at [15] onwards; Chew Soo Chun v PP [2016] SGHC 6 at [59] onwards.

      So life imprisonment for two prisoners, with a different impact on each of them, may be justified or not depending on the sentencing objective. I would however add that the sentencing objective for each of them could well be different, itself because of differences between the offenders’ circumstances etc.

      (Of course, the court is ultimately constrained by the range of possible sentences set out in the legislation.)

    12. Various words and phrases relating to the powers of the state

      ‘Separation of powers’

      When discussing the concept of separation of powers, it would be useful to distinguish between (a) saying something about the fact that the powers are separated; (b) saying something about what exactly those powers are. For example, some might seek to argue that judicial review of administrative action on the ground of irrationality is a violation of the principle of separation of powers. But what does this mean? If, for example, there is a legal rule that the executive has no power to be irrational, then the court’s examining decisions and quashing them if they are found to be irrational is no breach of the separation of powers.

      ‘Institutional competence’

      When discussing institutional competence, it may be useful to distinguish between absolute and relative institutional competence. Some of you asked: why is it said that the courts ought to abstain from judicial review in areas which engage polycentric socio-economic considerations, while the courts have no hesitation about developing the law on (say) vicarious liability, which has potentially significant impact on all manner of areas of life (e.g., to take just one example, the demand for insurance)? The answer may be that, in the field of vicarious liability, the courts are free to develop the law as they see fit, because, if not them, who else? The same cannot be said of areas where there is a statute regulating the topic in question, because the answer to ‘who else?’ is ‘the Legislature’.

      ‘Accountability’

      Various writers have made the point that ‘accountability’ does not merely mean giving an account. Anne Davies, for instance, writes of accountability as a process from someone (who?) setting standards, to obtaining an account (in what level of detail?), to someone (who?) judging the account against the standards, to some response to the findings. For more details, you may wish to consider:

      1. Anne Davies, Accountability: A Public Law Analysis of Government by Contract (OUP, 2001) 
      2. Nicholas Bamforth and Peter Leyland (eds), Accountability in the Contemporary Constitution (OUP, 2013) — available at the National Library
      3. Various works by Dr Ellen Rock, which are listed on her profile page: https://www.uts.edu.au/staff/ellen.rock
    13. Should it be compulsory for opposition MPs to be present when Parliament sits?

      Someone asked: what if no opposition MP turns up to Parliament? Would that make Parliament’s decisions less democratically legitimate?

      My response:

      There comes a point when certain matters are left up to the political process. In your situation, the law would already have taken a step toward protecting democracy, namely, giving opposition MPs (/NCMPs/NMPs) the opportunity to be there. The law does not go further and require them to be there; they are at liberty to be there or not to be there, and to suffer the political consequences accordingly. Indeed, if the law were that certain/all MPs must turn up or be punished, I can imagine that very few people would want to be an MP; and if the law were that MPs may turn up, but that Parliamentary proceedings cannot proceed without a certain number of MPs from a certain party, then those MPs from that party could effectively hold Parliament to ransom by threatening to simply refuse to turn up.

      Something similar can be said of the fact that there is no law requiring that there be any opposition MPs at all, or that someone contest every seat. The law has already done enough to uphold democracy by creating the condition under which people have the opportunity to stand in elections.

    14. When may the Executive withhold evidence on the ground of the “public interest”? 

      Strictly speaking, this is the wrong question to ask. “[I]t is important to remember that the decision ruling out such documents [or evidence more generally] is the decision of the Judge… It is the Judge who is in control of the trial, not the Executive”: Duncan v Cammell Laird [1942] AC 624, 642. So the question is: when may the Executive attempt to withhold evidence from production in court on the ground of the “public interest”, and when may the court allow such an attempt?

      There are at least two relevant legal rules.

      First, there is section 126(1) of the Evidence Act: “No public officer may be compelled to disclose communications made to him or her in official confidence when he or she considers that the public interest would suffer by the disclosure.” As we should know by now, the words “when he or she considers that…” do not mean that the public officer can refuse to disclose evidence of the communications on his/her own say-so. To the contrary, “the relevant public officer does not have an unfettered discretion to determine that the public interest would suffer by the disclosure of the relevant communications”: Chiu Teng Enterprises Pte Ltd v AG [2011] SGHC 77, [8]. Therefore, “the court can review the legality of a determination of public interest under s 126(1) on the usual grounds known to administrative law”: ibid. (See also Mah Kiat Seng v AG [2021] SGHC 202, [96].)

      Note that s 126(1) refers to “communications” made to a public officer: see e.g. Mah at [15]-[16].

      Note also that the state may waive the right to rely on s 126(1): see, for example, Mah at [91]-[95]. In that case, certain footage from a police officer’s body-worn camera showed someone making a report to that officer. That was a recording of “communications” made to a “public officer”, so s 126(1) applied. However, the AG “unilaterally [gave] a limited waiver of s 126”; it was willing to disclose the footage to Mah, but subject to the restrictions set out at [95] (for example, that the complainant’s identity be redacted).

      Second, there is common-law public interest privilege: “Documents which ought otherwise to be disclosed under the rules of civil or criminal procedure may only be withheld if the court concludes that the public interest against disclosure outweighs the public interest in the administration of justice.”: Mah Kiat Seng at [80]. This appears to be a flexible doctrine, in that the court will consider whether the reasons weighing against disclosure can be satisfied through the use of “appropriate safeguards”: at [99]. In that case, Mah sued the police in respect of alleged torts committed by various police officers at a police lock-up. The AG, representing the police, submitted — and the court agreed — that there was a public interest in withholding certain CCTV footage of certain areas in the lock-up due to “security risks” (at [99]). Yet the court ordered that that footage be disclosed to Mah, though in the following limited manner: Mah would not be able to take copies of the footage, but instead would be allowed to view the footage at the police station (at [104]). Note that the court did consider alternative possible safeguards, such as blurring out certain parts of the footage (at [100]-[101]).

      The Attorney-General may intervene in civil proceedings to assert public interest privilege: ARW v Comptroller of Income Tax [2019] 1 SLR 499 (CA), cited in Mah at [56]ff.

    15. Can the President be compelled to give evidence in court?

      Can the President of Singapore be compelled to give evidence in a witness in court proceedings? Answering this question has brought me down a delightful rabbit-hole, the contents of which I would now like to share with you.

      It would be useful to begin with the law in the UK. According to Halsbury’s Laws of England, there are two relevant rules. First, “A monarch has immunity at common law from being compelled to give evidence in court”.1 Second, “the monarch may not give evidence in her own cause”2 – for example, the monarch cannot give evidence in cases involving alleged crimes against the monarch, such as treason.3

      Robin Callender Smith has questioned whether these are really accurate statements of the law in the UK. He has argued that they are not necessarily so, and hence has criticised the decision in the case of Edward Mylius.4 In that case (decided in 1911), Mylius was convicted of having committed criminal libel against King George V, by publishing material alleging that he had a secret second wife. The Lord Chief Justice, Lord Alverstone, agreed with counsel for the Attorney-General that the court could not issue a subpoena to compel the King to give evidence.5

      By the way, since then, Callender Smith has also called attention to the case of Paul Burrell. Burrell was a butler who was charged with stealing various items from members of the royal family. However, the prosecutors halted the prosecution after receiving a statement by the Queen that she had given those items to Burrell for safekeeping. Evidently, the prosecution took the view that the Queen could not be compelled to give evidence. Callender Smith is presently attempting, under the UK’s Freedom of Information Act, to get more details on the reasons for the prosecution’s view.6

      One might think that, assuming the law of the UK is as set out above, Singapore would have corresponding rules relating to the President. I sreiously doubt this. The foundation of the rules in the UK appears to be the idea that the courts are, formally speaking, Her/His Majesty’s courts, and the monarch cannot be compelled by his/her own courts. Moreover, the monarch would be free to ignore a subpoena, given that the monarch “can do no wrong”. These ideas are alien to Singapore. Singapore’s Constitution does not say that the President “can do no wrong”. This is unlike, say, Malaysia’s Federal Constitution, which states that the Yang di-Pertuan Agong and the Rulers “shall not be liable to any proceedings whatsoever in any court”,7 and which explicitly preserves the Rulers of the states’ “sovereignty, prerogatives, powers and jurisdiction”.8 Moreover, Singapore’s Constitution does not say that the courts are “the President’s courts”. The President is not above the courts, not vice versa. Finally, I know of no statutory provision stating that the President may not be compelled to give evidence in court; this is in contrast to jurisdictions like Brunei where such a provision exists.9

      But even if the abovementioned rules in the UK could in principle apply to the President of Singapore, they could not apply in a case where the President is willing to give evidence without being compelled to do so.

      Callender Smith’s article also quotes a letter by Sir Arthur Bigge, personal secretary to King George V, which suggests that the Government could “declin[e] to allow His Majesty to appear”.10 I do not know about the precise basis for this power in the UK (assuming it exists). This does, however, prompt the question: Could the Government, or anyone else, stop the President from giving evidence if the President is willing to do so?

      When sections 4A-4B of the Attorney-General (Additional Functions Act), which were recently inserted by the Courts (Civil and Criminal Justice) Reform Act 2021, come into force, the AG will have the power to intervene in civil proceedings in the public interest. The power to intervene is not new law; it reflects the common-law position that the Attorney-General can (with the leave of court) intervene in civil proceedings in order to assert “public interest privilege”.11 In the 2019 ARW case, the Court of Appeal used that phrase to refer to rule in s 126 of the Evidence Act; here, we are concerned with s 126(1) which provides: “No public officer may be compelled to disclose communications made to him or her in official confidence when he or she considers that the public interest would suffer by the disclosure.” One might ask: what if the President is nonetheless willing to disclose such communications, or to give any other sort of evidence, without being compelled to do so? I believe the answer may lie in the common-law doctrine of public interest immunity,12 which, it stands to reason, the AG would be entitled to intervene in civil proceedings to invoke.

      Finally, what if the AG and the President (and/or, for that matter, the Government) take different views on whether the doctrine of public interest immunity is to be invoked? I would submit that the position is as follows. It appears that public interest immunity may be invoked by an executive officer in his own right,13 so the AG might not have the power to overrule that executive officer and decline to invoke the doctrine. As for the opposite situation – the executive officer doesn’t want to invoke the doctrine, but the AG does – I would submit that the AG would have the final say, given that it is the AG and not the President whom the courts have described as being the “guardian of the public interest”14 in this sense. If I am right, then, in this latter case, any clash between the AG, the Government, and/or the President on when to claim public interest immunity is a matter to be resolved behind the scenes rather than in court.

      1. Halsbury’s Laws of England, vol 29 (Crown and Crown Proceedings, 2019), para 64. ↩︎
      2. Ibid. ↩︎
      3. Joseph Chitty, ‘A Treatise on the Law of the Prerogatives of the Crown’, 377, cited ibid. ↩︎
      4. Robin Callender Smith, ‘The Missing Witness? George V, Competence and Compellability and the Criminal Libel Trial of Edward Frederick Mylius’ (2012) 33 Journal of Legal History 209 <https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2037498>. ↩︎
      5. Ibid, SSRN page 24. ↩︎
      6. Robin Callender Smith, ‘Will we discover when the Queen can be required to give evidence in court?’, Inforrm, 9 July 2019 <https://inforrm.org/2019/07/09/will-we-discover-when-the-queen-can-be-required-to-give-evidence-in-court-robin-callender-smith/>. (It is interesting to note the comments by one Tony Adler on this blog post; Adler claims to set out situations in which members of the Royal Family appeared as witnesses in court.) ↩︎
      7. Federal Constitution of Malaysia, Art 32(1) and 181(2). Even then, there are exceptions to this: see Art 182-183. ↩︎
      8. Ibid, Art 181(1). ↩︎
      9. Supreme Court Act (Brunei) s 34(1), cited in Prince Jefri Bolkiah v State of Brunei Darussalam [2008] 2 LRC 196 (PC) [11]. That case cited the Mylius case, and was cited in Callender Smith, ‘The Missing Witness?’, above, fn 110. ↩︎
      10. Callender Smith, ‘The Missing Witness?’, above, at 23. ↩︎
      11. ARW v Comptroller of Income Tax [2019] 1 SLR 499 (CA) [24]-[28]. ↩︎
      12. Mah Kiat Seng v AG [2021] SGHC 202. ↩︎
      13. Ibid, [90]. ↩︎
      14. ARW (above) [24]-[25], cited in Mah (above) [58]. ↩︎

    16. Kable and Liyanage

      I teach students about the Liyanage case, which most likely represents the law in Singapore. To grab students’ attention, I preface it by showing students the legislation in the Kable case, which is targeted at one named individual. But that’s all I use the Kable case for. I tell students not to read Kable as it will distract them from the point I am making using Liyanage. To elaborate:

      A student asked: What in Kable says that judicial power extends to making orders that are dispositive of a person’s legal rights and obligations, and that the legislature cannot exercise this power?

      The answer is that Kable itself does not say so. None of the judges in that case thought that the Community Protection Act was unconstitutional merely because it targeted only one person. This was despite their not denying that, as Brennan CJ (dissenting) put it, Parliament’s passing the Act was an “exercise of a power that is not purely legislative”. To the contrary, Dawson J held that under the law of New South Wales (unlike in Ceylon), it is not impermissible for the legislature to exercise judicial power; he distinguished Liyanage on that basis.

      Note that this was so as a matter of the law of New South Wales. It is an instance of the fact that, for reasons that need not detain us, the principle of the separation of powers does not apply with full force at the state level, so the judicial power of New South Wales is not necessarily vested exclusively in the judiciary of New South Wales. At the federal level, on the other hand, the rule is — as it was in Ceylon, and is in Singapore — that the judicial power of the Commonwealth of Australia may only be vested in the judiciary of the Commonwealth (specifically, in courts created under Chapter III of the Constitution of Australia). For more on this, see Peter Gerangelos, ‘Separation of Powers in the Australian Constitution: Themes and Reflections’ (2017) 29 SAcLJ 903. He refers to what I have just described as the “‘first limb’ of the separation of judicial power” (at [8(a)]), and elaborates on it at [15]ff. 

      And that’s why I wanted to discourage everyone from reading Kable in full. All I wanted was to use the statute in issue in that case it as an example of something that, like the legislation in question in Liyanage, would, if passed in Singapore, be an unconstitutional usurpation by the Legislature of judicial power. The Singapore courts have suggested that it would be. We see this from Mohammad Faizal bin Sabtu v PP [2012] 4 SLR 947 (HC), according to which s 33A of the MDA is not unconstitutional because, “unlike the legislation in cases such as Kable… and Liyanage“, it does not have the “effect of prescribing the punishment to be imposed on particular individuals or of directing the outcome of pending criminal proceedings” (at [47]).

      By the way, the reason why the majority in Kable thought that the Community Protection Act was unconstitutional because it forced the courts to perform executive functions — as Toohey J put it, the court was required to “participate in the making of a preventive detention order where no breach of the criminal law is alleged and where there has been no determination of guilt” — whereas the courts can only perform judicial functions. (Specifically, as Gummow J explained, the judicial power to order imprisonment only encompasses ordering detention following a finding of guilt. Only the executive can order that someone be imprisoned for other reasons.)

    17. A law for which ‘nobody asked’

      In class, someone raised the point that laws respond to the needs of society. I responded that I could think of at least one example of a law for which, to use a colloquial turn of phrase, nobody asked. 

      What I had in mind was s 12A of the Registration of Births and Deaths Act 1937, which allows the Registrar-General of Births and Deaths to “refuse the entry or alteration of a child’s name in a register of births under this Act” on various grounds. These include situations when the name is “obscene or offensive” or “contrary to the public interest”.

      This provision was introduced by the National Registration (Amendment) Bill 2016. During debates on Parliament on that Bill, Mr Leon Perera asked: “Since every act of legislation should be a response to a real-world problem or a problem that is likely to happen, my question is: have there been a significant or an increasing number of persons wanting to register their names that might be deemed to contain an impermissible reference to a rank or an award or deemed to be offensive or contrary to the national interest?”  

      Mr Desmond Lee, who had introduced the Bill, gave examples of names that the Registrar-General would have the power to refuse to enter onto the register: “Those that used titles like ‘Sir’, ‘Professor’, ‘President’; names that are offensive or that use expletives. One other example could be names that could possibly be deemed derogatory or highly offensive to any particular race or religion.”

      Speaking for myself, I can see the force in allowing the prohibition of a name such as “Shit” or “F*** You”, notwithstanding the public interest in the freedom to choose one’s child’s name. I can also see the force in allowing the Registrar-General to prohibit someone from naming their child “Adolf Hitler” (a name which some listicles I found online say is prohibited in, among other places, Malaysia, though I have yet to find an independent source).

      That said, I do note that Mr Lee didn’t squarely address Mr Perera’s question of why the Government saw fit to introduce the Bill that would insert s 12A. I am driven to the conclusion that this is a law for which ‘nobody asked’. That does not mean that the law is useless; I, for one, would not want someone to be named “Little Bobby Tables” (I hesitate to type the full form of his name here.*) The fact remains, however, that it is not clear what, if anything, the law was introduced in response to.

      In the coming weeks, we’ll be studying what one can do in the event that one wishes to challenge the Registrar’s decision to refuse to enter a name in the register.

      * If you don’t get that joke, but you want to, let me know, or, better yet, ask someone who is actually qualified to answer well. Don’t worry; it’s not in our syllabus.

    18. The terms “anxious scrutiny”, “light touch”, etc.

      The terms “anxious scrutiny”, “light touch”, “super-Wednesbury”, and “sub-Wednesbury” are not terms of art in Singaporean case law. It is not wrong to use them, but there is the risk that a court will not know what you are talking about when you use them.

      Moreover, for the sake of conceptual clarity, it is often better to explain exactly what these entail — for example:

      • “higher threshold which the applicant has to cross to make out unreasonableness”
      • “court suggested that [constitutional right] was a consideration which the decision-maker was bound to take into account: Vijaya Kumar“
      • “the decision-maker failed to take into account relevant evidence: Fong Thin Choo. The court can quash a decision if there is “manifestly insufficient evidence”: Axis Law v IPOS. As the applicant’s [fundamental right] is at stake, the court ought to be less easily satisfied that the evidence justifying [decision] is satisfied, as in Bugdaycay…”
      • “because national security is at stake, the court will not readily hold that the decision is irrational; however, the question remains whether the decision is ultra vires the statute, just as in Tan Seet Eng…”
      • “even accepting that the proportionality of the decision is non-justiciable, the court may still inquire into its rationality, as in Colin Chan v MITA“
      • “while the court cannot inquire into what national security requires, it can and must inquire into whether the decision is based on national security considerations (Chng Suan Tze at [93])”

      To reduce all of these possibilities down to just one or two words would not do justice to the wide variety of methods by which the law controls state action, as well as the doctrinal differences between them. It would also not be the most persuasive if these one or two words are not part of Singapore law.

    19. The word “shall”

      In class, I mentioned that the word “shall” is somewhat frowned upon, or, at least, controversial. See:

      For more interesting reading, simply search the Statute Law Review for “shall”.

    20. Oral vs. written advocacy in Singapore

      See Re Lord Goldsmith Peter Henry PC QC [2013] 4 SLR 921. At [29]-[31] (emphasis removed) the High Court remarked: 

      “the appellate process in the Singapore Court of Appeal has evolved to be more writing-centred than that in many other common law court systems. The written medium occupies a central role in the appellate process here […] 

      Court rules require the parties to make detailed written submissions in the form of Cases prior to the hearing of oral arguments. The Case is a formal document that must provide structured and comprehensive arguments. It affords counsel a platform to make rational, dispassionate and intellectual submissions on the merits of their cause and the finer points of law that are in play. In short, the parties are expected to place on record their entire panoply of submissions through the written medium well in advance of the appeal hearing […] 

      Oral arguments therefore ordinarily serve as a useful concluding supplement to the Case. During the oral phase, counsel are expected to present the essence of their Case. Counsel also assumes the role of assisting the court to clarify information that could give a clearer picture of the issues rather than functioning as traditional orators who have an expectation that they will be heard out at length according to their desires.”

      It would follow that lawyers arguing appeals must come to the courtroom prepared to answer questions from the court. By that stage, the court would have read the written arguments. Bear that in mind when considering the law on excessive judicial interference and reconciling it with what you might witness when you watch the hearing of an appeal.

    21. Mayors in Singapore

      Someone asked what Mayors in Singapore do, and whether they are redundant.

      To answer this question, it would be useful to start with the law stating what Mayors’ functions are. That law isn’t found in the Constitution. Instead, a search on Singapore Statutes Online leads us to the People’s Association (Community Development Councils) Rules. Rule 6(1) tells us: “The Board may designate the Chairman of the Council for a District to be the Mayor of that District where… the number of residents in the District is not less than 150,000.”

      So we next need to find out what the “Board”, a “Council” for a “District”, and the “Chairman” of the “Council” are. When we read rules 2 to 5, we find out that the word “Council” refers a Community Development Council, which may be appointed by the Board of Management of the People’s Association for a certain area in Singapore (called a “District”). Rule 3 tells us that all this may take place “[f]or the purposes of carrying out the objects of the Association… and the activities advantageous towards, or necessary or convenient for, the furtherance of such objects”.

      So the last piece of the puzzle is: what are the “objects of the Association”? We are told in section 8 of the People’s Association Act 1960. Those, in turn, are the functions of Community Development Councils, whose chairpersons may be given the title of “Mayor”. 

      So the question of whether Mayors are redundant is, essentially, the question of whether the functions of the People’s Association are the same as the functions of MPs. They are not, although they overlap. To be sure, there are conversations to be had about whether Mayors do effectively perform their functions, and indeed whether they should have those functions in the first place. But that question is different from the question of whether all of their functions are performed by someone else — which they are not.

      It is also worth considering what resources Community Development Councils have to perform those functions, compared to individual MPs. Take, for example, “the fostering of community bonding and strengthening of social cohesion among the people of Singapore”. It is likely that CDCs have more funds to (for example) organise events for this purpose than individual MPs do.

      Finally, it is worth considering that elected MPs are elected whereas CDCs, including their Mayors, are not. (Some of them happen to be elected MPs, but they were not elected as Mayors.) That means that elected MPs face the threat of being voted out. That, in turn, keeps them on their toes, and ensures that they effectively perform important functions such as acting as the bridge between individuals and the Government — functions which the People’s Association doesn’t serve.

    22. Having a legal power doesn’t mean that you have to exercise it to the fullest possible extent: a case study from the Road Traffic Act

      Section 121(2) gives the Minister for Transport the power to make subsidiary legislation, known as “rules”, “with respect to the precedence of vehicles and pedestrians respectively and generally with respect to the movement of traffic (including pedestrians) at and in the vicinity of crossings”. Section 121(3) gives us examples of such rules: for example, the Minister may make rules “prohibiting pedestrian traffic on the carriageway within 100 metres of a crossing”.

      Please note what section 121(3) does. It does not say that pedestrians may not cross within 100 metres of a crossing (other than at the crossing itself). Indeed, it is not directed at pedestrians at all. Rather, it is directed at the Minister. The purpose of sections 121(2)-(3) is simply to give the Minister the power to make certain subsidiary legislation for certain purposes.

      The Minister, exercising this power, made rule 3(1) of the Road Traffic (Pedestrian Crossings) Rules, which prohibits “any pedestrian, cyclist, mobility vehicle user or PMD rider” from crossing the road within 50 metres of a crossing (other than at the crossing itself).

      There is no contradiction between the two. The Act gives the Minister the power to make rules prohibiting crossing within 100 metres of a crossing. The Minister exercised that power, but not to the maximum possible extent: he only made a rule prohibiting crossing within 50 metres of a crossing. That is lawful. Just because the Act gave him a power doesn’t mean that he has to exercise that power to the fullest possible extent.

      You may also ask: didn’t the Act only give the Minister the power to make rules relating to “pedestrian traffic”? Does rule 3(1) of the Rules exceed these powers, by applying not only to pedestrians but also “any… cyclist, mobility vehicle user or PMD rider”?

      The answer is no. Section 121(2) of the Act gives the Minister a general power to “make rules… generally with respect to the movement of traffic (including pedestrians) at and in the vicinity of crossings”. This is a wide power, and it clearly allows the Minister to make rule 3(1) of the Rules. Section 121(3) gives us examples of types of rules that the Minister may make using this power, e.g. rules relating to pedestrians near crossings. They are just examples. We know they are just examples because of the words “Without limiting subsection (2)” — in other words, the examples in section 121(3) do not limit the general power to make rules in section 121(2).

    23. What does it mean to say that the AG is the “guardian of charities”?

      In class [that is, a class for students other than LLB students], when someone asked this question, I struggled a bit to summarise the relevant parts of the law on charities, which can be somewhat arcane. I just checked a couple of textbooks, and concluded that the law is even more arcane than I remember. So the best answer I can give, without using too much legal jargon, is this:

      The AG is a guardian of the public interest. Charities are supposed to exist for various aims relating to the public interest. Therefore, the AG may act in the interests of a charitable cause or the beneficiaries of a charity. For example, at common law, if a charity misapplies its funds, the AG can sue to stop this. (One would think that it is the supposed beneficiaries of the charity who should have the right to sue the charity. But the beneficiaries are often not in a position to do so – they may lack the means, etc. Besides, there may not be clearly identifiable human beneficiaries in the first place, e.g. in the case of a charitable trust that exists to advance a certain religion.)

      The common law of charities in Singapore has been modified by the Charities Act 1994. Nonetheless, the Attorney-General has certain functions under that Act. For example, under s 23 of that Act, if the Commissioner of Charities, having begun an inquiry into a charity, is satisfied that “there is or has been any misconduct or mismanagement in the administration of the charity” and that it is “necessary or desirable to act for the purpose of protecting the property of the charity or securing a proper application for the purposes of the charity of that property or of property coming to the charity”, then the Commissioner may do various things, such as removing a member of the charity’s governing board – but the Commissioner can only do this with the Attorney-General’s consent. This took place in 2013: see https://www.agc.gov.sg/docs/default-source/newsroom-doucments/media-releases/2013/agc-statement_ag’s-role-under-the-charities-act_approved_11-april-2013.pdf

      Another example of the AG’s role with regard to charities is Re Lot 114-69 Mukim 22, Singapore. In 1920, Lim Kit Fah sold a piece of land to five people jointly, including Ong Choo Kee. Some time after Ong Choo Kee died, his grandson, Ong Yew Kew, applied to court for a declaration that he (and not the other people to whom Lim Kit Fah had sold the land) had become the owner of the land outright. The Attorney-General objected, arguing that Lim Kit Fah had intended the land to be used for charitable purposes (namely, the advancement of religion); therefore, Ong Yew Kew could not own the land outright, but was bound to continue those charitable purposes. 

      The High Court initially concluded that the Ong family had occupied the land for so long, and the public charitable purpose had fallen into disuse for so long, that Ong Yew Kew was entitled to take over the land ([2001] 1 SLR(R) 811). However, on hearing further evidence, the High Court decided ([2003] 1 SLR(R) 773) that the land charitable purpose had not fallen into disuse: while the Ong family had lived on the land, the temple on the land had remained open to the public. (By the way, if I’m not mistaken, the temple in question is this one: https://kewongyah.org.sg/en/#section1) 

    24. “try again”, the butterly effect, and the Tan Seng Kee case

      In class, I mentioned that the Tan Seng Kee case was brought at least in part because of a comment by Tommy Koh on Facebook. I, for one, find it interesting that, in the world in which we live nowadays, comments on social media can not only be newsworthy, but also play a role in legal and political developments. I promised to substantiate these observations with facts, so here you go:

      1. On 6 September 2018, the Supreme Court of India handed down its judgment in Navtej Singh Johar v Union of India AIR 2018 SC 4321, in which it held that s 377 of the Indian Penal Code (“Whoever voluntarily has carnal intercourse against the order of nature with any man, woman or animal, shall be punished with imprisonment for life, or with imprisonment of either description for a term which may extend to ten years, and shall also be liable to fine.”) was unconstitutional “in so far as it criminalises consensual sexual conduct between adults of the same sex” (at [156(i)]). (The court’s reasoning is, in my view, largely inapplicable in Singapore.)
      2. The same day, Professor Simon Chesterman, who was (and still is) the Dean of the Faculty of Law at NUS, put up a Facebook post about the case.
      3. The same day, Professor Tommy Koh commented on the post: “I would encourage our gay community to bring a class action to challenge the constitutionality of Section 377A”.
      4. The same day, one Edward Tay replied: “Prof, hasn’t that angle been disposed of by Lim Meng Suang?”, referring to Lim Meng Suang v Attorney-General [2015] 1 SLR 26, in which Singapore’s Court of Appeal held that s 377A of Singapore’s Penal Code (“Any male person who, in public or private, commits, or abets the commission of, or procures or attempts to procure the commission by any male person of, any act of gross indecency with another male person, shall be punished with imprisonment for a term which may extend to 2 years.”) did not violate Art 9(1) or Art 12 of Singapore’s Constitution.
      5. The same day, Prof Koh replied: “try again”.
      6. On 10 September 2018, Mr Ong Ming Johnson (also known as DJ Big Kid) applied to the High Court for a declaration that s 377A violates Art 9(1), Art 12(1), and Art 14(1)(a) of the Constitution. On 12 September 2018, he told media that he had been “energised” by the decision of the Indian Supreme Court and decided to “take up” Prof Koh’s “challenge”.
      7. On or before 22 November 2018, Mr Choong Chee Hong (also known as Mr Bryan Choong) made a similar application to the High Court.
      8. On 21 September 2019, Dr Tan Seng Kee (also known as Dr Roy Tan) made a similar application to the High Court.
      9. On 30 March 2020, the High Court dismissed all three applications.
      10. The applicants appealed to the Court of Appeal. On 28 February 2022, the Court of Appeal declined to rule on the applications because it ordered that s 377A be “unenforceable in its entirety” (at [149]; emphasis removed). However, it made certain obiter remarks strongly suggesting that s 377A is unconstitutional. So, as Assoc Prof Eugene Tan later put it, s 377A is “on its last legs” and is in effect a “dead law“.
      11. On 21 August 2022, the Prime Minister announced during his National Day Rally speech that s 377A would be repealed for various reasons, including that “there is a significant risk of s377A being struck down, on the grounds that it breaches the Equal Protection provision in the Constitution”. He also announced that the Constitution would be amended so as to prevent the existing “definition of marriage [from being] challenged on constitutional grounds in the courts”. We currently await the text of the constitutional amendment. If it is passed, then it is not implausible that, sooner or later, the courts will have the occasion to rule on the precise effect of that amendment.

      The decision in Navtej Singh Johar was irrelevant to the decisions of the Singapore courts, as were Prof Chesterman’s Facebook post and Prof Koh’s comments. Nonetheless, it is interesting to observe the effect that these developments have had on where we are today.

    25. Writing clearly: some tips

      The ideal paper (if it exists) is easy to follow in the sense that the reader never has to wonder where the discussion is going or why something is being said. Therefore, it is good to provide ‘signposts’ to guide the reader along. ‘Signposts’ are indications to the reader of what arguments are about to be made, and where one point ends and the next begins. For example:

      “The case law reveals two approaches. The first is… For example… The second, unlike the first, focuses on… For example…”

      “What can explain the gradual shift in the case law from X to Y? There are three main reasons. First… However, this reason is of tangential relevance compared to the second… Yet this second reason cannot explain why… The main reason is that… This explains not only… but also…”

      Where possible, aim to state the argument upfront and then use sources to support it. Otherwise, the reader may be left wondering why cases and articles are being described or quoted from, and will have to keep wondering until the end of the paragraph/section. (In addition, there is a risk that you will end up providing excess/unnecessary detail about the cases.) Instead of:

      “Case A said… Case B said… This shows that X is true.”

      Try:

      “X is true. This is shown by case A… Similarly, Case B…”

      More detailed examples (which I came up with — these aren’t from actual students’ essays)

      Consider the following examples of responses to a question about “non-justiciability”.

      Avoid the following, which is not good because it does not make the main points clear, and wastes space providing unnecessary (and irrelevant) detail instead of presenting arguments:

      In Colin Chan v MITA, the appellants were ministers of the Jehovah’s Witnesses.  The Minister for Home Affairs ordered that their local chapter be deregistered and their publications be prohibited. The appellants sought judicial review of these orders. The Court of Appeal held that “the issue is not justiciable” and “there is very little room (if any) left for any doctrine of proportionality (assuming it exists) to apply, other than the well-established one of irrationality.” However, the Court of Appeal examined the orders and held that there was nothing “to show what might turn out to be an arguable case that the Minister had acted irrationally when he issued Order 405/1994”. [These are just quotations from the case; it is not clear what point is being made.]

      However, in Lee Hsien Loong, the High Court said that “Justiciability depends, not on the source of the decision-making power, but on the subject matter that is in question.” Hence, the High Court granted judicial review of a treaty. [Note: this is not only imprecise; it is incorrect.] This was a more flexible approach [more flexible than what?] as the High Court recognised that it was “here concerned with the interpretation of a treaty and a foreign statute in the light of some international instruments”. 

      The next case is Yong Vui Kong, where the Court of Appeal held that a clemency decision can be subject to judicial review. The Court of Appeal held that “the Supreme Court had jurisdiction to adjudicate on every legal dispute on a subject matter in respect of which Parliament has conferred jurisdiction on it” and this included “any constitutional dispute between the State and an individual”. However, the clemency power was “not a matter of legal right”. Thus, “the clemency power is a constitutional power vested exclusively in the Executive, it is not justiciable on the merits“. [This sentence is not grammatically correct.] However, there were ultimately legal limits to the clemency power, namely, bad faith, malice, and unconstitutionality. [At this point, it is not clear why all this detail has been provided.]

      Therefore, ‘non-justiciability’ should not be treated as an excuse to avoid engaging in judicial review. [Much more is needed to explain how the above statements of what the law is can lead to a conclusion as to what the law should be. Moreover, this sentence suggests that the courts do avoid engaging in judicial review, which runs contrary to the point that appears to have been made earlier.] The courts do apply a flexible and non-rigid approach toward justiciability as ultimately all power has legal limits.

      Instead, consider:

      Despite the use of the term “non-justiciability”, the Singapore courts have never deployed a doctrine of non-justiciability to hold that the exercise of some power is not subject to legal limits. In Colin Chan, the Court of Appeal stated that “matters of national security are not justiciable”, and yet went on to perform judicial review on the ground of irrationality. In Tan Seet Eng, the Court of Appeal explained Colin Chan as a case where the national security context called for a “less intense standard of review” — not no review at all.

      It is true that, in Lee Hsien Loong, the High Court appeared to suggest that the exercise of power may be rendered non-justiciable by virtue only of the “subject matter” in question. But this must be seen in the light of the High Court’s defining “subject matter” in terms of the particular question at stake in the particular case, not the general topic in question. While one might say that the issue in that case concerned international relations, the High Court defined the “subject matter” in more specific terms, the effect of an international treaty for the purposes of “determin[ing] the domestic legal obligations” as to the manner in which process was to be served. In other words, even as the High Court appeared to use the binary labels of ‘justiciable’/’non-justiciable’, it applied this terminology in a flexible manner by requiring the issue being labelled to be described in specific, granular terms.

      This same approach, viz. applying the doctrine of non-justiciablility sparingly by requiring the question before the court to be stated particularly, can explain the courts’ reasoning in other cases even before Tan Seet Engtacitly deprecated the terminology of justiciability in favour of an explicitly “flexible” doctrine of deference. For example, we may reconcile the statement in Colin Chan that “matters of national security are not justiciable” with the court’s performing irrationality review as follows: while the proportionality of the decisions in question was non-justiciable, their rationality was justiciable. Similarly, in Yong Vui Kong, the Court of Appeal did not make a blanket statement that an exercise of the clemency power was non-justiciable; rather, the Court of Appeal said that an exercise of the clemency power is “not justiciable on the merits” (emphasis in original) — in other words, aspects of the exercise of that power other than the merits of a clemency decision are justiciable. (While it is not clear what precisely “merits” refers to, the point for present purposes is that the Court of Appeal did not label every clemency-related issue non-justiciable.)

      In short, it should not be thought that, until Tan Seet Eng, the Singapore courts’ use of the terminology of ‘non-justiciability’ had resulted in an unwarranted eagerness to refuse to impose and enforce legal limits to executive powers. Our critical attention therefore ought to focus on what those limits are, and not the question of whether at all there are any limits.

    26. Organising an essay: an example

      It is not necessarily the case that one needs to (a) spend about a paragraph at the beginning of the essay defining the terms used in the question; (b) use the rest of the space to answer the question. It may well be that far more than an introductory paragraph or two may be necessary to provide definitions. In fact, it may well be that the entire essay is an exercise in definition. If, for example, you are asked whether judicial review is anti-democratic, your essay might take the following form:

      1. It depends what is meant by ‘anti-democratic’…
      2. First, ‘anti-democratic’ could mean having the aim of undermining democratic processes. By this measure, judicial review is not anti-democratic.
        1. One might think it is anti-democratic. After all…
          1. Case study
          2. Case study
        2. However, in truth…
          1. Case study
          2. Case study
      3. One might instead say that ‘anti-democratic’ refers to having the effect of undermining democratic processes.
        1. This is not necessarily the case.
          1. Case study
          2. Case study
        2. That said, there are situations in which judicial review has the effect of…
          1. Case study
          2. Case study
      4. Is judicial review being ‘anti-democratic’ in this limited sense justified? ‘Anti-democratic’ connotes something undesirable… So it might be misleading to apply the label ‘anti-democratic’ if there are good reasons for the law to de-prioritise democratic ideals or the wishes of a democratically-elected body.
        1. Good reasons might include…
          1. Reason A
            1. Case study
          2. Reason B
            1. Case study
          3. But not reason C
            1. Case study
        2. Even then, judicial review would be ‘anti-democratic’ in the sense of undermining democratic ideals/institutions in certain discrete instances, but that is justified as a means to promoting the broader ideal of democracy…
      5. Conclusion:
        1. The label ‘anti-democratic’ is misleading because… Instead, a better label is…
        2. All this having been said, I have explained why one might be tempted to label judicial review ‘anti-democratic’. The law needs to pay attention to X, Y, and Z to avoid being what I have described as being undesirably anti-democratic.
    27. Various tips on writing research papers, answering essay questions, thinking about the law, etc.

      Other blogs

      I would like to recommend to you the blog of Dr Dimitrios Kyriazis: https://www.thelawprof.com/writing-legal-essays-1. There are several posts there about writing essays. They focus on essays about the law, but the advice there is really applicable to any sort of essay.

      I would particularly recommend a blog post by Professor Mark Elliott, which a colleague kindly shared with me: https://www.thelawprof.com/writing-legal-essays/remember-to-argue-guest-post-by-professor-mark-elliott. I have always found Professor Elliott’s writing clear and smooth.

      Consider this Instagram post by Juliette Han: https://www.instagram.com/p/Cxng-BkOjdA/

      Miscellaneous points

      Feel free to critique materials you come across, including the views of courts and academics (including me). You are free to explain not only that you take a certain view because X says so, but also that you agree with X’s view but not for the reasons that X refers to.

      In general, it is good, not bad, to acknowledge potential flaws in or possible criticisms of one’s arguments. It would be better to confront such possible criticisms than to give the impression that one is trying to get away with letting them slip by unnoticed.

      Consider categorising things. For example, when unpacking the assertion that there is “diversity” in the law, consider different types of diversity: a diverse range of legal doctrines? a diverse range of approaches to different sets of facts? 

      Something may be principled even if the underlying principle is not expressly articulated.

      Embracing a concept doesn’t translate directly to performing certain activities. Consider the idea of embracing the concept of human dignity. For some, that concept means abolishing the death penalty; for others, it means that the condemned person walks to the gallows instead of being dragged there; for others, it could mean encouraging the death penalty for those who have committed crimes against dignity; for yet others, it could mean retaining the death penalty, but varying the method of execution. (This point is my paraphrase of certain ideas which I was exposed to by a lecture by Professor Jeremy Waldron, in case you’re interested in reading more about human dignity.)

      If you say that something should be abolished, does it need to be replaced with something else?

      The law sets out minimum standards, not ideal standards to aspire to. For example, if you have never committed a serious crime, that is not something to boast about, any more than washing your hands is something to boast about. Acting unlawfully is not bad, but acting lawfully doesn’t make you good/excellent. (The phrase “hygiene factor” comes to mind.) So, if all that can be said in support of an idea is that it would not be unlawful, then perhaps we should think twice about whether it is such a great idea.

      If you criticise something for being insufficiently detailed, consider whether it is (a) possible and (b) desirable to provide further detail.

      Even if something is not binding on X, X can still choose to follow it.

      It is possible to argue that, in order to do task A better, someone should also be made to do task B. 

      Guiding the reader

      It is difficult to prescribe a structure for any essay. Instead, I would like to offer this tip: Your reader should rarely, if ever, be left wondering ‘why is this person saying this?’ or ‘why is this point suddenly coming out of left field?’ or ‘what is this idea that this person keeps mentioning but hasn’t unpacked?’ To avoid this, you should not only sequence what you have to say in a logical manner — for example, if you need to provide background knowledge in order to set the stage before making a point, then do so — but also provide useful signposting. Phrases like “There are two arguments… First… Second… One might argue that… But this would miss the point…” or “It depends what one means by the right to life. If one means… But that would hardly be satisfactory because… Instead, the right to life, in its fuller sense, means…” would show how each statement leads to the next/follows from the previous one, and would thus help the reader to follow the argument.

      Answering ‘yes/no’, ‘good/bad’, etc. questions

      You need not answer what appears a yes/no question with a simple yes or no. ‘It depends’ can be a valid and useful answer. However, you don’t want to just say ‘it depends’ and leave it at that. Rather, explain: what is ‘it’ and what does it depend on? 

      Your answer could be something like ‘if X means A, then yes; if X means Y, then probably no’; ‘yes in the short term; no in the long term; the right time to transition between the two is…’; ‘yes, provided that the following conditions are met’; ‘it doesn’t matter whether or not one does X; what is more important is Y’; etc. If the question asks whether more should be done to ensure X, the answer could be: yes, more should be done to ensure X; no, enough is being done to ensure X; no, X is bad and it should be removed, not promoted; it’s impossible to ensure X, and we can at most encourage X; etc.

      ‘This is the wrong question to ask’ or ‘a simple yes/no answer would be too simplistic’ can also be valid answers, but of course you will have to explain whatever your answer is.

      Not everything is black-and-white

      Be careful of extremes. For example: You might argue that discretion is bad because predictability is good. But one might as well argue that discretion is good because flexibility is good. Of course, the question of what we make of discretion is not just a binary question of whether it is ‘bad’ or ‘good’ — so don’t act as though it is.

      There was an essay question about restorative justice. Strong essays ecognised that the question of whether restorative justice is to be embraced is not a binary one: the principles of restorative justice can bear different weights depending on contextual factors such as the severity of the crime, and a process is not hopelessly deficient merely because the role played by the victim is not the biggest possible. They also considered variations on the theme: they distinguished between (for example) making victim-offender mediation compulsory vs. encouraged vs. mandatory.

      Consider what your aims are, and why they are your aims

      Some argued that restorative justice would be a good thing because it would reduce recidivism. That may be true, but is it the only aim of restorative justice? Be careful not to send the message that the criminal justice system should pay more attention to victims, not for the victims’ sake as an end in itself, but as a means to the offender’s rehabilitation.

      Meaningful comparisons

      Suppose you are asked to consider whether Singapore’s Parliament should become more like the UK’s. Clearly, no person who suggests that could reasonably mean that (for example) the average height of the members of Singapore’s Parliament should be made closer to that of the members of the UK’s, or that Singapore’s MPs should dress more like MPs in the UK. Rather, in order to answer the question, one must consider: which aspects of the UK’s Parliament are the salient ones which one would even think are worth imitating?

      Don’t rule out comparison because of differences that don’t, well, make a difference. Suppose you were making an argument about the AG’s prosecutorial functions. Could you say that a comparison with the UK would be inapposite, because there the AG doesn’t have prosecutorial functions? No, because the UK does have a public prosecutor, and you could very well compare Singapore’s AG qua public prosecutor with the UK’s public prosecutor (who is called the Director of Public Prosecutions). Of course it would be wrong to compare based on job titles: it would be difficult to make a meaningful comparison between, say, the President of Singapore and the President of the USA, because they are not even close to being functionally equivalent. But we can very well compare institutions which serve a similar function: it would make perfect sense to compare Singapore’s Prime Minister with the President of the USA, or Singapore’s President with the King of the UK, or Singapore’s AG qua PP with the UK’s DPP, or Singapore’s AG with the UK’s (AG plus DPP).

      Different names for the same thing; the same name for different things

      Sometimes, other legal systems use a term but use it to mean something else from how we use it. Take, for example, the phrase “strict liability” in criminal law. As far as I can remember, in English law, there is no general defence of due diligence to offences with no mens rea element; whereas, in Singaporean and Canadian law, there is, and if there were no due diligence defence then we would use the term “absolute liability”. In other words, what we would call an offence of “absolute liability”, the English would call an element of “strict liability”. So anyone who was writing an essay comparing strict liability offences in Singapore law and other legal systems would need to bear this in mind, searching for the functional equivalent of the Singaporean doctrine of strict liability elsewhere, and not merely searching for doctrines that are named “strict liability” elsewhere.

      Similarly, it is possible that a label such as “the rule against fettering”, “natural justice”, etc. refers to something different in Singapore than in some other jurisdiction.

      Seeing arguments, phenomena, etc. in context

      In discussing cases in which novel arguments were made to courts, consider why these arguments may have been raised. For example, if I recall correctly, Prof Swati Jhaveri has made the point that perhaps what prompted the applicant in Chiu Teng to attempt to argue for a novel ground of judicial review was that $33 million (not a small sum) was at stake. What could prompt litigants to attempt to argue that an error of fact has taken place? or that taking the national pledge is a “ceremony in a religion”?

      When making an observation about a phenomenon, consider the circumstances that may have given rise to that phenomenon. For example, it may be that the number of applications to the court of a certain sort has increased because something in the law has changed to make it easier to make such an application. Or it may be that the number of applications for judicial review has increased because the rate of regulatory activity by the executive / the range of executive powers created by the legislature has increased. Be careful not to consider only one possible explanation for a phenomenon to the exclusion of others. For example, as Chan CJ points out in ‘From Angst to Empathy’, the reason why judicial review applications sometimes fail is not necessarily that the law stacks the odds against the applicant; it could simply be that those applications are unmeritorious, or that there are other means by which disputes are resolved (see [14]ff). (A more prosaic example: one might be tempted to look at a (hypothetical) recent increase in rates of casino crime and conclude that the cause is moral decay in society, but the more accurate explanation might simply be that recently more casinos have come into existence.)

      Assessing things

      In assessing anything (an attempt, a balance, a rule, etc.), be sure to explain (and justify) what standard (sometimes colloquially called a ‘yardstick’) you are measuring it against. For example: if you say that a rule is “insufficient”, how are you measuring sufficiency, and why have you chosen that method of measurement?

      In assessing a balance between A and B, you may consider not only the outcome of the balancing exercise, but also the method by which it is sought to be struck, as well as whether A and B are the correct things to balance in the first place.

      How much detail should one go into?

      In writing about any matter, the level of depth required is simply what would be relevant in answering the particular question. For example, if you are answering a question on whether the Chevron doctrine should apply in Singapore, it could be useful to discuss whether a particular difference between US and Singaporean constitutional law is an argument in favour or against Singapore law following Chevron. If it is necessary to provide some detail to explain what this difference is and why it is relevant, so be it. On the other hand, it would not be a good idea to engage in a bare discussion of the US Constitution without explaining how it is relevant to the question.

      “It depends”

      If your thesis is ‘it depends on X’, then you could consider organising your essay along the lines of the various possibilities. For example:

      • If X is A, then the conclusion is…
      • If X is B, then the conclusion is…
      • If X is C, then the conclusion is…
      • X cannot be A because…
      • That leaves B or C. X is now B… But in the longer term, X could change to C if…

      Making assumptions

      Suppose you think that an assumption would be questionable to make, and you say so. Can you then go on and rely on that assumption anyway? The answer is not necessarily no. It is perfectly fine to take one of the following views:

      1. The assumption is open to challenge, but not on grounds that matter for present purposes
      2. The assumption is open to challenge, but on balance I think it is justified / we’re better off sticking with it
      3. I think X because the assumption is wrong. But even if we were to accept it, X is still true, because…

      Of course, it depends on what you mean by “challenge”. It’s one thing if there are some possible arguments against making the assumption, but you can still see the value in accepting it (even if only reluctantly). But if (for example) the assumption is clearly very wrong, then one would think that any argument founded on that assumption must also be wrong. 

      Problems and solutions

      Some would have us believe that there is no use in one pointing out a problem if one doesn’t have a perfect solution. That is not true. Even explaining that a problem exists is a useful thing to do. Moreover, even if one cannot offer a complete, perfect, well-thought-out solution, it would still be helpful to offer some remarks about what a solution may look like, what a solution will require, etc.

      Transplants

      Sometimes, someone (e.g. a court) borrows one idea from one context and deploys it in another context. One way to respond may be to say that the idea is ill-suited to the latter context. But we should dig deeper and ask: Why would one even think to attempt such a transplantation? What is the transplanter trying to achieve?

      When you are discussing a line of cases, don’t just comment on each case in isolation from the others, as though you are writing a series of case notes. Rather, consider saying something about how each case builds on/represents a break from the previous one; paves the way for the next one; contains the seeds of a problem which later presents itself; etc.

      When you don’t know enough

      If you think some kind of data would be useful in testing a hypothesis/answering a question, but that data isn’t available, and you have indeed put in the work to see if that data is available, then feel free to point that out — feel free to acknowledge gaps in knowledge, e.g. “… It does not help that there is no data about…” Consider inserting a footnote to state briefly what you did to find such data — that’s a way to rebut any objections from readers that you haven’t looked hard enough.

      Confront hard cases

      There was a question calling on students to discuss the idea that “religion and politics should not mix”. Some essays pointed out that there are religions which themselves teach the value of accommodation and tolerance of other religions. This is well and good, and it may appear that such a religion can sit comfortably within a framework of secular governance because there are few (if any) sites of potential conflict between the religious belief and the state goal. But consider:

      • If a religion tells believers that (say) God commands them to submit to secular state authorities, are the believers submitting because they think they should, or because God has told them to? If it is the latter – if the religion dictates that religion and state are separate – then one might argue that the religion purports to occupy a higher position in the hierarchy of norms, in that it is the religion that gets to say which norms are higher.
      • What if there is a religion that doesn’t teach that – say, a religion whose core belief is that other religions are wrong, that the state is to be resisted, etc.? It will not be so easy to say that everyone can just smile and get along.

      “More should be done”

      When answering a question about whether “more should be done” about something, consider: what is presently being done? If the answer is that nothing is being done, or that we don’t know, that itself can be a useful observation (provided that it’s accurate!)

      “More should be done” — by whom?

      Note that it’s possible to do something without identifying it by name. It’s possible for a legal system to have features of restorative justice without using that phrase. So don’t just, say, search the literature for “restorative justice” in Singapore and conclude that it doesn’t exist because nobody has used that phrase. Instead, consider what “restorative justice” means, and consider whether it exists/should be embraced in substance.

      Ask yourself: Who cares, and why?

      Essay questions generally do not call for the repetition of information, without more. That is not to say that information is not important; it is: an essay should not make assertions without substantiating them. However, the role of pieces of information is to support arguments. A good rule of thumb is to, whenever you write about a piece of information (such as some legislation or a case), to ask yourself why you are doing so and what the relevance to the topic is. Then make sure that, in your writing, you make clear what the relevance to the topic is. Better yet, state the relevance before narrating the information, so that the reader is told in advance what the main point is, rather than the main point emerging only toward the end.

      The skills involved in dealing with an essay question are the same skills as those that come into play when you have to respond to someone who makes a loaded statement. Suppose you are making submissions to a court, and the judge remarks that what you are proposing sounds “anti-democratic”. The purpose of such a remark is, to quote Mohammed Ali bin Johari v PP [2008] 4 SLR(R) 1058 (CA) at [175(c)(iv)], to “assist counsel and their clients to be cognisant of what is troubling the judge”. At that point, what the judge wants to hear is not one case after another, but rather an explanation of why what you are saying is not, in fact, “anti-democratic”. A response to the judge’s remark would therefore need to engage with the concept of being “anti-democratic”.

      Similarly, suppose you are arguing that your client was not given a fair hearing, and that you are required to engage with the assertion from the Dow Jones case at [59] that a hearing need not be given to a person who has “no right but simply a privilege” before that privilege is taken away. One possible tactic could be to cite cases that say that even one whose mere “privilege” is at stake is to be given a fair hearing. But another tactic could be to question what a “right” is, and to argue that your client did have a “right” at stake. Yet another could be to argue that, even though your client did not have a “right” at stake, he did have something more than a mere “privilege”. The skill of engaging with the concepts of ‘rights’ and ‘privileges’ and asking what these words mean, rather than merely citing cases which use these words, is the same skill that one has to exercise when responding to the assertion that the constitution “focus[es]” on A but neglects B.

      “They Say/[but] I Say”

      There is a book called They Say/I Say: The Moves That Matter In Academic Writing available in the library. Its main point is this:

      “the underlying structure of effective academic writing — and of responsible public discourse — resides not just in stating our own ideas but in listening closely to others around us, summarizing their views in a way that they will recognize, and responding with our own ideas in kind. Broadly speaking, academic writing is argumentative writing, and we believe that to argue well you need to do more than assert your own position. You need to enter a conversation, using what others say (or might say) as a launching pad or sounding board for your own views.” (at page 3)

      The writers note that “in the real world we don’t make arguments without being provoked”, and hence academic writing is not just a “process of saying ‘true’ or ‘smart’ things in a vacuum” (at page 3). So “[i]f your own argument doesn’t identify the ‘they say’ that you’re responding to, it probably won’t make sense” (at page 4).

      The book provides many examples, but, for present purposes, one will suffice: at pages 4-5, the writers discuss Martin Luther King’s ‘Letter from a Birmingham Jail’:

      “You deplore the demonstrations taking place in Birmingham. But your statement, I am sorry to say, fails to express a similar concern for the conditions that brought about the demonstrators.”

      Advising a client (real or hypothetical) hen the law is not clear

      1. When the law is unclear, it is a good idea to point out the unclarity. Useful phrases might include:
        1. As can be seen from… it is unclear whether the test is A or B. However, it does not matter. If the test is A… If the test is B…
        2. … depends on whether the court holds that the test is A or B. If the test is A… whereas if the test is B…. Therefore, we should attempt to persuade the court that the test ought to be A, possibly on the grounds that, as pointed out by Professor Craig…
        3. While an argument may be made that the test ought to be A, there is little point pursuing the matter because Mr X would not succeed even on the more liberal A test…

      Don’t use ambiguous words

      Please be careful when using labels such as “light-touch”, “super-Wednesbury“, etc. Their meanings may be unclear and you may risk being misunderstood.  For example, as one of you pointed out, the term “super-Wednesbury” has been used by different authors to mean completely different things. The label “light-touch”, moreover, does not, as far as I can recall, appear in the case law (unlike the term “anxious scrutiny”); one therefore cannot assume that it will be understood as a term of art.

      On a related note, please beware of the danger of using phrases which are clearly shorthand for something but do not make clear what that something is. A potential example we have come across is “rule of law” as used in popular discourse. Another is “due process”: exactly what kind of process are we saying is due to people?

      ‘Due’

      The idea of a ‘right to due process’ ,without more, is circular, because to say that something is ‘due’ to you simply means that you have a right to it. The real question is: what kind of process is ‘due’, and why?

      ‘How’

      If I ask you “How can I make good moussaka?”, all of the following could be possible answers:

      1. By following a recipe.
      2. By ensuring that I slice my eggplants longitudinally.
      3. With a little help from my friends.
      4. Carefully.
      5. With an oven.
      6. By being sure to use the best ingredients.

      This is because ‘how?’ can mean many things, from ‘through what process?’ to ‘using what?’ to ‘by whom?’ to ‘in what manner?’ to ‘with what principles in mind?’

      I would suggest bearing the many possible meanings of ‘how’ in mind when engaging with questions that ask ‘how?’.

      Approaching questions critically

      Essay questions in law schools are not meant to test one’s recall of knowledge, without more; this applies a fortiori to take-home research assignments and to questions asked as part of an open-book examination. Rather, the skill being tested is that of deploying pieces of knowledge to make arguments. It is therefore necessary to begin by commenting on precisely what one is being asked to argue about. This may well involve stating that the question is ambiguous or unclear; stating that the question is fallacious (for example, because it has been asked on the assumption that there is a strict dichotomy between X and Y, when in truth that dichotomy is a false one); or stating that the question makes multiple assertions, some of which are correct and others of which are not.

      It is hoped that the following examples will illustrate possible lines of enquiry which one might pursue in response to a question:

      • “Is X too much…?” — How are we to measure the ‘amount’ of X? What is the ‘ideal amount’ of X?
      • “Assess X’s role in bringing about Y.” — How has X helped to bring about Y? Are there ways in which X has detracted from Y? What else has helped to bring about Y?
      • “… the doctrine of…” — What is that doctrine? Is there only one doctrine, or does it consist of multiple doctrines?
      • “… by design…” — Who is the designer? What were the designer’s aims? What other choices did the designer have?
      • “X is inherently anti-democratic…” — Is X inherently anti-democratic, or is it possible for X to be done in a non-anti-democratiic manner? Is there a difference between being anti-democratic and being non-democratic?
      • “… a court should defer to the expertise of the democratically elected, democratically accountable institution…” — Is that institution both democratically elected and democratically accountable? Is it always? What do democratic credentials have to do with “expertise”?
      • “X is undesirable because it is Y” — Is X undesirable? Is X Y? Is Y the reason why X is undesirable?

      Miscellaneous feedback on particular essay assignments

      In one year, I observed as follows:

      The best essays squarely addressed the question asked through argument. Most essays did at least pay lip service to the question asked, in that they acknowledged certain terms used by the question that required a definition. Beyond this, what was needed was a sustained effort to address the question. Weak responses tended to say something about what the question meant, but then go on to discuss something not entirely related, only answering the question proper right at the end.

      A number of essays went significantly over the word limit; marks were deducted for this, particularly when it was not necessary for the essay to go over the word limit in order to make the points that it sought to make. While no marks were deducted in respect of essays which were shorter than the word limit, it tended to be the case that these essays could have benefited from more detail.

      In general, when answering questions such as the ones that had been asked, it is not a good idea to provide lengthy passages that state facts (including developments in the law) but do not explain what argument is being made using those facts. The strongest essays contained only a small section (if any at all) that contained an overview of certain facts. Instead, such essays were driven by arguments; facts were only used to support arguments. This approach meant that only relevant facts were stated, and no more. Weaker essays, by contrast, contained lengthy passages that narrated facts, but contained little analysis; the result sometimes came across more like an encyclopaedia entry than a paper that sought to advance an argument. 

      While the word limit was a real challenge, it could often have been made easier to meet this challenge by condensing or eliminating passages that were merely descriptive so as to make room for analysis. 

      Some essays contained lengthy passages devoted to summarising others’ ideas. The use of others’ ideas is, of course, not in and of itself wrong, but what is needed is an application of those ideas to the particular question asked.


      In another year, I observed as follows:

      The strong essays were well-researched (and some were very well-researched indeed).

      They were organised sensibly so that the argument was easy to follow. There were signposts and transitions that guided the reader along a smooth path rather than making the reader bash through a dense jungle.

      They contained well-substantiated arguments; there were no unjustified assumptions. For example, when they asserted that “many people” believed something, they cited a source to back that up. They did not assert that there is definitely causation when there is only evidence of correlation.

      They did not just tell the reader facts; they used these facts to show the reader something, and they made clear to the reader what that thing was. At the same time, they did not make more of the facts than was justified: for example, they did not treat a single anecdote as though it were data.

      They provided sufficient explanation to let the reader understand the significance of certain facts, while not dumping so much factual detail on the reader that the significance of those facts was lost. For example, strong essays did not name-drop [a government scheme] without explaining what that is; they did not just assert that a law does something, but rather went into detail about which part of the law does what, and precisely how; yet, when they stated the implications of Singapore entering into [an international treaty], they did not provide multiple paragraphs stating the law of international treaties in the abstract.

      If they used slogans or platitudes, they took them as the starting point of analysis and unpacked them; when they did use slogans or platitudes as a conclusion, they argued for the conclusion, and did not treat them as self-evident; they recognised the eloquence and brevity of such statements without treating these attributes as necessary indicators of substantive soundness or correctness.

      On the meaning of “begs the question”, see here.

      Read questions carefully

      Suppose there is a question about whether Article 15 of the Constitution (on freedom of religion) means that conscientious objection to national service should be recognised. That question is not merely about whether conscientious objection should be recognised; it is about whether it should be by reason of Art 15.

      Whether something is “adequate”

      When considering whether something has been “adequate”, there are at least three questions that we can consider asking:

      1. What is the thing whose adequacy we are measuring? (In the case of legislation, is it only the legislation itself, or also the process leading up to the enactment of the legislation?)
      2. What is/are the standard(s) (‘yardstick(s)’) by which to measure adequacy?
      3. How does the ‘thing’ measure up against those standards?

      If you struggle with point 2, try asking about the different sources of standards. Who sets the standards by which X is judged? There are at least two possibilities to consider:

      1. Standards set by others.
      2. Standards set by X itself.

      Taxonomy

      Several of you have asked questions relating to taxonomy, e.g.: When a decision-maker takes into account irrelevant considerations, should we say that it has acted illegally or that it has acted irrationality? If proportionality is accepted as a ground of judicial review, is it a species of illegality; a species of irrationality; or something that exists alongside these categories? Is the law on procedural legitimate expectations better thought of as being part of the law of procedural fairness or part of the law of legitimate expectations? etc.

      As in other areas of law (such as resulting trusts; constructive trusts; grounds on which restitution of unjust enrichment may be sought; and whether there are still such things as conditions and warranties in contract law), I am afraid that I can offer no easy answers. What I ask you to do is to consider why taxonomy matters.

      Suppose you have two categories, A and B. Suppose you want to argue that the courts (/whoever) are wrong to say that X falls within category A; you want to say that X should fall within category B instead. What is the point of saying this? Is it just a matter of applying a label? Does it just mean that, if you were writing a textbook, you would write about X in chapter B instead of chapter A — a mere question of rearrangement? If so, then, one may ask: is there really any practical point in quibbling over categorisation?

      An example on when categorisation does matter is this: when there is some rule (R) that applies to category A but not B. In that case, it makes a difference which category X is in, because that would make a difference as to whether rule R applies to X or not. Take, for example, the question of whether you would call a mushroom a “vegetable” or a “fungus”. Why does it matter? One reason why it matters is that, under section 7 of the Control of Plants Act, one needs a licence to “import for sale, supply or distribution” a “vegetable”.* That is one reason why classification matters. If I had not mentioned that statute, you might legitimately ask: who cares whether a mushroom is a “vegetable” or a “fungus”?**

      Of course, remember that categories aren’t necessarily mutually exclusive. Something can be both a vegetable and a fungus.

      One might ask: Isn’t intellectual honesty/neatness a good enough reason in and of itself to categorise something accurately? Perhaps. But that presupposes that intellectual honesty/neatness demands that there be multiple possible ways of categorising that thing in the first place. In other words, it would be somewhat pointless to stop at saying that a lime must be categorised differently from a lemon because they are different colours. This is because, one may ask: what is the point of classifying them by colour in the first place? Why not just have one category of ‘citrus fruits’ (or even just ‘fruits’) and be done with it?

      More generally, when we design a system of taxonomy, it is because we have a certain aim in mind:

      • Is a tomato a fruit or a vegetable? It is probably the former in a system of classification designed to serve biologists, but the latter in a system designed to serve chefs.
      • Would it be useful to divide living things into three categories: “plants that are small and easy to care for at home”; “animals that are small and easy to care for at home; and “other living things”? This would be useless for biologists, but perhaps useful for a person who is seeking companionship and/or a sense of responsibility in life.

      *For the avoidance of doubt, I express no view on whether a mushroom is a “vegetable” within the meaning set out in section 2 of the Act. 

      **Yes, I know that the obvious answer is: “a biologist”. But I hope that the point I am trying to make is sufficiently clear.

      Systems of taxonomy are not set in stone. Think of the classification of various types of plaintiff in occupiers’ liability cases in See Toh Siew Kee. There, the court said that the traditional classification into invitees, licensees, and trespassers arose for largely historical reasons, viz. that that classification arose before the idea of a general tort of negligence developed. Again, there was a time when Donoghue v Stevenson was seen as part of the law on liability for defective products rather than part of the general law of negligence. Similarly, on the account of Sedley in Lions Under the Throne, there was a time when what we would now call administrative law cases were once classified under the law of local government, the law of licensing, etc.

      With this — the idea that taxonomies can change over time — in mind, think next of, say, the idea in Anwar Patrick Adrian that White v Jones may be seen as part of the Albazero line of cases. This was, at least to me, very interesting because it raises the issue of to what extent there is a rigid distinction between contract law and tort law.  The next question we should ask is: well, why is there this distinction in the first place? It would, of course, make a difference if we were trying to come up with a complete and doctrinally sound account of the law of obligations in Singapore generally (sed quaere, of course, whether such a category exists; or should the ‘law of obligations’ not be seen as distinct from the ‘law of property’? etc.) It might make a difference if there were some conflicts-of-laws issues involved laying down different rules for contractual vs. tortious claims. Again, it might make a difference if, hypothetically, the Limitation Act provided for different limitation periods for contractual claims and tortious claims. On the other hand, in the absence of such issues, it would not make a difference in a simple case in which you are trying to sue a lawyer who was negligent in drawing up a will: you would simply say to the court, ‘I should win because this case is analogous to Anwar Patrick Adrian‘.

      Similarly, when would you need to ask about the classification of the grounds of review? If the classification makes no sense, you should feel uneasy, just as Daniel Tan, Paul Craig, Mark Elliott, et al. (and many, many others, whose works are too numerous to feature on our reading lists) have been bothered enough to be moved to write articles on such topics. At the same time, you should ask whether injustice is thereby done; or whether, despite doctrinal infelicities, one is able to go to court, say ‘That is Wednesbury unreasonable’, be more or less reasonably understood, and have justice done for one’s client. (I am not sure whether there is a clear answer to this question: I would leave it to you to form an opinion with the materials that we have.)

      ‘Checks and balances’

      When you speak of ‘checks and balances’, consider: what risk is being checked against? So when you speak of replacing one ‘checking’ mechanism (A) with another (B), don’t only ask whether A is ‘more’/’less’ effective (as though you were comparing A and B quantitatively); ask: do A and B guard against the same kind of risks in the first place?

      Even if someone can’t guarantee that something is prevented from happening, that person can still serve as a check against the risk of it happening. You can check against the risk of something happening, not only by making sure that it doesn’t happen, but also by making it harder for it to happen, making it slower to happen, nudging the person who makes it happen to reconsider their decision, making it more obvious to the public that it is happening, etc.

      Changing from one thing to another

      When you propose that something new be introduced, don’t just stop at commenting at the merits of that thing in and of itself; instead, consider the merits of that thing compared to the status quo. If someone is proposing introducing X, but we already have Y, and Y is a perfect substitute to X, then perhaps it is worth stating, not only that X is good (even if it is), but also that X is ultimately unnecessary.

      When considering the possibility of changing from one thing to another, consider what would thereby be gained and what would thereby be lost.

      Is there ever a reason to keep something that we know is sub-optimal, rather than replacing it? Possible reasons include:

      1. The process of change would be difficult/impossible
      2. We can’t think of a better way
      3. We can think of a better way, but we can’t implement it
      4. While the status quo is bad, the underlying aim it aims to achieve is good — so we should think of partial reform rather than replacement

      Question: In light of limited space / limited time / the limited nature (if any) of the materials and case law available in Singapore, should questions be ‘read down’ / addressed narrowly?

      Answer: You can make decisions as to your focus, but you cannot ignore parts of the question or change it.  (If I may, you may interpret the question, but must not in substance rewrite it.) In this regard, you may find it useful to explain, in your introduction, what you will (and will not) be focusing on and why.

      On a related note: If a reasonable reader would think that a complete answer to the question must address A, B, and C, and if you only address A and B, then you should explain why you are not addressing C. Otherwise, one might get the impression that the paper does not completely answer the question because it doesn’t address C.

      This is, by the way, similar to other general skills such as explaining one’s methodology and stating one’s assumptions.

      What could be a valid reason for omitting C? Possibilities include the following (all of which, of course, would need to be substantiated):

      • At first glance C is relevant to the question, but actually it is not.
      • C is actually a subset of A and/or B.
      • The paper illustrates a broad principle. It is trite that that principle applies to C. What is less clear is whether the principle also applies to A and B. Therefore, the bulk of the discussion will focus on A and B.
      • C is theoretically relevant, but it hardly arises in practice.

      Note that all of these require an explanation of why it is justifiable not to discuss C. It is not enough to say that you are not discussing C because of time/space constraints. This is because one might retort: if you want to save time and space, why not omit A or B instead of omitting C? Even if your aim is to save space, something needs to say about why C is the thing that is omitted in order to achieve this aim.

      Choosing a case study to focus on

      Some of you asked: Can I pick one specific case study to focus on? For example, in writing about educating people about Singapore’s constitution, can you just focus on one aspect of the constitution? In writing about restorative justice, can you just focus on cases involving one particular offence/type of offence/type of offender/type of victim?

      The answer is: The questions speak in general terms, so ideally you should say something about the general point. However, you are free to focus on a small number of case studies — but use them to illustrate the general point. For example:

      • E presents an interesting case study in the complexities of doing X. A, B, C, and D are straightforward, but E (like F and G) shows that X is far trickier than appears at first glance.
      • It is useful to use A as a proxy for X generally. After all, A is the most common/most serious case of X / other cases like B, C, and D are similar to A.
      • It is clear that X encompasses A, B, and C. But how far can we push the idea? Would X also encompass D and E? This is useful because it tells us about the conceptual limits to X, and when it is time to shift focus away from X and toward Y instead.

      The fact that you have space/time constraints is not itself a reason to focus on (say) A. It is a reason to focus on one thing, but one may ask: why A, and not B or C? Why compare the law of Singapore with that of (say) the UK, out of all the countries in the world? 

      Consider whether a reader might criticise your choice of A as unrepresentative because A is an aberration.

      You are free to discuss a hypothetical situation — say, to criticise a law because it would be insufficient to address the problems arising in a hypothetical case, C. But consider whether the reader might criticise your hypothetical example as being too far-fetched, unrealistic, implausible, etc. — and if so, consider pre-empting such an argument and responding.

      On the other hand, don’t ignore outliers. That would be like saying that statelessness isn’t a problem in Singapore because there are only a few hundred stateless people compared to the total population of millions: a difficulty isn’t any less of a difficulty merely because few people face it; if you are tempted to think otherwise, imagine that you were one of those few hundred! It may be a policy choice to devote less resources toward tackling a problem faced by relatively few people, but that does not make it a small problem or a non-problem. 

      (I’m going off on a tangent, but I’m reminded of this blog post by Gautam Bhatiacriticising a 2013 decision of the Supreme Court of India by which the court refused to declare their equivalent of Singapore’s former s 377A of the Penal Code unconstitutional. We don’t need to concern ourselves with the details, but one thing the Supreme Court said was that “a miniscule fraction of the country’s population constitute lesbians, gays, bisexuals or transgenders and in last more than 150 years less than 200 persons have been prosecuted (as per the reported orders) for committing offence under Section 377 IPC…” Bhatia had this to say:

      “The Court then makes a truly bizarre observation, holding that because only 200 people have been prosecuted, there is no discrimination. Two hundred people. Suppose the State randomly catches hold of 200 people and shoots them dead. This, according to the learned Judge, would not be an Article 21 violation because, well, it’s just 200 people. Our Constitution does value human life – but only when it’s more than 200 people. Remember that.”

      Please don’t hesitate to cite others’ work, quote from others, or state facts presented by others…

      … but remember to do something with what you cite. You might say “Professor X says that Y because Z”, but explain how that figures into your argument: is Professor X correct? is her conclusion wrong because her methodology is wrong, or because her reasoning is fallacious, etc.? is her view inapplicable in Singapore because Z is not true in Singapore? does her view raise the question whether Y is true in Singapore, given that something like Z appears to be true in Singapore?

      Also, sometimes there are limits to what you can infer from data — and that’s fine. Consider, for example:

      • “Professor X has found that in England, Y because Z. In Singapore, Z is not strictly true, but government statics suggest that something similar to Z appears to be true. This raises the question: is there the risk of Y in Singapore? 
      • Perhaps not, but we cannot dismiss the possibility. The stakes are high: [elaborate]
      • It is about time, then, that somebody conducted a study to see whether Z is true in Singapore. Of course, that is easier said than done. [elaborate]”

      Making an original point that builds on others’ work

      Even if your point is original but has drawn inspiration/support from elsewhere, it is perfectly fine to say so, acknowledging the support without detracting from the originality of the point. For example:

      1. “A discusses point X. His discussion of point X relies, in great part, on the doctrine of parliamentary sovereignty in the UK, which does not apply in Singapore. Nonetheless, in the Singapore context, one may modify point X by…”
      2. “In the context of discussing X, A states that… Taking this to its logical conclusion, in the case of Y…”
      3. “A points out that… But something more is needed to apply this line of thought in situation Y…”
      4. In a footnote: “Inspiration for this point is drawn from A, who states that…”
      5. “A similar point was first made by A, albeit in the context of…”

      What does it mean to “critically discuss” a statement?

      When you are directed to “critically discuss” a statement, you are free to not only to examine the issues raised in the statement, but also (if appropriate) express a view on the correctness of the statement. The point is to respond thoughtfully to the statement in a manner that you consider appropriate.

      Question: To what extent is it necessary to define concepts mentioned in the questions? What if the very definitions of these concepts is questionable/contested/contestable?

      Answer: In general, you should address the definitions of concepts mentioned in the question, particularly when your answer turns on these definitions. Also, in general, sometimes definitions are necessary by way of introduction; in other cases, the crux of the paper is really about what definition applies/ought to apply. If you have to choose one out of several possible definitions, then make clear that you are choosing and explain why that is your choice. Similarly, if the (possible) definition(s) is/are contestable, unclear, or of dubious utility, then say so and explain why. 

      “If”

      The word “if” is your good friend.

      Consider: 

      • The first question is… The answer could be A or B. I think it is A because… If so, then… But if the answer is B, then that leads us into the next question… 
      • At first glance, that question is an important one. But it is actually immaterial. If the answer is A, then… But if the answer is B, we reach the same result, namely…
      • The answer really depends on what one means by “helpful”. If we take “helpful” to mean A then… On the other hand, if we take helpful to mean B, then… For present purposes, let us consider definition A. Definition B is superficially attractive, because… But the better view is that…
      • The law in Singapore is A, according to which the applicant’s claim would fail. But even if the law were B (as X has argued it should be), the applicant’s claim would still fail because…

      If you think/write like this where appropriate, you will demonstrate both that you can express a view, and that you are capable of considering alternative views.

      What kind of arguments can you raise when answering a hypothetical problem question?

      Anything appropriate. In particular, when you are asked to pretend that you are arguing a case before a court (/advising someone as to whether to do so), you are free to:

      • state what the law is, and state how it would be applied
      • argue that a previously decided case should be distinguished, and hence has no relevance to the present case
      • argue that a previously decided case is wrong and should not be followed
      • argue that there is no established law on a certain matter, and then propose what the law should be

      Of course, not all of these arguments will be equally readily accepted by a court. The third, in particular, has the limitation that a lower court is bound by a higher court’s decisions. So you should still try to use the skill of prioritising your arguments.

      For these reasons, it is not true that, when answering a hypothetical question, you are bound only to state what the law is, and not what you think the law should be.  (And, even when you are doing the former, it is — as you should know by now — often far from straightforward what the law is.)

      Question: Do you have to provide “both a thesis and anti-thesis”?

      Answer: You simply have to write a good essay. A good essay, among other things, follows the following advice by Mark Elliott:

      “In order to set out your argument in a persuasive manner, you need to deal both with relevant points that support your argument and with relevant points that appear to challenge your argument — and, in dealing with the latter points, you need to show why they do not in fact fatally undermine your argument. In other words, the approach I’m suggesting here doesn’t mean that you should adopt a blinkered approach, paying no attention to counterarguments: rather, you need to deal with them in a way that shows that, having thought about and weighed them in the balance, you are in a position to show why your argument stands in spite of them (or why your argument can be adapted in a way that accommodates such points).

      All of this points towards a further matter: namely, that advancing an argument in your essay does not mean that you need to (or should) be argumentative in the sense of adopting a strident tone that brooks no debate or compromise. Rather, advancing an argument in the way I’ve suggested here means being thoughtful and persuasive: taking the reader with you on a journey that demonstrates that you have looked at the relevant material, carefully thought through the issues raised by the question, and arrived at a view that you are able to justify and defend through well-reasoned and suitably evidenced argument.”

      This applies to writing all sorts of essays, including during an exam. 

      I would add one more point: A “relevant poin[t] that appear[s] to challenge your argument” doesn’t necessarily mean a possible point that appears to totally contradict your argument. For example, you might concede that one of your arguments has a potential weakness, but that is not fatal to your argument; or you might point out that some of your arguments will hold only under certain conditions, but that those conditions are likely to hold; or you might say that the opposing position is not totally wrong, but just on balance less strong.

      Similarly, sometimes, you might argue that a certain statement is not true, yet you might concede that the statement, while false on the whole, does contain a grain of truth.

      If you are concerned about time constraints, remember that everybody is under time constraints; your task is not to write the best possible answers in the world, but rather to write the best answers you can write given time and other constraints.

      General comments on style and grammar

      It can be hard for the reader to keep track of idiosyncratic abbreviations.

      Headings should guide the reader by providing a sneak preview of what each section or sub-section of the essay is about. Hence, a heading like “Evaluation” or “Short Answer” does not help much. 

      While it is good for the introduction to provide a summary of the answer to the question, this can be done more elegantly than through a section labelled “Short Answer”.

      A statement that “begs the question” is one that contains circular reasoning, in that it assumes (or merely asserts) what it purports to prove. “Begs the question” does not mean the same thing as “raises the question”.

      Avoid dangling modifiers.

      When you use a verb, you should generally make clear who is ‘doing’ the verb. The easiest way to do this is to make it a point to use the active voice (e.g. “The High Court said that…”) Using the passive voice (e.g. “It was said by the High Court that…”) is sometimes clunky, but not in and of itself wrong. What is undesirable is using the passive voice in a manner that does not mention the subject — for example:

      • “It was said that…” — who said?
      • “It is thought that” — who thinks?

      If you are arguing something, there is no need to preface it with “It is argued that” or “This essay argues that” if the context makes clear that what you are saying is your argument and not someone else’s. For example, consider: “In X v Y, the High Court held that… But the High Court fell into the trap of A…” There is no need to say “It is argued that the High Court fell into the trap of A”. In fact, if you do say that, the reader is left to wonder who “argued that the High Court fell into the trap of A”.

      Typically, an author defines certain words near the start of a piece of writing before proceeding to use those words. If you quote only the words but not the definition, the reader will often be left wondering what the words mean.

      When quoting others’ words, use quotation marks or an indented block quotation. If you must change the original words, use square brackets. Use “emphasis in original”, “emphasis added”, or “emphasis removed” as necessary.

      For example, if I said “I want to go to the zoo.”, you could write:

      Ong said that he “want[ed] to go to the zoo”.

      Be wary of online citation generators, which can produce results that are not useful to the reader (for example, citations that do not tell the reader where a certain article might be found, or when a certain website was published). Instead, I recommend that you try to learn a style such as the SAL style or OSCOLA and get into the habit of producing citations manually; doing so will soon become second nature, which will be particularly useful if you go into litigation.

    28. Feedback on a student presentation about access to a lawyer before/during police investigations

      Like all procedural rules, the rule that one doesn’t have immediate access to a lawyer on arrest must be seen in the context of other rules. Even if one doesn’t have access to a lawyer during police questioning, that doesn’t mean that someone can be convicted on the basis of police statements that were unlawfully obtained. We see this from the Kadar case, as well PP v Dahalan bin Ladaewa [1995] 2 SLR(R) 124 (which the Thursday team also mentioned). More generally, the court will not convict a person merely based on what they said to the police: the ultimate question is whether all the evidence convinces the court beyond a reasonable doubt that the person is guilty.

      There is also s 258(3) of the Criminal Procedure Code 2010, which says that statements obtained in certain circumstances are strictly inadmissible as evidence. One could argue that the scope of s 258(3) is too narrow, but that doesn’t mean that there should be a right of access to counsel during questioning: it could just as well mean that s 258(3) should be broadened. Further, as we see from Kadar and Dahalan, the courts can exclude a statement by the accused person on other grounds – for example, because the statement was recorded in an irregular manner that rendered it unreliable, because (as in Dahalan) the person did not understand English well and did not have access to proper translation, or because (as in Dahalan) the person was under the influence of drugs and therefore “not in a fit state to be interviewed” ([77]).

      Note also that even if a piece of evidence is admissible, the court might not necessarily give heavy weight to it. I won’t fault anybody for not having known this distinction, but it is one worth noting now. For example, the courts have held that a statement given by a person who is tired is admissible, which means that the court can consider it. But, nonetheless, the court might choose to give it little weight, for example, by holding that it is unreliable.

      Conversely, even if there is access to a lawyer before/during police questioning, that doesn’t mean that the lawyer is free to (for example) encourage the arrested person to lie; assist the arrested person in destroying evidence; etc. Those things are crimes, and also go against the Legal Profession (Professional Conduct) Rules. 

      It would also be helpful to consider possible variants of a proposed rule that one have a right of access to a lawyer during police questioning. In the first place, can the lawyer say anything? Or could the lawyer be allowed only to observe (and note down anything the lawyer thinks is improper, so that the lawyer can mention it in court in due course)? If the lawyer objects to a certain line of questioning, would it follow that the person being questioned does not need to answer it? Who would evaluate the lawyer’s objection?

      Note also that – as [one team of students] stated – it’s possible to argue that there should be access to a lawyer during police questioning in certain circumstances but not others, etc.

      A right of access to a lawyer wouldn’t necessarily mean that the state is obliged to provide and pay for a lawyer. It could simply mean an reasonable opportunity to find and engage a lawyer at one’s own expense.

      I really appreciated [one team’s] discussion of whether the answer to the question should change depending on whether the process of recording the statement is captured on video. I suppose the answer to that question depends on who will have access to the video and when.

      Something could have been said about this pamphlet which is given to people who have been arrested.

      Both teams cited a few Singaporean cases; here’s a summary of what the courts in Singapore and Malaysia have said:

      • The Malaysian courts, citing Indian cases, have held that the constitutional right to “consult” a lawyer doesn’t refer to having the lawyer with you when being questioned by the police. This, they say, is because the word “consult” means to seek legal advice, and legal advice is by its nature confidential; therefore, a “consult[ation]” with a lawyer can’t possibly take place in front of the police. (I add that the word “be defended by”, in legal parlance, refers to being defended in court, so it’s not relevant here)
      • Moreover, you don’t get a confidential consultation with the lawyer immediately upon arrest. The Malaysian courts have offered two reasons for this: (a) that may, paradoxically, delay your implied right to a speedy trial; (b) that needs to be balanced against the police’s need to discharge their duties efficiently. I’m not convinced by either argument. I don’t understand the first argument: even if you have a right to a speedy trial, shouldn’t you be allowed to waive this right? As for the second argument: while the police’s duties are important, they – unlike the right to counsel – do not have constitutional status.

      The Singapore courts have endorsed the second argument. The High Court in Leong Siew Chorsaid that the accused person has the right to consult a lawyer only after a “reasonable time”, and the “reasonable time” requirement is meant to accommodate “the duty of the police to follow up on new leads quickly and to gather swiftly whatever evidence was available lest it disappears or is destroyed”. The Court of Appeal tentatively agreed: it felt unable to express a conclusion given that Leong’s lawyers did not make detailed arguments on the precise benefits that access to counsel would have brought to Leong. (To put it in context: The lawyers argued that the effect of not having access to counsel was that the police statements were unreliable. This needed substantiation, which was lacking.)

      The next milestone is the James Raj case, where the High Court reflected on what a “reasonable time” is. The High Court noted that the courts, in previous cases, had assumed that “the rationale for granting a ‘reasonable time’ was to afford the police a degree of latitude in carrying out their investigations.” Hence, the court took the view that the burden of proof is on the police to prove that “giving effect to the right to counsel would impede police investigation or the administration of justice”. Further, the court went into considerable detail in assessing the police’s arguments about what a “reasonable time” was, and concluded that there was no evidence that, here, “access to counsel in this case would have hindered investigations”. In other words, the High Court did accept that there is only a right to see a lawyer after a “reasonable time”, but it suggested that a “reasonable time” is no more than the police can show is necessary. Moreover, the High Court affirmed that the law is not that you only get to see a lawyer once investigations are totally complete (at [14]). It even said that it is “at least arguable” that the “reasonable time” refers only to “any necessary or unavoidable delay occasioned by practical or administrative concerns, eg, having to transport the arrested person to the place of remand or having to contact the counsel of the arrested person’s choice”.

      Unfortunately, the Court of Appeal stopped at saying: “With great respect to the Judge, the position on the issues raised in the present application has been clear and settled since the decision of this court [i.e. the court of Appeal] in Jasbir Singh.”

      Speaking for myself, I do like the approach of the High Court in James Raj, since it – at the least – recognised that there is a conversation to be had about what is a “reasonable time” and what is not. As Tham Lijing puts it, “access to counsel is not necessarily an ‘all or nothing’ dichotomy”. For example, he says, there could conceivably be “flexible access to counsel depending on the particular circumstances of the case – eg one hour for the lawyer every week to explain to the suspect his basic rights and options”. Professor Ho Hock Lai makes the more general point (in ‘Recent (Non-) Developments in an Arrested Person’s Right to Counsel [2014] SJLS 267 at 281) that the “fairest way forward” would be for the courts to elaborate on what precisely a “reasonable time” is. I would add that perhaps, in certain situations, it might be “reasonable” to allow access to a lawyer even before police statements are taken.

      One final thought: There is no law saying that the police must allow access to a lawyer during questioning. But there is no law saying that they mustn’t. In other words, they have the discretion to do so – or, more generally, to allow other persons to be present during the questioning. This must be so; otherwise, it would be unlawful to have the Appropriate Adult Scheme. If – as I have argued – that discretion exists, what might be the legal limits to that discretion?

    29. Feedback on a student presentation about modern slavery

      As [one team of students] pointed out, ‘traditional’ slavery means human beings being treated as property. Modern slavery, then, shares certain features in common with ‘traditional’ slavery, but what? Perhaps the following might be food for thought – I’m no historian, but the following is what I remember from when I was a student:

      • In Roman law, there was a distinction between ‘persons’ and ‘things’: ‘persons’ have legal rights and duties, whereas ‘things’ are the object of legal rights and duties. Slaves were ‘things’, not ‘persons’.
      • This had several legal consequences. For example, a slave could not own property, because property ownership is a kind of right, and only ‘persons’ can own ‘things’; a ‘thing’ can’t own another ‘thing’. A slave might have been allowed to have some personal possessions (including being paid), but that was only because the ‘owner’ of the slave allowed it; the slave had no legal right to those things. If you killed a slave, the ‘owner’ could sue you for causing property damage.
      • The defining feature of slavery was not quite (a) doing menial work or (b) being treated badly. In ancient Rome, a doctor or an accountant could have been a slave. They could have been treated well. But they did not have the right to be treated well. What they had, they had (for the most part) because of the beneficence of their ‘owners’ and not because they had a legal right to it. (That said, I should point out that I do not know what proportion of slaves did what kind of work, how they were treated, how much they were paid, etc. I suppose it would be fascinating to study the interplay between the law of slavery and the socio-economic forces determining how slaves were treated in practice – but that’s for another day.)
      • One right that slaves lacked was autonomy. That said, slavery was not the only form of legal status that took away a person’s autonomy. Consider the system of feudalism in operation in (at least) England. Nowadays, there are comparatively few landowners; often, people rent land (and/or buildings) from the owners (‘landlords’) in exchange for money. But under feudalism, the landlord was more literally a lord, with not only economic but also legal power, including the power to make rules governing the manor. As a tenant, you did not simply pay money; you had to perform certain services, such as performing military service for the lord, paying the lord in the form of grain, etc. You could pay for these things – say, hire someone to do the military service, or buy grain from someone else – but the fact remained that you were liable to do these things.
      • Moreover, there wasn’t just one owner and one tenant; there could be a whole chain of ownership, with the king/queen owning all the land in England; lords below that; and so on… with serfs right at the bottom. Serfs were especially un-free; in fact, “unfree” was the technical term to describe them. Not only were they obliged to do certain things (like pretty much everyone else – even lords were subordinate to the king/queen); they didn’t even have a fixed job.

      Of course, modern slavery need not have all the features of traditional slavery or serfdom – otherwise, we would just call it “slavery” or “serfdom”. That said, merely working for someone isn’t slavery/serfdom; neither is merely doing unpleasant work. If you asked me to boil slavery/serfdom to just one point, it would be a lack of control over one’s life.

      Of course, there can be degrees of lack of control. If, for example, someone forces you to work by threatening to kill you if you don’t, we would have no hesitation in saying that you have no choice. But suppose that person threatens to, say, make it harder for you to purchase food, or restrict your movement? Suppose that if you don’t work then you won’t have a home/food – would that count as slavery? Would it matter if the threat of homelessness/hunger/etc. isn’t imposed deliberately by one person, but is instead the product of a combination of socio-economic circumstances?

      As both teams rightly pointed out, there can be a law about something without using a certain term for that thing. But, as we have seen earlier in the course, and as both teams reminded us, labels can matter. The word “slavery” has certain connotations that no other word has. Also, by framing the problem as one of “modern slavery” rather than multiple problems of trafficking, forced labour, etc., we prevent thinking in silos. On the other hand, one potential problem with focusing on “modern slavery” is that, by focusing our attention on the worst/most egregious/most obvious situations that are akin to slavery, we may lose sight of other situations that are less egregious but still bad, more subtle or obscure, etc.

      It’s possible for a law to target something without explicitly outlawing that thing. For example, some of the mandatory conditions attached to work passes under the Employment of Foreign Manpower Act might have the effect of making modern slavery easy to detect, reducing the likelihood of conditions in which exploitation can occur, etc. […] Even a requirement that workers have time off work can help, in that it can be an opportunity for the worker to seek help if they are being exploited. (I do acknowledge the point that it can be difficult to detect breaches of the requirement that employers give workers time off work, but there are some solutions, e.g. surprise inspections including interviews with the worker in the absence of the employer – which I believe MOM sometimes does, at least for migrant domestic workers.)

      I appreciate [one team’s] pointing out the role of NGOs – we shouldn’t forget the role that they can play in governance. This was a good example of how one can go into more detail than stopping at saying that the government ‘should do more’, etc. […] If there had been more time, it would have been good for both teams to say a little more about what precisely the Anti-Slavery Commissions in the UK/Australia/etc. do, and whether they have improved things. The Commissions might, for example, organise campaigns, but one may ask: didn’t campaigns exist before that too?

      That said, while labels matter, they aren’t the only thing that matters. For example, while there isn’t a body named ‘Anti-Slavery Commission’ in Singapore, there are various bodies that do some of the things that the Commissions in the UK/Australia do – consider, for example, the MOM’s ACE Group? (I’m not saying that their work is perfect; I’m just saying that if you want to draw a comparison, they are the entities in Singapore which you may want to consider drawing comparisons with.)

      Sometimes, we talk about the need for those in power to take into account the needs, views, etc. of people. But some people may have difficulties making themselves heard – whether because of systemic issues or other reasons – and NGOs can help in that regard.

      Note that it’s possible for a country to do what some international treaty requires, without necessarily signing up to that treaty.

      There are laws requiring that certain commercial entities report certain matters; don’t confuse this requirement with an obligation to investigate:

      • Section 54(4) of the UK’s Act requires each such entity to report “the steps the organisation has taken during the financial year to ensure that slavery and human trafficking is not taking place — (i) in any of its supply chains, and (ii) in any part of its own business”, or to issue “a statement that the organisation has taken no such steps”. This is not an obligation to investigate anything.
      • Similarly, in Australia, certain entities have to “describe the risks of modern slavery practices in the operations and supply chains of the reporting entity” and “describe the actions taken by the reporting entity and any entity that the reporting entity owns or controls, to assess and address those risks, including due diligence and remediation processes” (Modern Slavery Act 2018 (Cth), section 16), but this appears to me to stop short of an obligation to take any particular action to address the risks.
      • Besides, there is the question of what the consequences of non-compliance with reporting requirements are, which in turn may have an impact on the rate of non-compliance (which the Wednesday team discussed). It might have been useful to present some examples of reports by businesses and comment on how useful they are.

      One team referred to the finding by Research Across Borders that 60% of domestic workers are “exploited”. As with all statistics, we must consider the methodology, as did the MOM in its response. I don’t have the time to consider this point in greater detail, but I would add this: The “60%” figure, taken in isolation, might be of limited usefulness, if only because it purports to encompass multiple types of “exploitation” – which may well be solved in different ways. Even taking the Research Across Borders report at face value, I would suggest that more specific findings – for example, the rate of late salary payment – might be more useful.

      One last point – [one student] asked a very interesting question: what if certain acts of exploitation take place overseas? Can Singapore law do nothing? The answer is not necessarily no. For one thing, as a matter of international law, it is generally accepted that States are entitled to regulate not only conduct within their territories, but also conduct of their nationals anywhere in the world. (See, for example, section 376C of the Penal Code 1871.) Also, even if a contract is entered into overseas, that doesn’t mean that the Singapore courts don’t have the jurisdiction to hear and decide disputes involving that contract – there’s a complicated series of rules, which form part of the body of law known as “private international law” or “the conflict of laws”. (Relevant factors include where the parties/witnesses are, and where the contract is to be performed – which itself depends on the precise content of the contract.)

    30. Feedback on a student presentation about Sentences for Enhanced Public Protection

      It is true that – as both teams pointed out – it is the court, not the executive, who gets to decide whether someone is to be sentenced to SEPP. But once it has, it is the executive who gets to decide when they are released.

      If a person is sentenced to life imprisonment, and has served 20 years of the sentence, then the Minister must consider whether a “remission order” – essentially, parole – is to be made. (I’m not sure whether the basis for this is, as the Thursday group said, only good behaviour in prison: I’d like to think that the risk to the public is also a relevant factor. In fact, I would go so far as to say that behaviour in prison is relevant only as an indicator of risk to the public if released.) If a remission order is not made, the Minister must re-consider at least once every year after that. (See section 50P of the Prisons Act 1933.) What’s the difference between this and SEPPs? The differences, it appears to me, are:

      • If someone who has been sentenced to life imprisonment is released under a remission order, the only condition is that they must not be convicted of a criminal offence and sentenced to, essentially, anything other than just a fine (section 50S). By contrast, a person serving an SEPP may, when released on licence, be subject to more conditions, such as attending counselling or other activities, mandatory drug testing, curfews, electronic tagging, etc. (section 50ZI).
      • If someone has been sentenced to life imprisonment is released under a remission order, but they breach the condition (by committing the offence), it is up to the court to decide on the consequence – namely, whether to impose an “enhanced sentence”, which may or may not be jail for life (s 50T). By contrast, under the SEPP system, it is up to the Commissioner of Prisons, who has two options:
        • First, the Commissioner may recall the person to prison (s 50ZK). This is only for up to 14 days (but the Commissioner can recall the person earlier for the purpose of investigating – typically, I suppose, when the person is suspected of having committed an offence while in prison – and this investigation period doesn’t count toward the 14 days). (I don’t know whether this option was available in the UK.)
        • Second, the Commissioner may report the matter to the Minister, who may decide to revoke the release order – which means that the person will be locked up indefinitely (s 50ZL).
      • Under the SEPP system, the Minister may release the person unconditionally (s 50ZG(7)). By contrast, it appears to me that there is no such option for people who have been sentenced to life imprisonment but who have been released under a remission order.

      As I mentioned to [one team of students]: when comparing two people’s sentences, it is necessary to avoid comparing apples and pears. [The team compared one case where someone was sentenced to indefinite imprisonment of at least 8 years, with another where someone has served 20 years of an indefinite prison sentence.] But the comparison is not between 8 years and 20 years. It is between [8 years to infinity] and [4 years to infinity], and not between [8 years] and [20 years].

      A similar point: it is not enough to argue that the SEPP system will not suffer from the same delays as the IPP system in the UK because Singapore’s courts are more efficient than the UK’s Parole Board. The comparison should be between (a) the UK’s courts and Singapore’s courts; or (b) the UK’s Parole Board and Singapore’s Minister for Home Affairs plus Detention Review Board.

      Were the problems with the IPP system in the UK mainly because of certain features of the system, or more because of how certain actors behaved? Take, for example, the phenomenon that courts in the UK were too trigger-happy when it came to imposing IPPs. Would you criticise the law in the UK for making it too easy for the court to do so? Or would you criticise the courts for being too keen to do so (when they could have chosen not to)? Or both?

      I wonder: When considering whether to release someone, should the Minister start with the presumption that the person is to be released unless he thinks there is reason not to (e.g. evidence that the person has not been rehabilitated)? Or should the starting point be that the person will not be released, unless the Minister thinks there is a positive reason to do so (e.g. evidence that the person has been rehabilitated)?

      It would have been good if both teams had considered the differences between the Parole Board in the UK, and the Minister for Home Affairs in Singapore. The latter will take into account the views of the Detention Review Board, but is not bound to follow them.

      How will the risk of re-offending be assessed? [One team of students] rightly pointed out problems with certain methods of assessment, especially machine learning-based ones. I’d just like to remind everybody that there has been no public statement on whether such systems are or should be used. I don’t know if COMPAS is used in Singapore; for a view of what is used, see this article.

      It may be true that the Minister may naturally have the incentive to err on the side of caution and keep SEPP prisoners locked up for the sake of protecting the public. But that comes at a cost, and the Minister also has the incentive to reduce costs.

    31. Feedback on a student presentation about juries

      It is often said that juries are independent, etc. But so are judges. There are also threats to juries’ independence, but again the same can be said of judges. Juries cost public money, but so do judges. Perhaps, then, we should not ask whether a jury system would be good, but rather whether it would be better than the status quo.

      What about playing on juries’ susceptibilities to particular modes of persuasion? Judges are human too, and it may well be that different types of argument or different methods of presenting a point may appeal to different judges. One may respond that judges are duty-bound to be wary of being bamboozled, but so are juries. Perhaps the difference is that judges get specialised training – which brings to mind [one team of students’] idea that there could be training for jurors.

      [On students’ argument that juries can “temper justice with mercy”:] How? If the answer is that juries are supposed to render a fair decision, so do judges. If the answer is that the accused person is tried by more than one person, well, that could be addressed by having cases heard by panels of judges rather than a single judge. Some may argue that judges retain the power to acquit someone even if they believe the person is guilty – this is known as “jury nullification”, and is controversial because it is, in essence, an act of disobedience to the law. Moreover, where there is a range of possible sentences, there is already room for “mercy” in the selection of a sentence from that range.

      Both teams pointed out that juries’ decisions as to guilt may be influenced by the possible sentence. That is true, but as a counterpoint, I would point out that judges have also said things like “There is no question that as a modality of punishment, capital punishment is different because of its irreversibility. For this reason, capital cases deserve the most anxious and searching scrutiny.” (Kho Jabing v PP [2016] 3 SLR 135 at [50]) and “A capital charge carries the ultimate penalty for an accused person and the court’s scrutiny in respect of this must not only be rigorous, but punctilious and searching.” (Imran bin Mohd Arip v PP [2021] SGCA 91 at [86]) This is not to say that the courts decide whether to find someone guilty based on the sentence that would be imposed in the even that they do; the point is simply that courts are mindful of the need to take extra care when someone’s life is on the line.

      Both teams discussed the possibility that there is no impetus for the jury system to be re-introduced: there is no public pressure, etc. That is not a reason not to bring back the jury system; rather, it is the absence of a reason to bring back the jury system. While we’re on the topic, there is a lesson here. You could categorise some of the arguments into, for example, ‘push’ and ‘pull’ factors: there is a difference between a reason that the status quo is unsatisfactory, and a reason why the solution is the jury system (instead of something else).

      Both teams could have said a bit more about was the role of judges in a jury system. […] If there were more time, both teams could also have considered issues such as whether a judge can/should be able to overrule the jury’s decision.

      Note that one feature of jury systems is that juries’ deliberations are secret, which means that you can’t find out what the jury’s reasoning process is. Contrast this with judges, who are duty-bound not only to make a decision but also to state the reasons for their decision.

      On what grounds can one appeal against a decision reached by a jury? It’s not as simple as saying that the jury misappreciated the evidence, gave undue weight to one piece of evidence, etc. This is simply because it’s not possible to know what the jury’s precise reasoning was! Therefore, in general, one can only succeed on an appeal if (a) the decision of the jury is so perverse that, in the face of the evidence, no reasonable jury could possibly have reached that decision; or (b) the judge had misdirected the jury as to what the law is. Note that different legal systems could conceivably have different rules on when one can appeal.

      I briefly shared some thoughts about other possibilities. For example, one can imagine a court consisting of a panel of trained laypersons – like Magistrate’s Courts in the UK. Or there are courts which consist of a combination of professional judges and laypersons, as in Japan.

      It’s hard to design legal institutions, isn’t it? One last thought: If a system is flawed, it could be because the design is flawed, or because the way in which it is implemented is flawed. By bearing in mind this distinction, we can better evaluate criticisms of the system. For example, consider the argument that jurors might not do a good job because they don’t want to be there. There could be two ways to address that problem: (a) get rid of the jury system altogether; (b) make jury service voluntary. Neither is perfect, but my point is that we shouldn’t forget to consider the second option.

    32. Discrimination and Article 12(2) of the Constitution

      What does “discrimination” in Art 12(2) mean?

      There isn’t an established definition of “discrimination” in Article 12(2) of the Constitution. Therefore, if you wish to raise an Art 12(2) claim, it will be for you to define “discrimination”. This can be done with whatever aid can be gleaned from Lim Meng Suang.

      One would think that Lim provides some guidance, but does it? Consider the following extracts and whether they are consistent with each other. In Lim, at [57], the court appears to use the word “discriminatory” simply to mean ‘tending to differentiate’. And at [88], the court says that Art 12(2) contains the “specific and applicable principles relating to the right to equality”, suggesting that “discrimination” refers simply to inequality – but then at [92] it is said that Art 12(2) “furnishes specific legal criteria as to what constitutes discrimination” (emphasis added; emphasis in original omitted) – read Art 12(2) and see if you can work out what these criteria are.

      By way of comparison, I would point you to the following examples of definitions of discrimination:

      1. Art 1 of CEDAW and Art 1 of the ICERD define “discrimination” in terms of the “effect or purpose” (emphasis added) of “any distinction, exclusion or restriction”.
      2. Title VII of the USA’s Civil Rights Act 1964 discusses concepts such as “intention to discriminate”, “disparate impact” (as opposed to disparate treatment); etc.
      3. The UK’s Equality Act 2010 defines “direct discrimination” and “indirect discrimination” clearly and with some precision, although this it not uncontentious: see, e.g., Sandra Fredman’s commentary on R (E) f Governing Body of JFS [2010] 2 AC 728 (UKSC).
      4. Art 9 of South Africa’s Constitution does not define “discrimination”, but does distinguish between “fair” and “unfair” discrimination; prohibit “unfair” discrimination; and, in the case of “discrimination” on certain identified grounds, raises the presumption that that discrimination is “unfair” and therefore prohibited.

      I have included the fourth item on this list because, even though it does not define discrimination and seemingly equates it to unequal treatment, it at least lays down (if I may adapt some words from Lim) some concrete legal concepts, viz. the presumption that such unequal treatment is unlawful if it is on one of a number of specified grounds. 

      By contrast, what does Art 12(2) add to our analysis? It does tell us that there is a potential issue of causation: “on the ground only”.  But (like SA’s constitution) it does not, on its face, tell us what “discrimination” means: does it require malicious intent or is it enough that the thing done has the effect of putting one group at a disadvantage as compared to another? There are also a few activities singled out for protection, e.g. carrying on a trade – but why have these been singled out? Is this singling out merely “declaratory and aspirational”, or does it mean that these activities are to be protected more than others? etc. The point is that, even though Art 12(2) is more specifically worded than Art 12(1), there is certainly much more to be said about what precisely Art 12(2) requires.


      Is it a problem that the grounds of discrimination listed in Art 12(2) are limited?

      Arguably. But this is a legislative choice with which many jurisdictions around the world have grappled: for a sampling, see Lim at [96]ff, which contains extracts of constitutional/statutory provisions from various constitutions – note how different jurisdictions identify different grounds on which discrimination is to be specifically prohibited, e.g. gender; caste; age; disability; sexual orientation.

      One can think of many more: e.g., should the law specifically prohibit discrimination against pregnant women (as compared to non-pregnant women)? What about obesity? Conscience? Culture (see http://www.saflii.org/za/cases/ZACC/2007/21media.pdf – note the reference to “South Indian Tamil Hindu culture”, as opposed to simply the Hindu religion or the Indian race)?

      And this is to say nothing about the definition of such grounds: should having HIV be considered a disability? (Note also how different jurisdictions identify different activities/areas of life for especial protection: see, e.g., how India’s constitution specifically mentions access to public amenities.) The wide range of such possibilities explains the court’s conclusion on this point at [185].

      Further reading for those who are interested:

      1. Sandra Fredman, Discrimination Law (2011) (library call number: K3242.F73 2011), especially chapter 3 (this is quite a readable book that provides an overview of the laws in a number of jurisdictions)
      2. Advanced reading: John Gardner, ‘On the Ground of Her Sex(uality)’ (1998) 18 Oxford Journal of Legal Studies 167

      What is wrong with unequal treatment? — a thought experiment

      As we have discussed in class, one problem with the sort of statute that is caught by Art 12(1) is that there is simply too remote a connection (if there is any connection at all) between the object of the statute and the way in which the object is achieved through differentiating between different people. But what else is wrong with unequal treatment? In Ramalingam Ravinthran, we see references to bias, irrelevant considerations, etc. in the discussion of Art 12(1).

      In this light, think about the following extract from Baroness Hale’s speech in Ghaidan v Godin-Mendoza at [130]:

      “My Lords, it is not so very long ago in this country that people might be refused access to a so-called “public” bar because of their sex or the colour of their skin; that a woman might automatically be paid three quarters of what a man was paid for doing exactly the same job; that a landlady offering rooms to let might lawfully put a “no blacks” notice in her window. We now realise that this was wrong. It was wrong because the sex or colour of the person was simply irrelevant to the choice which was being made: to whether he or she would be a fit and proper person to have a drink with others in a bar, to how well she might do the job, to how good a tenant or lodger he might be. It was wrong because it depended on stereotypical assumptions about what a woman or a black person might be like, assumptions which had nothing to do with the qualities of the individual involved: even if there were any reason to believe that more women than men made bad customers this was no justification for discriminating against all women. It was wrong because it was based on an irrelevant characteristic which the woman or the black did not choose and could do nothing about.”

      Consider the following thought experiment. Suppose one wanted to make such arguments to the Singapore courts. How would one fit such arguments into the scope of Art 12?  Consider: would one have more success raising such arguments under Art 12(1) on the authority of Ramalingam Ravinthran; or is it that, following Lim Meng Suang, such arguments, to the extent that they have something to do with the nature of race specifically, are best raised under Art 12(2) (and, to the extent that they have something to do with the nature of gender specifically, are to be pronounced ‘extra-legal’)?  Are the courts likely to accept (assuming that Art 12 is held to have horizontal effect) that a prospective tenant’s race is logically irrelevant to the landlord’s decision; or will they require evidence to show that it is irrelevant on the facts of this case?

      1. Would the traditional grounds of judicial review have succeeded in R (Patel) v General Medical Council [2013] 1 WLR 2801?

        A student asked about whether, in Patel, the applicant could have succeeded on traditional grounds of judicial review without reference to the concept of legitimate expectations. My response:

        In Patel, the court said that the policy was not irrational. The court also said that the application of the policy to Patel was unlawful (or, more precisely, the decision to implement the policy with immediate effect without any transitional framework was unlawful). But it did not explicitly say that the reason for this unlawfulness was irrationality. If anything, it said that a problem was that the General Medical Council had not taken into consideration the (mandatory) relevant consideration of the impact of the policy change on people like Patel: at [81].

        (Yes, the court used the phrase “Wednesbury grounds”, but it appears that the court was using the word “Wednesbury” as an umbrella term to refer to all the grounds of review mentioned in the Wednesbury case. This is not the same as Wednesbury unreasonableness, i.e. irrationality as a concept distinct from failure to take into account mandatory relevant considerations.)

        It is true that failure to take into account a mandatory relevant consideration was a well-established ground of judicial review even before the courts expressly articulated the concept of substantive legitimate expectations. But it is thanks to the concept of substantive legitimate expectations that the court even thought to consider promises as a relevant consideration in the first place.

        The court went on to say that, if it stopped there, then the remedy would only be an order to the GMC to reconsider its decision: at [82]. But the court went on to say that “it was not open to the GMC to change its policy in 2006 or again in 2010 without adopting some transitional provision that would cater for the case of this claimant”: at [85]. Is this irrationality review? The court didn’t use the word “irrationality”. But was it irrationality review in substance?Again, perhaps, but the very reason why the court would have thought to analyse separately the policyand the timeline according to which the policy was implemented is, again, the fact that a promise had been made to Patel (and others). 

        Or, of course, we could think of Patel as involving the application of a novel ground which was referred to in Coughlan as “abuse of power”. As you rightly say, we “struggle to define” this. But can we not say the same about many other grounds of review too? Can we not say, as you say, that Coughlan and Patel were decided on the basis that it would be irrational to take a certain course of action given the promise that had been made? After all, we also struggle to define irrationality. Maybe, when the courts in Coughlan and Patel were saying that the decision would not be “irrational”, what they meant was that it would not be “irrational” under the ‘normal’ Wednesbury threshold, but that this threshold would have to be heightened in cases involving the breach of a promise to someone? In other words, may we see “abuse of power” as a specific form of irrationality? Can we say — to use Paul Daly’s method of analysis — that promise-breaking is an indicium of unreasonableness?

        In the end, whether we consider substantive legitimate expectations to be a new ground of judicial review or a principle that affects how we apply the traditional grounds, the effect is the same: (a) a promise made to someone matters, in that the lawfulness of executive action stands to be judged in light of the promise; (b) the promise is, however, not determinative of the case. 

      2. The ‘red light’ and ‘green light’ models of the state, and judicial review

        In response to a student’s question about what the point is:

        You are free to argue that a model is not useful for a particular purpose, and that there exists a more precise way in which to see things. 

        On the other hand, to quote (out of context) Spandeck v DSTA at [80], might it be that “[i]ts very presence suggests that it has some substantive content that is capable of being expressed in terms of legal principles”? Can we, “[r]ather than denouncing it as a mere ‘label’… strive to infuse some meaning into it, if only so that lawyers who advise litigants and even law teachers [students?] can make some sense of the judicial formulations”?

        One might argue that this reasoning is problematic, in that it is premised on a bare appeal to history, and has the effect of entrenching any label that forms part of the status quo. But what this reasoning entrenches is only the label, not what the label applies to. Hence, in Spandeck, the Court of Appeal kept the label “proximity” but redefined its content from the ground up.

        So I wonder: Even assuming that the content of the traffic-light model is completely flawed, might there be something useful in at least the outline of the model — the premise that there are, broadly speaking, two types of path that a system of administrative law can take? Might something in the analogy of traffic lights be useful, such as the idea that the purpose of a model is to give directions to someone?

        What Spandeck did was to list several factors that tend toward/against the finding that X owes a duty of care to Y and group them all under a category called “proximity”. In so doing, what Spandeck added to the list of factors was the insight that, as they all deserve to be put in the same category, they all have something to do with each other. So the idea of “proximity” ends up not only encapsulating various legal principles, but also reminding us that it makes sense to group these legal principles together. Might the same be said of the various legal principles which, say, Chan CJ has grouped under the label “green light”?

        Finally, here are some quotations from Carol Harlow and Richard Rawlings, Law and Administration (3rd ed, 2009) (internal citations omitted) — the book which first set out the ‘red light’ and ‘green light’ theories:

        “… the breakdown — or perceived breakdown — of the doctrine of ministerial responsibility… brought cries of ‘elective dictatorship’. It is not surprising, therefore, to find many authors believing that the primary function of administrative law should be to control excesses of state power and, more precisely, subject to the rule of the law courts. Light-heartedly, we have called this conception of administrative law ‘red light theory’ because of its emphasis on control.” (23)

        “In the period between the two world wars an alternative tradition grew up, which we have called ‘green light theory’. In using this metaphor, we do not wish to suggest that green light theorists favour unrestricted or arbitrary action by the state; what one person sees as control of arbitrary power may — as Carr suggested — be experienced by another as a brake on progress. But while red light theory… favours strong judicial control of executive power, green light theory sees in administrative law a vehicle for political progress and welcomes the ‘administrative state’. In saying this, we must remember that both red and green light theories originated in earlier eras and try to understand their historical context. Both were, as Taggart reminds us: ‘forged on the anvil of the emerging welfare state…’” (31)

        “Because they look in at administration from outside, lawyers traditionally emphasise external control through adjudication. To the lawyer, law is the policeman; it operates as an external control, oft en retrospectively. But a main concern of many green light writers was, as already suggested, to minimise the influence of courts: courts, with their legalistic values, were seen as obstacles to progress, and the control that they exercise as unrepresentative and undemocratic.” (37)

        “Red light theory prioritises courts; green light theory prefers democratic or political forms of accountability.” (38)

        “Discussing red light theories, we talked of ‘control’ through courts. We did not stop to unpack the word. Control can be symbolic or real; it can mean to check, restrain or govern. Griffith and Street clearly sensed latent ambiguities, remarking that ‘A great deal turns on the meaning which is attached to the word “controls”. Banks control a river; a driver controls his car. The influence of a parent over a child may be greater than the power of a prison guard over a convict.’ Here the ‘controls’ are direct and internal rather than indirect and external. To extend our metaphors, however, a river bank may be inspected by an officer of the water board… to see that it is in good repair; a policeman may stop the driver and caution him for speeding; a health visitor may advise the child’s parents to exert a different kind of influence; and the prison guard may be questioned by the board of visitors. These are all external controls, but they are not judicial.” (39)

        Do these quotations (which I have selected in the hope that they capture the gist of what Harlow and Rawlings were saying) set out a more useful distinction than what the red/green distinction has become in Singapore? 

        Should a court take a different approach to the review of executive power depending on the existence and effectiveness of alternative means of control?

      3. Does the test in Chiu Teng v Singapore Land Authority describe when a procedural legitimate expectation will be recognised?

        No, only a substantive legitimate expectation.

        Some students have been confused by the statement at [114] that “there ought to be no difference in principle between procedural and substantive legitimate expectations”, read in isolation. But:

        1. When the court said “there ought to be no difference in principle between procedural and substantive legitimate expectations” at [114], that was immediately preceded by reasoning from “first principles” about whether legitimate expectations generally should be protected (at [111]-[113]).  So it appears that “there ought to be no difference” means that, if one is to be protected, then so is the other, because there is no difference in the reasons for protection – not necessarily that the test is the same.
        2. True, the court then went on at [115] to cite the part of Mount Sinai that said that “in some cases it is difficult to distinguish the procedural from the substantive”.  This was taken from Binnie J’s judgment in Mount Sinai at [35] (https://scc-csc.lexum.com/scc-csc/scc-csc/en/item/1879/index.do). But in that case, Binnie J went on to say at [36]-[37] that there can still be a “principled basis” for distinguishing between the two; and at [38] that substantive relief is not available in such cases under Canadian law. In other words, the point was not that it’s difficult to distinguish between procedural and substantive legitimate expectations so we should give up trying to distinguish, but rather that it’s difficult to distinguish so we should be careful to distinguish in a principled manner.
        3. Finally, when laying down the test at [119], the court said: “… and substantive relief should be granted under the doctrine subject to certain safeguards…” — so the test was meant to apply only to substantive legitimate expectations.

      4. The phrase “improper purposes”

        A student identified a possible ambiguity in the phrase “improper purposes”: “purposes” can mean (a) the purpose that the power-holder pursued; (b) the purpose served by the exercise of the power. My thoughts:


        The cases (or, more often, textbooks describing those cases) generally appear to use those words (and their variants) to refer to one’s aims, and not the outcome. We see this in classic English cases like Wheeler v Leicester ([1985] 1 AC 1054 at 1081C), Padfield ([1968] 1 AC 997 at 1032E), and Shell ([1988] 1 All ER 938 at 952d). These cases do not always use the phrase “improper purposes”, but that phrase is used by leading textbooks to describe them: see, e.g., De Smith’s Judicial Review at [5-113] and [5-117], which discusses Padfield.

        The phrase “improper purposes”, as conventionally used, could refer not only to acting for improper motives (e.g. avoiding political embarrassment, as in Padfield), but also acting for a purpose for which one doesn’t have the legal power to act. Consider the M M Pillay case, in which the argument was that the subsidiary legislation on restricted zones was ultra vires the Act in that the Act “did not confer any power on the Minister to make rules to prohibit, except on payment of a fee, motor vehicles from using any road” [4]. 

        You are right to ask whether the decision itself goes some way toward achieving the purpose set out in the statute. To my mind, this could mean one of two things:

        1. A case in which there is no legal power to make a decision (/take some action) that is aimed at achieving a certain purpose — this has been discussed above.
        2. Irrationality, in that, though the decision falls within the range of what the decision-maker is legally empowered to do, it does not fall within the range of what a reasonable decision-maker would do. The argument would be that no reasonable decision-maker would make a decision that goes no way at all toward achieving certain ends, or that no reasonable decision-maker would make a decision that goes no way at all toward achieving its ostensible ends.

        While the English phrase “improper purposes” could refer either to (a) improper motives; (b) acting for a purpose not authorised by statute; or (c) a lack of effectiveness in achieving certain ends, the phrase is generally used in administrative law to refer to (a) and (b) only. 

        My view is that it makes sense to distinguish conceptually between these three, not least because the justifications for each of them may be different. 

        I would use the phrase “improper purposes” to refer to (a) and (b), if only for the sake of convention. As for (c), it is one type of irrationality, but it is not the only one. A lack of logical connection between means and ends is not the only thing that could make a decision irrational.

        I do, however, see the problem you allude to with the World Development Movement case. That case is often criticised for drawing a distinction between “sound” and un-“sound” developmental purposes, when the statute (arguably) made no such distinction. Perhaps a more straightforward way to make sense of the case is simply to say that the decision was irrational.

        Finally: If there is evidence suggesting that a decision was irrational, it may be that only a statement of the decision-makers’ reasons can rebut that evidence. In the absence of such a statement of reasons, the court may conclude that the decision was indeed irrational. See Craig at [12-041] and De Smith at [7-111]ff, and the cases cited therein (especially Lonrho).

      5. Staying the execution of a sentence of death on the ground that the law has changed

        Some of you asked: Suppose someone has been sentenced to death on the basis that the law was A. The courts have since changed the law, making it now B. What can that person do? (For the avoidance of doubt, the following does not constitute legal advice.)

        The starting point is that, at common law, while the Court of Appeal has the power to re-open a concluded criminal appeal (Kho Jabing v PP [2016] 3 SLR 135 (CA)), it will not do so just because the law has changed (R v Jogee [2016] 2 WLR 681 (UKSC) at [100], cited in Kho Jabing at [58], and explained by Gobi a/l Avedian v PP [2020] SGCA 102 at [26]). After all, whatever arguments prompted the court to change the law in B’s case could just as well have been made in A’s case (Kho Jabing at [58]).

        Now, an application to the court to re-open a concluded criminal appeal is called a “review application”: see sections 394F-394K of the CPC, which were applied in Gobi. For present purposes, the important provisions are these:

        • “The applicant in a review application must satisfy the appellate court that there is sufficient material (being evidence or legal arguments) on which the appellate court may conclude that there has been a miscarriage of justice in the criminal matter in respect of which the earlier decision was made”: s 394J(2).
        • Here, “legal arguments” are “sufficient” if they
          • had never been canvassed during the proceedings (s 394J(3)(a))
          • “could not have been adduced in court earlier” “even with reasonable diligence” (s 394J(3)(b))
          • are “compelling, in that [they are] reliable, substantial, powerfully probative, and capable of showing almost conclusively that there has been a miscarriage of justice in the criminal matter in respect of which the earlier decision was made” (s 394J(3)(c))
          • are “based on a change in the law that arose from any decision made by a court after the conclusion of all proceedings relating to the criminal matter in respect of which the earlier decision was made”: s 394J(4).
        • The court may only conclude that there has been a “miscarriage of justice in the criminal matter in respect of which the earlier decision was made” if that decision was “demonstrably wrong” or “tainted by fraud or a breach of the rules of natural justice, such that the integrity of the judicial process is compromised”: s 394J(5).
        • “[D]emonstrably wrong”, for present purposes, means that there must be a “powerful probability”, and not just a “real possibility”, “that the earlier decision is wrong”: s 394J(6).

        A person who wishes to rely on such an argument should immediately apply for a stay of the execution of the death sentence: see generally Syed Suhail bin Syed Zin v AG [2020] SGCA 122 at [80]-[84].

      6. Freedom of religion and the “sovereignty, integrity and unity” of the state

        I would like to point out the following extracts from Bijoe Emannuel v State of Kerala AIR 1986 Ker 32 <https://indiankanoon.org/doc/1657841/> (in which the High Court of Kerala dismissed a challenge by Jehovah’s Witnesses to a rule requiring students to sing the national anthem), which bear some similarity to certain passages from Colin Chan v PP:

        “By the Constitution (42nd Amendment) Act, 1976 the Preamble to the Constitution was amended and the words “socialist secular” have been inserted and the word “integrity” has been incorporated after the word ‘unity’. By the same Amendment Act, Part IV-A containing Article 51A was newly added under the caption ‘Fundamental Duties’.

        These provisions would show that the Parliament was more concerned about the unity and integrity of India. By the introduction of the aforesaid changes in the Constitution I am of the view, that the Parliament has reasserted that it is the fundamental duty of every citizen to strive hard to uphold and protect the sovereignty, unity and integrity of our motherland. All such rights, fundamental or otherwise must therefore subserve the sovereignty, integrity and unity of this great nation.” (emphasis added)

        and, later:

        “As observed by the Supreme Court in Swamiar’s case (AIR 1954 SC 282)… [f]reedom of religion guaranteed under our Constitution… is not absolute to be exercised independently of the Fundamental Duty prescribed by the same organic instrument. In short, the religious beliefs, must reconcile with the right of a State to employ the sovereign power to ensure orderly living ‘without which constitutional guarantees of civil liberties would be a mockery’… The conduct of the members of Jehovah’s Witnesses in not showing unqualified respect to Constitution, the National Anthem etc. would endanger the security of the Nation in that the same will develop, among the citizens a tendency to ignore the mandates of the Constitution thereby defeating the object discernible from the Preamble and other provisions of the Constitution, to accomplish which, the people gave the Constitution to themselves. It therefore follows that if religious practices based on religious faith and belief protected by the State, run counter to public order, morality, health or a policy of the Government to uphold the sovereignty, integrity and unity of the nation or social welfare embarked upon, then the said religious practice must give way for the benefit of the people and the nation as a whole.” (emphasis added)

        Note that Article 51A(c) of the Constitution of India has stated, as of 1976: “It shall be the duty of every citizen of India — to uphold and protect the sovereignty, unity and integrity of India”.

        And Article 25(1) states: “Subject to public order, morality and health and to the other provisions of this Part, all persons are equally entitled to freedom of conscience and the right freely to profess, practise and propagate religion.” (“[T]his Part” refers to Part III of the Constitution, which pertains to “fundamental rights”.)

        Please note that the decision of the High Court of Kerala was overruled by the Supreme Court of India: AIR 1987 SC 748 <https://indiankanoon.org/doc/1508089/>.

      7. Freedom to practise religion: the “essential practices” test

        The history of the ‘essential practices’ test has been usefully set out by Gautam Bhatia in this blog post: https://indconlawphil.wordpress.com/2015/08/19/essential-religious-practices-and-the-rajasthan-high-courts-santhara-judgment-tracking-the-history-of-a-phrase/. (He builds on these ideas in his article in Global Constitutionalism: https://search-proquest-com.libproxy.smu.edu.sg/docview/1835779220/fulltextPDF/10B42EC62EE9428FPQ/1?accountid=28662) 

        Bhatia traces the “essential practices” test to the following statement by B R Ambedkar in India’s Constituent Assembly in 1948: 

        “The religious conceptions in this country are so vast that they cover every aspect of life, from birth to death. There is nothing which is not religion and if personal law is to be saved, I am sure about it that in social matters we will come to a standstill. I do not think it is possible to accept a position of that sort. There is nothing extraordinary in saying that we ought to strive hereafter to limit the definition of religion in such a manner that we shall not extend beyond beliefs and such rituals as may be connected with ceremonials which are essentially religious. It is not necessary that the sort of laws, for instance, laws relating to tenancy or laws relating to succession, should be governed by religion.”

        Bhatia continues:

        “Ambedkar’s use of the term ‘essentially religious’, therefore, was in response to a very specific concern. He was worried that unlike in the West, with its seemingly clear demarcation between the City of God and the City of Man, there was no aspect of Indian life which was untouched by religion. Consequently, insofar as the Constitution protected religion and personal laws, there was a very real risk that it would entirely hamstring the State’s power to pass social legislation. He was, therefore, adamant that there must be a separation between religious activities, and secular activities tinged with religion. The latter could have no constitutional immunity from legislation. In Ambedkar’s formulation, it is clear the word ‘essentially’ qualified ‘religious’, and was designed to separate the religious from the secular.”

        I could go on quoting from the blog post, but I may as well ask those of you who are interested to read it in its entirety.

      8. The history of the right to propagate one’s religion

        The history of the right to religious propagation would appear to be as follows.

        It is rooted in the idea raised during India’s Constituent Assembly’s debates that “certain religions like Christianity and Islam were proselytizing religions and that they should be permitted to propagate their faith” (see B Shiva Rao et al., The Framing of India’s Constitution: A Study (1968) at p 261).

        Concerns about this were raised, such as that “religious propagation had been responsible for the unfortunate division of the country into India and Pakistan” and “religion was a private affair between oneself and one’s creator and it had nothing to do with others and therefore the right to propagate religion was wholly unnecessary”.

        Nonetheless, the word “propagate” was retained, apparently since it would not have any “dangerous implications” as the state would remain secular, the right to propagation would be available to adherents of all religions, and there were still limitations such as public order (ibid at p 267).

      9. Feedback on students’ essays on the Singapore Presidency

        The question asked whether “recent developments in the political arena in Singapore” have rendered the Presidency “now irrelevant”. It was insufficient only to discuss the “recent developments in the political arena”, or to argue that the Presidency is “irrelevant”, without discussing whether there is a link between the two.  

        Given the reference in the question to “a different era in Singapore’s constitutional and political history”, it would have been useful to discuss the particular “recent developments in the political arena” that have marked a change from the past “different era”.

        It was necessary to distinguish between arguments that the (popularly) elected Presidency was “irrelevant”, and arguments that the Presidency itself was “irrelevant”. Strong answers discussed the potential “irrelevan[ce]” of the President’s functions separately from the potential “irrelevan[ce]” of the fact that the President is popular elected.

        Students were welcome to examine the connotations of “relic” and “era”, and hence challenge the assumption that the time at which the elected Presidency was so long ago – and so alien to the present – that it could be said to fall within a different “era” from the present. Students were also welcome to argue that the statement that the elected Presidency is “now irrelevant” implies that it was once relevant, and to challenge the truth of this implication.

        In weaker essays, the only “recent developments” discussed were developments relating to the Presidency itself. These essays sometimes stopped at asserting that, given these developments, the Presidency cannot be said to be “irrelevant”. The weakness in this approach was this: if the changes prevent the Presidency from becoming “irrelevant”, then there must have been something that would, but for the changes, have made the Presidency “irrelevant”, and so an account of the changes would have been incomplete without an account of this something.

        Certain weaker essays criticised the institution of the elected Presidency but fell short of showing that these criticisms rendered the elected Presidency irrelevant. 

        Some essays asserted that, because certain risks against which the Presidency is meant to guard have not recently eventuated, the Presidency is irrelevant. However, one may ask: is the point of checks and balances not to guard against the possibility of risks eventuating?

        It is not true that the President is the only check against the risk of the legislature frittering away public funds. At the same time, it is not true that the Presidency must be “irrelevant” because it is possible for people to vote out a spendthrift government: is Singapore to tolerate profligacy for up to five years?

      10. The concept of the burden of proof

        I feel I should say something about the meaning of “burden of proof”, “legal burden”, and “evidential burden”. Let us take as our starting point Britestone Pte Ltd v Smith & Associates Far East, Ltd [2007] 4 SLR(R) 855 (CA) at [60]:

        “… at the start of the plaintiff’s case, the legal burden of proving the existence of any relevant fact that the plaintiff must prove and the evidential burden of adducing some (not inherently incredible) evidence of the existence of such fact coincide. Upon adduction of that evidence, the evidential burden shifts to the defendant, as the case may be, to adduce some evidence in rebuttal. If no evidence in rebuttal is adduced, the court may conclude from the evidence of the plaintiff that the legal burden is also discharged and making a finding on the fact against the defendant. If, on the other hand, evidence in rebuttal is adduced, the evidential burden shifts back to the plaintiff. If, ultimately, the evidential burden comes to rest on the defendant, the legal burden of proof of that relevant fact would have been discharged by the plaintiff.”

        Now, let me try to explain this in simpler terms.

        The “legal burden of proof” — or, for short, the “burden of proof” — means “the obligation to persuade the trier of fact that, in view of the evidence, the fact in dispute exists”: Britestone at [58]. For example, if I sue you for breach of contract, I bear the legal burden of proving that the contract exists; that you breached it; and that your breach caused me loss. If I fail to prove all of these facts, then my claim must fail, for I have failed to discharge my burden of proof. Similarly, if the Public Prosecutor prosecutes me for theft under s 378 of the Penal Code, it bears the legal burden of proving that I moved property which was in the possession of someone else; that I moved it with the intention of taking it dishonestly out of that person’s possession; and that I moved it in order to effect the taking. If the Public Prosecutor fails to prove all of those facts, then he has failed to discharge his burden of proof, I must be acquitted.

        When I bear the “evidential burden of proof” (also known as the “tactical burden”), this means that my “failure to adduce some evidence, whether in propounding or rebutting, will mean a failure to engage the question of the existence of a particular fact or to keep this question alive” (Britestone at [58]). Suppose that my opponent bears the legal burden of proof. Suppose further that my opponent has adduced evidence which would lead the court to conclude in his favour. In that case, if I fail to rebut that evidence, then my opponent must succeed. In that case, one would say that, in the light of my opponent’s evidence, I bear the “evidential burden” of rebutting that evidence.

        Let me illustrate this with an example. Suppose I am being prosecuted for voluntarily causing hurt, which is an offence under s 321 of the Penal Code. The Prosecution bears the legal burden of proving that I had the “intention of… causing hurt to any person”. Suppose the Prosecution adduces evidence that I make eye contact with someone, shout loudly, and then punch that person. Does this evidence, in and of itself, prove conclusively that I intended to cause hurt? Maybe; maybe not. But it does not matter. The point is that, as a matter of common sense, a person who has seen such evidence, is likely to think that I had the intention to cause hurt. In that case, the evidential burden is on me to rebut that evidence. If I do not, then the judge — like any reasonable person with common sense — would naturally conclude that I did have the intention to cause hurt.

        It is in this light that we must understand passages such as the following extract from the CA’s judgment in Ridzuan (at [51]):

        “an applicant who alleges that the PP’s decision declining to grant him a certificate of substantive assistance was made in breach of Art 12 of the Constitution would satisfy the evidentiary burden he bears if he can show two things – first, that his level of involvement in the offence and the consequent knowledge he acquired of the drug syndicate he was dealing with was practically identical to a co-offender’s level of involvement and the knowledge the co-offender could have acquired, and second, and more importantly, that he and his co-offender had provided practically the same information to CNB – yet only his co-offender had been given the certificate of substantive assistance. This follows from our holding at [40]–[43] above that the applicant does not have to produce evidence directly impugning the process by which the PP reached his decision; instead, he can discharge the evidentiary burden he bears by highlighting circumstances that raise a prima facie case of reasonable suspicion of breach of the relevant standard. The situation we described earlier in this paragraph, if it were to occur, will raise questions as to why only one co-offender and not the other was granted the certificate of substantive assistance.”

        This does not mean that the application for judicial review will succeed if the applicant shows one of the “two things”. All it means is that, if the applicant shows one of the “two things”, he thereby “raise[s] questions” such that, if the applicant’s arguments are not rebutted, then the applicant will succeed. In other words, the evidential burden of proof shifts to the respondent. However, it is still open to the respondent to rebut the applicant’s arguments; if the respondent successfully rebuts them, then the evidential burden shifts back to the applicant, in that, if the applicant does not rebut the respondent’s rebuttal, then the respondent will succeed (i.e. the applicant’s application for judicial review will fail). In short, even if A has raised enough evidence to shift the evidentiary burden to B, that does not mean that A will ultimately be victorious, because it is still open to B to rebut A’s evidence, and A to rebut B’s rebuttal, and B to rebut A’s rebuttal of B’s rebuttal, etc.

      11. Where do the legal limits to executive powers come from? Do they all come from legislation?

        A student asked about the following line from Short v Poole: “… the Court might well say, ‘Parliament never intended to give authority to make such rules [of subsidiary legislation]; they are unreasonable and ultra vires’.” This suggests that the reason why irrational executive action is unlawful is that the Legislature decided to make them unlawful. Is that really so?

        My response is as follows. By “[R]”, I was referring to the rule that irrational executive action is unlawful.


        The fiction that judicial review is all about enforcing legal limits which the Legislature had (impliedly) intended to put in place is arguably just that — a fiction that is put in place to uphold the doctrine of Parliamentary sovereignty in the UK. By resorting to this fiction, the courts can avoid being accused of subverting the Legislature’s intention; they claim to be upholding the Legislature’s intention. This is why, in cases like Short v Poole, what we now call judicial review for irrationality was said to be a matter of enforcing an (implied) legislative intention that executive power not be exercised irrationally. Similar, in Edwards v Bairstow [1956] AC 14 (UKHL) at 36, Lord Radcliffe said that, if there is a decision which “no person acting judicially and properly instructed as to the relevant law could have come to”, then the court will infer that that decision was the product of some “misconception of the law”. 

        But since then, English judicial thinking has shifted away from this fiction, at least when it comes to irrationality. In GCHQ at 410H-411A, Lord Diplock said about irrationality review: “resort I think is today no longer needed to Viscount Radcliffe’s ingenious explanation in Edwards v Bairstow… of irrationality as a ground for a court’s reversal of a decision by ascribing it to an inferred though unidentifiable mistake of law by the decision-maker. ‘Irrationality’ by now can stand upon its own feet as an accepted ground on which a decision may be attacked by judicial review.”

        This leads one to wonder whether it can really be said that irrationality review is simply an application of rules created (tacitly) by the Legislature, and gives rise to what, in the English literature, are known as the ‘modified ultra vires’ and the ‘common-law’ theories.

        All these are theories of where the legal limits to powers come from. In my article, I did not need to say where [R] came from; it could either come from the intention of the Legislature, or just from the common law. In the latter case, I suppose the question would be: what gives the common law the power to create such rules? 

        One possible answer is: the common law has the power to create rules by which the executive must conduct itself, for the same reason that the common law has the power to create rules by which (say) parties to contracts, drivers, and employers must conduct themselves. The common law is a source of law, so why can’t it contain laws about executive conduct? One may retort: it can’t because that would clash with the statute. But is there really a clash? Suppose a statute doesn’t explicitly say that some executive body must give a certain person a fair hearing before deciding to take his licence away. A person who subscribes to the ‘common law’ theory of judicial review would say that the Legislature has had no intention as to whether a hearing must be held; and, therefore, when the court adds the requirement that a hearing must be held, the court is merely supplementing the statute rather than contradicting it. So the answer to the question would simply be: [R] is a rule of the common law.

        A real problem for a common-law theorist would be a statute that, for example, explicitly says that X must not be heard before (say) his driving licence is taken away. If the common law introduces a rule saying that X nonetheless must be heard, then the common law would be contradicting the statute, which is forbidden. Sometimes, the courts would be available to wriggle out of this problem by creatively ‘interpreting’ the statute and saying that, on its proper interpretation, it does not really mean that X must not be heard — but, of course, the courts can only go so far in ‘stretching’ statutory language before they end up disobeying it altogether.

        But — and this is Prof Kenny Chng’s point (in this article) — we have a way out: it might be the Constitution that furnishes the “justification of the entire practice of judicial review” (at 312; emphasis omitted). While this might not change the fact that the “legal source of the rules applied in judicial review” is “the common law” (311), it would help to answer the questions of (a) why the common law is justified in developing the grounds of judicial review — in other words, the norms that the executive must follow — as they now are, and (b) why these norms may, in certain cases, override legislative intention.


        The next question: “where exactly in the Constitution are these norms located?” (312) First, Chng suggests: “Article 93 of the Singapore Constitution can be interpreted as a warrant for the judiciary to draw upon rule of law principles in their exercise of judicial review.” (312) Second, he suggests: “rule of law ideals [are] implicit in the concept of constitutionalism itself”. To my mind, to make these points, he must be suggesting that not acting irrationally is one of the “rule of law principles” —  a proposition that is debatable depending on one’s view of how the concept of the “rule of law” is to be interpreted”. He admits this (at 313), but responds (at 313), citing Luc Tremblay’s work (at fn 123), that there are “antecedent rule of law norms” which are not found in the Constitution itself but which are “required to supply legitimacy for the written constitution”.

        To these, I might add a third possibility, which I base on dicta in Chng Suan Tze at [82]: that Article 12(1) of the Constitution requires that the exercise of executive power not be “arbitrary”. Of course, the missing logical link is to say that irrationality is a type of “arbitrar[iness]”; support for this latter proposition may be drawn from Timothy Endicott, ‘Arbitrariness’ (University of Oxford Legal Research Paper Series No 2/2014) <https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2378858> at 18:

        “An arbitrary decision in general is one that is not distinguished, by reasons in favour of it, from an unreasoned choice.”

        See also Jeyaretnam Joshua Benjamin v PP [1989] 2 SLR(R) 419 (HC): 

        “If the discretion vested in the licensing officer under the Act were absolute and untrammelled, there might be a case for arguing that the licensing scheme is unconstitutional, as then the licensing officer would have been vested with arbitrary power to deprive a citizen of his constitutional right to freedom of speech and expression contrary to the objects of Art 14(2)…”

        So the answer to the question might well be: ‘some concept of non-arbitrariness implicit in the Constitution’. Of course, this answer is not perfect — for example, it leaves open the question of how this general concept of non-arbitrariness translates into precise legal rules (is acting disproportionately arbitrary?).

        A thinner answer might be that the definition of “judicial power” in Art 93 includes the power to develop English-style common law, which would simply mean that judicial review is justified by (a) the common-law model; coupled with (b) Art 93, which allows the courts to create and develop rules of the common law. 

        I could go on and on, but I will end with this: as we will explore in class, it is not necessarily the case that the various theories of why judicial review is possible are mutually exclusive. For example, it might be that the ultra vires model explains some grounds of judicial review; the common law model explains others; and the Art 93 model explains yet other rules of administrative law (e.g. furnishing the basis for courts ignoring ouster clauses).

      12. The separation of powers, the Liyanage case, and the Mohammad Faizal bin Sabtu case

        The separation of powers

        In class, I described the separation of powers as being a “value” or a “virtue”. On hindsight, I should have used clearer and less vague language. The point I was trying to make was that the separation of powers is not something that is either absolutely present or absent. Rather, powers can be separated in various ways, to various extents, and for various purposes. Hence, there is (for example) nothing intrinsically wrong with having members of the executive also be members of the legislature; with having both the legislature and the judiciary act as checks on executive power; or with a situation in which the legislature, executive, and judiciary all make law. Rather, I would suggest that we ought to (a) think of the aims of separating powers; and then (b) think about how far the way that powers are separated in our legal system succeeds in achieving these aims. What we should not do is attempt to answer questions like “Does Singapore have a separation of powers?”, because that pre-supposes that there is some ideal archetypal manner of separation of powers, when in truth there is not.

        The other point I wished to make by using the word “value” — but I think I will not have made it clearly until we begin Topic 2 — is that the separation of powers is not only a description of how state institutions are designed, but also something which state institutions bear (and must bear) in mind when they go about their duties. We will see this most clearly when we finish our introduction to judicial review, which will illustrate how the judiciary takes stepps to avoid itself violating the separation of powers.

        There is a useful book chapter on this for those who would like a more detailed (but still manageable) account of the separation of powers: Aileen Kavanagh, ‘The Constitutional Separation of Powers‘, ch 11 in David Dyzenhaus and Malcolm Thorburn (eds), Philosophical Foundations of Constitutional Law (2015: OUP). In short, Dr Kavanagh makes the following points:

        1. It is not true, that one power-holder must correspond perfectly to exactly one function of the state. For example, the courts do not only apply the law, but also make it.
        2. It is not true that each branch of the state is to “confine itself” to one function. “In order to carry out their respective roles, each branch must take account of the role and responsibilities of other branches.” (228)
        3. Rather, the main point behind the separation of powers is “to allocate power and assign tasks to those bodies best suited to carry them out” (230) and to “put mechanisms in place to correct for potential abuse of power and jurisdictional overreach” (230). For example, the courts are the ones who resolve legal disputes because the courts have the features that best suit them to doing so, such as  independence (231).
        4. Both of these tasks are “underpinned by the deeper value of coordinated institutional effort between branches of government in the service of good government” (235). Sometimes, the branches of the state are inter-dependent (235); and they have to show “inter-institutional comity” (235-236).
        5. We must consider that there are different varieties of (for example) law-making power. For example, while both the courts and the legislature make law, they do so in different ways. John Gardner calls them “legislative powers of law-making” and “judicial power of law-making” (quoted at 232).
        6. Instead of striving for a “pure” separation of powers, our account of separation of powers ought to take into account the “necessary and desirable interconnections, interdependence, and interactions between the branches of government” (238): “lawmaking, law-applying and law-executing are collaborative tasks where each organ of government must cooperate with the other organs in an interactive setting” (239). (But Kavanagh does not say that this is to come at the expense of the need for specialism and mutual checks and balances, as described in point 3 above.)

        Does Faizal contradict Liyanage?

        One of you asked the following question: Faizal says that the judicial power includes to pass a sentence on a particular offender. But that is what the legislation in question in Liyanage did. Is there a contradiction?

        My answer is that there is no contradiction.

         The problem with the legislation in question in Liyanage was that it prescribed a punishment to be imposed on specific individuals. By contrast, in Faizal, “[t]he enhanced punishments under s 33A(1) appply generally to all offenders who fulfil the prescribed conditions” (Faizal at [47]). Note that the legislation in question in Faizal did not apply to every drug offender; it only applied to some. This underscores what the Privy Council said in Liyanage: that not every law which is “ad hominem” and/or “ex post facto” must “inevitably usurp or infringe the judicial power” (289G). The Privy Council identified various factors that made the legislation an unconstitutional attempt to usurp judicial power:

        1. The legislation was “clearly aimed at particular known individuals”, as opposed to the “generality of the citizens” (289F).
        2. The motive behind the legislation was to “secure the conviction and enhance the punishment of those particular individuals” (290D), by making it much easier for them to be convicted; in substance, therefore, it was the Legislature which was attempting to convict them, whereas in truth only the courts have the power to convict.
        3. The legislation aimed to “ensure that the judges… were deprived of their normal discretion as respects appropriate sentences” (290G).

        It is this third point that, at first glance, clashes with Faizal, because Faizal says it is possible that judges have no sentencing discretion at all. But there is, in fact, no clash, as we see when we focus on the word “normal”. The problem in Liyanage was that, as a result of the legislation, the courts would have a narrower range of sentences to choose from for the alleged participants in the coup than for everyone else. Faizal itself affirms, at [45], that such legislation would be unconstitutional in Singapore as well: “no written law of general application prescribing any kind of punishment for an offence, whether such punishment be mandatory or discretionary and whether it be fixed or within a prescribed range, can trespass onto the judicial power” (emphasis added). 

        I note in passing that the first and second points, taken together, make it unclear whether the crux of the problem was that the effect of the legislation was to target selected individuals, or that the motive of the legislation was to do so.  Nonetheless, it would appear that the High Court in Faizal would have agreed with both points. To my mind, we can see this from the High Court’s treatment of the Australian case of Totani. At [57], the High Court remarked that one reason why section 33A of the Misuse of Drugs Act was not unconstitutional was that it “does not involve the imposition of a sentence absent a judicial finding of guilt” and was “not directed at achieving a particular sentencing outcome”. 

        I would also draw to your attention the following points made by the Supreme Court of Ceylon in Liyanage <https://www.lawnet.gov.lk/1977/12/31/the-queen-v-d-j-f-d-liyanage-and-others/>, with whose conclusion on the issue of ministers’ power to select judges the Privy Council agreed:

        1. The Supreme Court refused to allow a situation in which the executive could, by reserving the power to select the judges who would hear a particular case, “prevent a particular judge or judges from exercising any part of the strictly judicial power vested in them by the Constitution as judges of the Supreme Court” (358). This seems to suggest those who exercise judicial power must at least have the opportunity to decide cases. To my mind, this is a relatively weak point, compared to the next point:
        2. If the executive had the power to nominate judges to hear particular cases, then there would be a suspicion that the court is not independent of the executive (360). 

        By the way, for an example of legislation that was even more blatant than that in Liyanage in usurping judicial power, see Kable v Director of Public Prosecutions (1996) 189 CLR 51 (High Court of Australia) <http://www7.austlii.edu.au/cgi-bin/viewdoc/au/cases/cth/HCA/1996/24.html>, extracts of which are as follows:

        “(3) This Act authorises the making of a detention order against Gregory Wayne Kable and does not authorise the making of a detention order against any other person.

        (4)  For the purposes of this section, Gregory Wayne Kable is the person of that name who was convicted in New South Wales on 1 August 1990 of the manslaughter of his wife, Hilary Kable.”

        Liyanage and Article 12(1)

        Finally, a couple of you asked whether the legislation in Liyanage, if enacted in Singapore, would have fallen foul of Article 12(1) of the Constitution (on equality). I think the answer is that it would fall foul of Article 12(1) for the same reasons why it would fall foul of Article 93. In short, to pass muster under Article 12(1), legislation must be intelligible and logically linked to an aim which is not one that no reasonable person would consider to be a legitimate aim of legislation. The legislation in Liyanage was intelligible, and logically linked to the aim of punishing the persons involved in the coup. The reason why the legislation would fall foul of Article 12(1) is that no reasonable person would consider singling them out for especial punishment to be a legitimate aim of legislation; but the reason why this is so is simply that to do so would be a legislative usurpation of judicial power.

        A follow-up on Liyanage

        One of you asked: Suppose the legislation in Liyanage applied retrospectively, but generally — in other words, suppose it had been retrospectively deemed to come into effect on 27 January (or earlier), but would not be limited in application to only alleged coup-related activities which took place on or around 27 January. Would that legislation still be unconstitutional? In other words: what is the crux of the Privy Council’s reasoning: was it that the legislation was retrospective, or that it was not general in application?

        I can’t say for sure, because the court did not make clear exactly what about the legislation made it unconstitutional. Nonetheless, my guess is that the court might still have held such legislation unconstitutional. This is because there are parts of the judgment that suggest that what made the legislation unconstitutional was not only the effect of the legislation, but also the legislature’s motive (289E: “aimed”; 290A: “true purpose”; 290F: “it was not only the likely but the intended effect… true nature and purpose”). And the Government’s White Paper made this intention clear (278A-C). So the court might have held that, even if the legislation were ostensibly general, the legislature’s motive would still render the legislation an unconstitutional usurpation of judicial power.

        In addition, note that that sort of legislation, if enacted in Singapore, might fall foul of Article 11(1) of our Constitution, which provides: “No person shall be punished for an act or omission which was not punishable by law when it was done or made, and no person shall suffer greater punishment for an offence than was prescribed by law at the time it was committed.” This might well be the most straightforward argument why the legislation is unconstitutional. (I’m not sure whether the Ceylonese Constitution had such a provision.)

      13. Parliament and the (Parliamentary) executive: some basics

        A first-year student asked me about what it means to say the executive “enforces” the law; what precisely there is to be said about legislation in Parliament after that legislation has been passed; and the implications of the fact that ministers are also MPs. Here is my response (in the form of an audio recording): https://mediacast.smu.edu.sg/media/Response%20to%20questions%20on%20executive%20and%20legislature/1_pubz410p

      14. A decision of the Indonesian Constitutional Court on candidacy in elections

        A student asked about this development in Indonesian constitutional law. I wondered about the precise legal basis of the Indonesian Constitutional Court’s decision. I soon discovered that it wasn’t perfectly clear to me what precisely the court had decided. My response:


        This question intrigued me because I’ve never read anything about Indonesian law before. My immediate thoughts were:

        1. What is the precise basis for the Constitutional Court’s decision? What aspect of the electoral laws was unconstitutional, and precisely what part of the Constitution did it violate?
        2. If the Constitutional Court makes a decision on the constitutionality of a law, can the Legislature insist on reversing that decision? In some countries, the answer to a similar question is yes.  For example, under section 33 of the Canadian Charter of Rights and Freedoms, the Legislature can insist that a piece of legislation is to have force even if it violates certain rights in the Charter. (Section 33 is sometimes called the “notwithstanding” clause, because it allows a law to have force notwithstanding the fact that it is incompatible with certain rights). So if a Canadian court declares legislation as void because it violates one of those rights, the legislature can re-enact it anyway and give it force by invoking the “notwithstanding clause”.
        3. If the Legislature has no legal power to reverse the Constitutional Court’s decision, is there some reason why it can get away with doing so anyway?

        So I started by looking at the law in question and the Constitutional Court’s decision. I don’t understand any Bahasa Indonesia (and even if I understood Bahasa Melayu — which I don’t — I wouldn’t rely on that when precision is required). So I downloaded a copy of the decision and used Google Translate to translate it into English. (All quotations below are from that machine-generated translation; the usual caveats apply, of course.) As I understand it, the main part of the decision starts at paragraph 3.8; previous paragraphs only summarise the history of the litigation, the parties’ arguments, and other miscellaneous matters.

        I was unable to locate a transcript of legislative debates on the proposal to enact legislation to reverse the Constitutional Court’s decision. (Someone please let me know if you know where I may find one!)

        So, in the rest of this note, I will consider only the Constitutional Court’s judgment. I found it worth doing this because I suggest that it is a useful example of the kind of problem of constitutional interpretation that we often see in constitutional democracies. We might well return to this example later in the course when we study the concept of a Constitution; but even if we don’t, we will certainly consider similar examples.

        Relevant legislation

        I saw that the debate centred around Articles 40 and 41 of PERPU 1 of 2014, as amended by UU 10 of 2016.

        I got the sense the that the relevant laws were Articles 40 and 41 of PERPU 1 of 2014 (as amended). These, as I understand them (again from a Google translation), set out two ways in which a person may become a candidate in an election for governor, regent, or mayor.

        First, under Article 40, a political party or coalition of parties can register candidates if the party/coalition of parties:

        1. controls at least 20% of the seats in the province’s Regional House of Representatives (DPRD); or
        2. gained at least 25% of the votes in the last DPRD election and controls at least one seat in the DPRD.

        Second, under Article 41, an individual can register as a candidate if the individual has “support” from a certain percentage of the population of the province. This percentage ranges from 6.5% to 10%, depending on the population of the province. (I understand from this news story, linked to on this Wikipedia article, that candidates show “support” by getting people to sign support forms accompanied by photocopies of their identity cards.)

        The applicants’ arguments

        As I understand it, the applicants’ main argument was this: Under Article 40, why must the party have gained any seats in the DPRD at all to have candidates run for Governor/Vice Governor? Shouldn’t it be enough that they have gained a certain number of votes? After all, under Article 41, an individual candidate can run for election as Governor/Vice Governor as long as they have the support of a certain percentage of the population of the province – which is easier to get than to qualify under Article 40.

        The applicants argued that the law violated the constitutional principles of “popular sovereignty” and the “rule of law”, as set out in Article 1(2)-(3) of the Constitution, as well as the constitutional right to engage in “collective struggle” for one’s rights (Art 28C(3)) and equality before the law (Art 28D(1)).

        The majority of the court’s first point: the argument from democracy

        The majority of the court court said that the law in question “clearly limits the fulfillment of constitutional rights of political parties participating in the election who have obtained valid votes in the election even though they do not have seats in the DPRD, thus reducing the value of democratic regional head elections…” This, said the court, ran contrary to Art 18(4) of the Constitution, which said that the Governor “shall be elected democratically”, for two reasons. First, even if a party did not control any seats in the DPRD, the fact the people voted for it had to count for something; there was a need to “respect the people’s voice in the election”. Second, if only parties that controlled seats in the DPRD could have candidates run for Governor, then there would be the risk of having only “a single candidate”, which could “threaten the democratic process”.

        At first sight, I found this quite unconvincing. Yes, Article 40 limits the ability of political parties to participate in gubernatorial elections. But Article 41 also limits the ability of individuals to participate. And, for that matter, there are other laws limiting who can stand in elections: for example, there is a minimum education requirement. Surely the court didn’t mean to say that it’s wrong to have any limits on who can run for election at all?

        On closer inspection, the court did not say that. It did (as I understand it) affirm that the Legislature has the power to regulate such details, but subject to the minimum requirements set out in the Constitution.

        The key question is this: how are these minimum requirements – such as that the governor must be “elected democratically”, or that the principle of “popular sovereignty” is to be upheld – to be interpreted? The answer must depend on the interpreter’s ideas regarding democracy, what counts as a democratic system, etc. The answers to these questions may well be different in Indonesia in other systems. But even within Indonesia (or within any legal system), different people – and institutions – may have different answers. It is disagreements like these that explain the Indonesian legislature’s desire to reverse the Constitutional Court’s decision, as well as various other constitutional clashes.

        Justice M Guntur Hamzah’s disagreement with the court’s first point

        Indeed, there was a difference between different judges within the court. Justice M Guntur Hamzah, disagreeing with the majority, took the view that (as I understand it) while the Constitution states that the governor is to be “elected democratically”, it does not follow that there must be a certain model of democracy; the Legislature is free to choose one, as long as “the regional head [that] is elected is a representation of the people’s voice in the region”. This, he said, explains why it’s OK for the Governor of the Special Region of Yogyakarta to be chosen by a means other than direct election (I understand that it is an inherited position, similar to that of a hereditary monarch?) He added – and I see the force in this – that if the framers of the Indonesian Constitution intended for a particular method of electing the governor in a province, they would have spelled it out in the Constitution – as they had done in the case of the President and Vice-President (in Article 6A).

        I see the force in this view, but I do note that I, by virtue of my limited legal education and training, am inclined toward a certain view of democracy and a certain view of how Constitutions (and laws generally) are to be interpreted. This is the point where I will play the ‘I am not an expert in Indonesian law’ card and stop talking about this point.

        The court’s second point: the argument from rationality

        Speaking for myself, I find the majority’s second point to be more convincing, but only a bit more convincing. The majority said that the law was irrational, given that it was harder for a pair of candidates to run for Governor/Vice Governor as party members (under Article 40), compared to running as individuals (under Article 41). The court essentially made the two sets of the requirements the same. The court lowered the requirements for running for Governor/Vice Governor as members of a party, down to the level of the requirements for running as individuals.

        At first glance, this sounds compelling: why should it be harder for candidates to run as representatives of political parties, than in their own right? But if the two possibilities were perfect substitutes for each other, then, one may argue, the majority’s reasoning doesn’t make sense, because in practice no candidate would be disadvantaged – if their party didn’t get any seats in the DPRD, they could simply run as individuals.

        Moreover, if the two possibilities aren’t perfect substitutes for each other, then the majority’s appeal to equality may be unconvincing. There may well be several possible reasons why the requirements to run as a member of a party should be different from the requirements to run as an individual. I can’t articulate them precisely without studying Indonesian constitutional law in more detail (which I’m not in a position to do now).

        Justice M Guntur Hamzah’s last remark on the “motive and timing of the application”

        Justice M Guntur Hamzah’s last point was that the application to the Constitutional Court had been made “after the vote counting and results acquisition stage” – as I understand him, he said that the applicants made the application because the DPRD election results had just been released, and the applicants realised that because of those results, they wouldn’t qualify under Article 40 to nominate candidates for the Jakarta gubernatorial election. (Did I mention that the Constitutional Court case had been initiated by the Labour Party (Partai Buruh) and Gelora Party (Partai Gelombang Rakyat Indonesia / Indonesian People’s Wave Party)?)

        I have no comment whatsoever on either of these political parties, nor do I make any comment on their motives for bringing the case. Indeed, one may argue that whatever their motives were, they were entitled to attempt to persuade the Constitutional Court to correct what they thought was a violation of the Constitution. But I do note Justice Guntur Hamzah’s point that (as I understand it) electoral rules should not be changed, especially by the court, close to an election (for which he cited a US Supreme Court case called Purcell v Gonzalez). This links to a broader point: even if the Constitution demands that a law be declared void (or interpreted differently), what about the people who have planned their lives on the basis that the law was in force? Shouldn’t their expectations count for something?

        One may also make a comment about the general phenomenon – and, again, I make no comment on the two Indonesian political parties in question – of politicians engaging in constitutional litigation, not so much because they believe in a certain understanding of constitutional principles, but simply because they want some advantage for themselves (e.g. a better shot at being elected). Then again, one may ask: why would anyone go to court – or, indeed, lift a finger to do anything – except if there is something in it for them?

        Conclusion

        In theory, the courts have the authority to interpret the Constitution. From the point of view of democratic theory, this is justified by the fact that the Constitution is a product of democracy; the court is merely upholding the will of the People. Put that way, the court is not acting un-democratically; to the contrary, it is upholding democracy. But the Legislature which wants to act contrary to the Constitution claims to be doing the exact same thing: after all, the Legislature consists of people who are also democratic representatives. Who is right?

        It depends. One possible way forward is to ask: Which ‘People’ should take priority: the People who formed the 1945 Constitution, or the People today? Some argue that the very idea of allowing a court to strike down legislation made today based on a Constitution made long ago, means that the country is governed by the “dead hand of the past”.

        Then again, the People who formed the Constitution were not doing the same thing as the People today. The former were creating a system of democracy, within which today’s legislature is one player; the former were creating the grand rules of a system (like Roblox?) that was built to last, while today’s legislature just creates rules under that system (like individual Roblox games) that come and go, even when the Constitution remains. Moreover, the People who formed the Constitution (the argument goes) meant to lay down hard limits (like Roblox’s community standards), which are meant to safeguard certain interests no matter what governments (or Roblox game-makers) come and go.*

        So the really big question is: who should have the power to change the rules of the entire system itself?

        * I don’t know anything about Roblox other than what I heard about from some YouTube video I watched some time ago. I think it was this: https://www.youtube.com/watch?v=vTMF6xEiAaY  

      15. Judicial review for error of fact: is it simply review for failure to take into account a mandatory relevant consideration, namely, ‘the truth’?

        If I understand correctly, [the student who asked this question is] saying that ‘the truth’ is a mandatory relevant consideration, and so the idea of error of fact is simply an instance of the idea that a decision-maker must take into account mandatory relevant considerations. That is true insofar as we say that the relevant consideration is ‘the need to know the truth’.

        But the problem is that, often, the error of fact doesn’t consist merely of a failure to take into account ‘the truth’ in the abstract, but rather a failure to take into account a particular piece of evidence / to pursue a particular line of inquiry. As the court in E said at [62], often the statute will say nothing — neither explicitly nor implicitly — about that particular piece of evidence / line of inquiry.

        The assumption in E is that “[f]ailure to take account of a material consideration is only a ground for setting aside a decision, if the statute expressly or impliedly requires it to be taken into account” (at [62]). We know that statutes lay down general rules/principles, and are not addressed to individual cases. Hence, when the court in E used the word “consideration”, it must have used that word to refer to a variable or a factor phrased in general terms — say, ‘the potential conflict between the decision and the law of trademarks’ (Jessie Tan) or ‘the degree of public outcry about the offences in question’ (Venables) or ‘the identity of the person who is complaining to the Law Society’ (CBB — High Court held that it was irrelevant). So far, so good.

        By contrast, in certain other cases, the courts use the language of irrelevant considerations to describe what is in substance an error of fact. They do this by using the word “consideration” to refer to a particular piece of evidence — as in Maripan Ponnusamy (“err in failing to take into account Dr Teh’s evidence”: at [34]), Venables and possibly Fong Thin Choo (at [56]-[57]). What are we to make of such cases? The court in E would probably describe them as involving errors of fact. The reason why these cases have used the language of relevant considerations is that they used the word ‘consideration’ to mean something broader than, as E suggests, a factor which is “expressly or impliedly” mentioned in the statute.

        Of course, this latter set of cases can be rationalised as cases involving irrelevant considerations too. We can say that the alleged error in Maripan Ponnusamy was a failure to ‘take into account all relevant medical expert evidence’. We can say that the problem in Venables was that “public protests about the level of a tariff to be fixed in a particular case are legally irrelevant” (526G). We can say that the problem in Fong Thin Choo was that that the Director-General had failed to take into consideration ‘all relevant documents or testimonies of witnesses’.

        If we can, in this manner, say that cases involving errors of fact are just cases about irrelevant/mandatory relevant considerations, then what does E add? I can think of at least two answers.

        First, E is intellectually honest. It addresses the problem that the conventional notion of relevant/irrelevant considerations doesn’t tell us anything about the level of generality with which a consideration should be framed. Indeed, in E, the court wasn’t creating concepts out of thin air; what it was doing was trying to come up with a framework that satisfactorily explains many other past English cases which in substance involved review for error of fact but which struggled to explain the boundaries of this ground of review. By contrast, if we just stop at reading Fong Thin Choo, we are left wondering precisely how far the court can go.

        Second, E paves the way for saying that a decision can be quashed on grounds relating to evidence that emerged after the decision was made. By contrast, it is conceptually problematic to say that a decision-maker who did not take into account such new evidence had omitted to take into account a mandatory relevant consideration. 

      16. The terms “ultra vires” and “substantive review”, and the grounds of judicial review of executive action

        When a court says that an executive decision was “irrational”, is this a comment about the decision-making process?

        One might think that a statement that the ‘output’ of a reasoning process is irrational is merely an implied statement that the reasoning process itself is defective. The courts used to suggest this, as in Edwards v Bairstow [1955] AC 14 (UKHL) at 36 (“it may be that the facts found are such that no person acting judicially and properly instructed as to the relevant law could have come to the determination under appeal”; emphasis added). But the courts have since turned away from this analysis – GCHQ at 410H-411A:

        “resort I think is today no longer needed to H Viscount Radcliffe’s ingenious explanation in Edwards v Bairstow… of irrationality as a ground for a court’s reversal of a decision by ascribing it to an inferred though unidentifiable mistake of law by the decision-maker. ‘Irrationality’ by now can stand upon its own feet as an accepted ground on which a decision may be attacked by judicial review.”

        Taxonomy

        I take your point that review for error of fact can be described as “substantive review”, in the sense that a decision tainted by an error of fact is void on the ground that it may be wrong in substance. Indeed, several writers take this approach. For example, De Smith’s Judicial Review describes review for error of fact and irrelevant/mandatory relevant considerations as “substantive review” (see 8th ed, [11-022] onwards).

        I also take your point that, taking the idea of “ultra vires” to its logical conclusion, every ground of review is part of review for ultra vires-ness. (This is what I was trying to allude to when I said that “illegality” is something of an arbitrary label, in that all judicial review is review to determine whether something is legal or illegal.)

        It appears to me that the problem arises from the meaning of the phrase “ultra vires”. If we take “vires” to mean the set of all legal limits to someone’s powers, then judicial review on any ground is judicial review for ultra vires-ness. On the other hand, if we take “vires” to refer only to some legal limits (as you suggest), then it makes sense to say that ultra vires-ness definitely includes Tan Seet Eng­-type illegality; might include illegality in the sense of taking into account irrelevant considerations (given the fiction that it is the statute that impliedly specifies what considerations are relevant/irrelevant); and cannot include review for error of fact.

        I cannot say that one use of the word “vires”, and hence the phrase “ultra vires”, is correct and the other wrong. This is because the case law does not disclose clear use of those terms one way or another. A similar debate has arisen involving the word “jurisdiction”. Some cases use that word to refer to “entitle[ment] to enter on [an] inquiry”, and hence say that, once a decision-maker has “jurisdiction”, it is “as much entitled to decide [a] question wrongly as it is to decide it rightly”: see the discussion in Anisminic Ltd v Foreign Compensation Commission [1969] AC 147 (UKHL) at 171C-E. Other cases, such as Anisminic itself, suggest that a decision that breaches any of the legal limits to the decision-maker’s powers is a decision made outside the decision-maker’s “jurisdiction”.

        So, if we take as our starting point the model of “ultra vires” vs. “intra vires”, then the question of taxonomy is something of a question of semantics.

        That’s why I like the model which Daniel Tan alludes to (in ‘An Analysis of Substantive Review in Singapore Administrative Law‘), of input/process/output. Within each of these categories, some things are permissible (or, if you insist on speaking in Latin, “intra vires”) and others not (“ultra vires”).

        Applying this model, the difference between error of fact and irrationality becomes apparent. A statement that the decision-maker has committed an error of fact is a statement about the decision-making process. In fact, strictly speaking, we need not know what the eventual decision is in order to judge whether an error of fact has taken place in the course of decision-making. By contrast, a statement that a decision is irrational is a statement about the decision itself; and, if the decision is irrational, it is no defence to say that the decision-making process has been perfect. It may be that, more often than not, the reason why a decision is irrational is because the decision-making process is problematic in some way. But the following: difference remains: in order to pronounce a decision irrational, we need not identify any particular aspect of the decision-making process that was problematic.

      17. Judicial deference: miscellaneous notes

        The courts do not defer at all to the Executive’s views on the legality of what the Executive does. However, the courts are themselves the ones who pronounce on the legal boundaries of executive power. When the courts decide where those boundaries are to be set, they do so having regard to whether they can police these boundaries in a manner that pays heed to the need for deference in appropriate cases.

        Take, for example, the law on irrationality. The courts insist (so they say) that they set the threshold of irrationality very high because, if it were lower, then the act of applying the test would in substance amount to the courts second-guessing the executive decision-maker on matters on which the court is not competent to rule. So even if you would not describe the application of the test as deferential, the formulation of the test is based on the idea of deference.

        Perhaps the same may be said of some aspects of illegality, such as: refusing to hold that certain questions of fact are questions of precedent fact, on the ground that the executive decision-maker is better suited to answer them (R (A) v Croydon London Borough Council); and holding that certain considerations are neither necessarily relevant nor necessarily irrelevant, but rather that their relevance (if any) is itself a matter for the decision-maker.

        That said, it is true that the concept of deference did develop in the context of debates about proportionality, and, by extension, substantive review generally. (In short, while the proportionality test allows the court to strike down legislation/executive action on the ground that it goes further than is necessary to achieve a certain aim, the court will defer to the legislature/executive on what is necessary to achieve that aim.) So we should not be surprised if deference comes into play more prominently in cases involving substantive review.

        As for Zamir, I take your point that the House of Lords was convinced (albeit wrongly) that its approach was compelled by the legislation, such that all it was doing was to put the statute into action. (You have rightly suggested that the act of implementing a statute is not, strictly speaking, deference; after all, a court has no power to depart from what a constitutionally valid statute, properly interpreted, says; whereas deference is about having power but choosing not to exercise it.) The reason why I described Zamir as a case involving deference to the Executive is this: when the court chose to take the interpretation that it did, it did so for reasons relating to the immigration officers’ institutional competence. (See 949B: 

        “The immigration officer… has to act upon documentary evidence and such other evidence as inquiries may provide. Often there will be documents whose genuineness is doubtful, statements which cannot be verified,  misunderstandings as to what was said, practices and attitudes in a foreign state which have to be estimated. There is room for appreciation, even for discretion.

        The Divisional Court, on the other hand… is… not in a position to find out the truth between conflicting statements… It cannot possibly act as, in effect, a court of appeal as to the facts on which the  immigration officer decided.”)

        So the court thought that a measure of deference to the Executive was called for, and hence chose an interpretation of the legislation that would facilitate such deference to the Executive.


        What exactly does the doctrine of deference do? We don’t have a direct answer because the Singapore courts only started using the language of ‘deference’ in Tan Seet Eng in 2015. But that doesn’t mean that the concept of deference never existed; it always did. To my mind, the concept of deference, in the sense of the court recognising its limitations vis-a-vis the other branches, can have the following consequences — a greater degree of deference could mean:

        1. more “hesitan[cy]” in “holding a decision to be irrational” (Smith, cited in Tan Seet Eng at [92])  — this can mean that the threshold for irrationality is raised, or that the courts will require more by way of evidence in order to be persuaded that that threshold has been met
        2. a greater readiness to hold that a question of fact is not a question of precedent fact
        3. a greater reluctance to hold that certain considerations are irrelevant / must be taken into account
        4. (when applying a proportionality test — not explicitly in Singapore) a greater willingness to accept the executive’s/legislature’s assertion that it would not have been possible to achieve the aim while restricting rights to a smaller extent
      18. Applications to a judge to recuse himself

        A case in which a judge decided to recuse himself is Ong Wui Teck v Ong Wui Swoon [2016] 2 SLR 1067 (HC). Ong Wui Teck, a party who was acting in person, applied for Woo Bih Li J to recuse himself. Woo J held that no cause for recusal had been made out. Nonetheless, he recused himself:

        “I decided to recuse myself not because there was any merit in Mr Ong’s allegations against me but because, in the interest of justice, I was of the view that I should not hear his disputes with Mdm Ong in respect of the mother’s estate since I was contemplating making a complaint about Mr Ong’s conduct to the appropriate authorities.” (at [81])

        In this regard, it is useful to note the remarks of Choo Han Teck J in Re Raffles Town Club Pte Ltd [2008] 2 SLR(R) 1101 (HC) at [7], which Ong Wui Teck had cited:

        “A fair, honest, and impartial judge or member of a quasi-judicial body may be indignant that anyone should think him otherwise. But sometimes reasonable suspicion may arise from a mix of circumstances and unfamiliarity with the reputation of the judge or member of a quasi-judicial body. Under such circumstances, the judge or member of a quasi-judicial body concerned should recuse himself from the case should there be a genuine concern regarding his connection with the case or a party to the case. There is little else he can do. For instance, if he were to swear an affidavit to explain the circumstances and give a full account of the circumstances such as to show that he really has no grievance that might affect his judgment, he might still not fully remove all trace of odour from the air of impartiality. Indeed, the more he seeks to justify his position, the stronger the smell may grow. Hence, a judge or quasi-judicial officer in these circumstances will likely recuse himself and pre-empt any cause for concern.”

        I also mentioned in class that I thought it very unlikely that a recusal application could amount to contempt of court. I wish to clarify that I meant that the mere act of making a recusal application, without more, will generally not amount to contempt. But as the Ong Wui Teck case shows, the manner in which the application is made, or the (purported) grounds on which it is made, can amount to contempt of court. As the Court of Appeal said ([2020] 1 SLR 855 at [29]):

        “The Allegations arose out of the appellant’s recusal application. An applicant in a recusal application would naturally be required to set out the reasons he thinks justify a recusal. This may include an apprehension of bias. A recusal application is, however, not carte blanche or a “get out of jail free” card for an applicant to say what he pleases without any consequence. Any allegation made by an applicant in a recusal application must be backed by some evidence, or at the minimum, supported by rational basis. If so, such allegations will not attract liability for contempt, though whether they will actually succeed in securing a recusal would be another issue. An applicant cannot exploit a recusal application as an opportunity to engage in sarcasm or to violently abuse or ridicule a judge. The Judge opined, and we agree, that criticisms coloured to the extent that they attract the law of contempt will not be protected under the realm of recusal.”

        The Court of Appeal added (at [35]):

        “In a recusal application where the judge considers that the applicant’s assertion of bias has no basis and the applicant has made contemptuous allegations, the judge is placed in a very difficult position. If the judge dismisses the application, he may have reservations that the public may think he is biased. Yet, a judge will not want to be a judge in his own cause especially when the assertions of bias are so repeated and vituperative as to cause the judge to suspect that there has been a contempt of court. In these circumstances, it may be prudent for a judge intending to file a formal complaint of contempt against a party, to recuse himself from hearing matters relating to that party. This is at least until the conclusion of the inquiry or any proceedings arising from that complaint. It is prudent because it removes any suspicion of bias on the part of the judge, strengthens public perception of judicial impartiality and prevents the unnecessary protraction of litigation and wastage of resources arising from further recusal applications. This is not to say that in all situations where an application for recusal is made that the Judge should recuse himself or herself. It all depends on the circumstances of the case.”

        (For the avoidance of doubt, I am not related to any of the parties in the Ong Wui Teck case.)

      19. Constitutional amendments

        There is always a tension between the idea that a constitution (/Constitution) is meant to be enduring and the idea that it is meant to reflect what society holds important: what if the latter changes?

        One could attempt to calculate the ‘amendment rate’ of a Constitution. But numbers alone can be misleading. An ‘amendment rate’ would say nothing about the content or extent of the amendments. (Might it be, or example, that a change to the fundamental rights provisions would be of more moment than, say, a change to the name of a court?) Nor would it consider the various reasons for differences in ‘amendment rate’. One would be the vintage of the Constitution in question: one might think that a Constitution that has existed for a long time has already undergone all the amendments that the people want; on the other hand, it could also be that a jurisdiction has not amended its Constitution much (or at all) because it frequently replaces (or has recently replaced) its Constitution. Another would be the set of rules contained in the Constitution compared to other documents: for example, a legal system could have some — indeed, many — constitutional rules contained elsewhere than in the Constitution (consider, for example, how Singapore’s rules on the mechanism of voting for MPs are largely contained in the Parliamentary Elections Act and not the Constitution). 

        Therefore, an ‘amendment rate’ would, without more, not necessarily be an appropriate basis on which to think that some other jurisdiction is/isn’t similar to Singapore. One could attempt to, say, calculate the number of amendments per unit time, per capita, etc.; but this may or may not be sufficient to consider all the factors that determine the frequency, nature, and extent of the frequency of amendment.

      20. The terms “judicial review” and “unconstitutionality”

        Executive action can be reviewable on the GCHQ grounds (and whatever other common-law grounds there are, e.g., perhaps substantive legitimate expectations), as well as on the ground of unconstitutionality. Legislation is only reviewable on the ground of unconstitutionality.

        Note that “unconstitutionality” can refer not only to a breach of constitutional rights, but also a breach of anything in the Constitution, such as implied constitutional rights (see Daniel De Costa); the separation of powers (consider Liyanage, as well as the remarks in Nagaenthran (CA) about the constitutionality of ouster clauses); and, possibly, the rule of law (not yet seen in Singapore, as far as I can remember).

      21. On constitutions and constitutionalism

        Is a bad constitution a constitution?

        In your Legal Theory and Philosophy course, you will consider the assertion, sometimes made, that “an unjust law is not a law”. Thomas Aquinas famously made that assertion (though I’m not sure whether hew as the first to do so.) Gustav Radbruch, a German legal philosopher, agreed: in his 1946 paper ‘Statutory Lawlessness and Supra-Statutory Law’ (translated and reprinted at (2006) 26 Oxford Journal of Legal Studies 1), he argued that certain laws passed by the Nazis were not truly laws, and, therefore, the evil acts purportedly allowed by these laws could not be said to have been justified by law.

        The response of legal positivists is that the moral merits of a law do not determine its validity: they say that Nazi law is evil law, but is still law — see H L A Hart, ‘Positivism and the Separation of Law and Morals’ (1958) 71 Harvard Law Review 593.

        Hart’s point, as I understand it, is this: we should confront the fact that an evil law is a law; it is better to confront the fact that something cloaked in the mantle of legality can be evil than to blind ourselves through a no-true-Scotsman argument. As he put it in The Concept of Law at 210,

        “What surely is most needed in order to make men clear-sighted in confronting the official abuse of power, is that they should preserve the sense that the certification of something as legally valid is not conclusive of the question of obedience, and that, however great the aura of majesty or authority which the official system may have, its demands must in the end be submitted to a moral scrutiny.”

        And at 211:

        “to withhold legal recognition from iniquitous rules may grossly oversimplify the variety of moral issues to which they give rise”.

        I, speaking for myself, have sympathy with Hart’s views; I think an unjust, bad, weak, or non-democratic constitution is still a constitution. After all…

        … every legal system must, by definition, have a constitution

        In The Concept of Law, Hart said:

        “There are… two minimum conditions necessary and sufficient for the existence of a legal system. On the one hand, those rules of behaviour which are valid according to the system’s ultimate criteria of validity must be generally obeyed, and, on the other hand, its rules of recognition specifying the criteria of legal validity and its rules of change and adjudication must be effectively accepted as common public standards of official behaviour by its officials.”

        In other words, a legal system exists only if people — both ordinary legal subjects and officials — generally follow the rules of that system; if people, en masse, cease to recognise a legal system as creating rules binding on them, then that legal system ceases to exist. 

        In fact, Hart’s two conditions imply a third, namely, that the rules of recognition, change, and adjudication exist in the first place. These rules are constitutional rules. As John Gardner, ‘Can There Be a Written Constitution?’, in Oxford Studies in Philosophy of Law: Volume 1 at 162 <https://johngardnerathome.info/pdfs/writtenconstitution.pdf> put it:

        “A constitution is a conceptual l necessity of every legal system. In every legal system there are rules that specify the major institutions and officials of government, and determine which of them is to do what, and how they are to interact, and how their membership or succession is to be determined, and so forth. Without some such rules, as H.L.A. Hart explained in The Concept of Law, there is no legal system. These rules without which there would be no legal system (or some of them, invariably in combination with some other rules) make up the constitution of that system.”

        So, when people speak of a ‘constitution without constitutionalism’, they do not necessarily deny that it is a constitution; what they are criticising is the content of the constitution — just as when one speaks of an ‘evil law’, one does not necessarily deny that it is a law; what one is criticising is the law’s content.

        Most people who use the word “constitutionalism” use it to refer, not just to anything that limits power, but rather to the limitation of power to a substantial extent/in a meaningful sense. Merely saying that ‘the dictator is the only one who has tons of power’ doesn’t change the fact that the dictator has tons of power. It is cold comfort to say that there is only one dictator and not two. 

        The Constitution of Brunei

        Consider the Constitution of Brunei. It states that the Sultan holds the “supreme executive authority of Brunei Darussalam” (Article 4(1)) and “the power to make laws for the peace, order, security and good government of Brunei Darussalam” (Article 39). Article 84B immunises the Sultan from any court proceedings. Moreover, according to the Preamble, the reason why the Constitution exists in the first place is because the Sultan intended to introduce it; the Preamble asserts that, independent of the Constitution, the Sultan had “the rights and powers of OUR Prerogatives as Sultan and Yang Di-Pertuan…”, which was what gave him the power to enact the Constitution.

        Yet, the Constitution of Brunei does contain constitutional rules. It tells us things like what the sources of law are, and who (else) exercises law-making power and executive power. It is not democratic; nor does it contain all the constitutional rules. But it contains constitutional rules. 

        The Basic Law of Governance of Saudi Arabia

        One of you cited a website which says that the Constitution of Saudi Arabia is not legally binding. I would, with respect, be cautious about such an assertion. I found the assertion on Wikipedia, which cited Daryl Champion, The Paradoxical Kingdom: Saudi Arabia and the Momentum of Reform (2003) at 60. That book does not cite any source for the proposition that “[t]here is no legally binding written constitution”, but page 56 does refer to a “Basic Law of Government” which has a “constitution-like form”. Now, it may be that there is a higher law than the Basic Law — say, a royal decree. But that does not mean that the Basic Law is not legally binding. After all, there can be a hierarchy of legal rule. It may be that a type A legal rule is capable of overriding a type B legal rule. But that does not mean that type B legal rules are not legal rules.

        But let us suppose, for the sake of argument, that the Basic Law contains no legal rules at all. Even then, that would not mean that it is not constitutional in nature. Legal rules are not the only kind of rule; nor are they the only kind of constitutional rules — there are, as one of you pointed out, constitutional conventions.

        In addition, it could be that the function of certain provisions in the Basic Law is to put into writing rules that already existed in unwritten form.

      22. Feedback on students’ presentation: “Because Singapore is a democracy, there should be no judicial review.”

        The topic took the form ‘Because A is true, therefore B is true’. In response, one can ask, among others, questions like the following:

        1. Is A true?
        2. Is B true?
        3. Is there a causal link between A and B?
        4. Would B be true (or false) even if A were false (or true)?

        What was said in the presentation was mostly correct. What could have been improved was bringing democracy into the discussion of judicial review, and explaining how everything that was said contributed to addressing the topic. For instance, what was said about the ultra vires theory of judicial review could have been augmented by stating the implications for democracy, e.g., the ultra vires theory may be more palatable to those who see lawmaking by a democratically elected legislature as the preferred means of lawmaking because, according to this theory, all judicial review is is the act of applying the rules made by the democratically-elected legislature – it is not the courts, which are not democratically elected, making up rules.

        The presentation described how various constitutional provisions reflect certain democratic values. To complete the argument, it would be good to have said more on the extent to which judicial review ought to be the means by which those values are upheld. After all, other bodies are also charged with upholding those values, both in and of themselves and as reflected in the Constitution. Consider, for instance, the fact that MPs must take an oath saying that they will “preserve, protect and defend the Constitution of the Republic of Singapore” (see the First Schedule to the Constitution).

        The presentation touched on various aspects of democracy – voting; democratic values; popular participation in decision-making. This could have been made more explicit. The presentation described a lot; ideally, the relevance of each fact to the topic (for example, one may ask: what does justiciability have to do with democracy? What is the relevance of Thio Li-ann’s proposed approach to judicial review?) would have been discussed in more detail.

        As for the rule of law and the separation of powers, more could have been said on how these follow from democracy. If there is no necessary connection between these and democracy, then saying that judicial review upholds the rule of law/the separation of powers doesn’t quite address the question of how judicial review sits with democracy. If there is a necessary connection, this should be explained explicitly. Consider also: can there be a society with democracy but no judicial review (as Jeremy Waldron arguably envisages, at least under certain conditions: see the article referred to below)?

        For further reading, consider Aileen Kavanagh, ‘Participation and Judicial Review: A Reply to Jeremy Waldron’ (2003) 22 Law and Philosophy 451 <https://www.jstor.org.libproxy.smu.edu.sg/stable/3505148>. (Note that she and Waldron are focusing on judicial review of legislation on the ground of unconstitutionality.)

      23. Feedback on students’ presentation: Is there a better way for judges in Singapore to be appointed?

        In general, the problem with choosing judges is that whoever has the power to choose may choose someone like them, and so perpetuate his mentality. One way to reduce this problem is to reduce the extent to which that “mentality” can play a role in the judge’s decision-making. The more unambiguous the laws which the judge has to apply, and the less power the judge has to change the law, the less likely it is that the judge can decide cases in line with a certain ideology/agenda/mentality instead of in accordance with law.

        The group is to be commended for its sensitivity to the need to select appropriate comparative case studies (as opposed to, say, picking three countries at random to compare with Singapore). When addressing the question of whether there is a “better way” for Singapore, it is useful to consider (a) the extent to which the status quo in Singapore falls short; (b) possible alternatives; (c) whether those alternatives can be adopted in Singapore; (d) whether, if adopted, those alternatives will bring about an improvement. The group did well in describing the alternatives, their merits, and the feasibility of adopting them; more could have been said on their merits relative to the status quo. In describing the status quo, it would have been useful to distinguish between (a) problems that could arise; (b) whether those problems do arise.

        More space for analysis could have been created by organising the presentation in a different manner. Instead of:

        1. Country A
          1. Describe its system
          2. Evaluate its system
          3. Can/should Singapore adopt its system?
        2. Country B
          1. Describe its system
          2. Evaluate its system
          3. Can/should Singapore adopt its system?
        3. Country C
          1. Describe its system
          2. Evaluate its system
          3. Can/should Singapore adopt its system?

        the following might have been more efficient:

        1. Issue 1: who qualifies to be a judge?
          1. Status quo in Singapore
          2. Possibilities from Countries A, B, C…
          3. Can/should Singapore change?
        2. Issue 2: who shortlists potential judges?
          1. Status quo in Singapore
          2. Possibilities from Countries A, B, C…
          3. Can/should Singapore change?
        3. Issue 3: who makes the final choice?
          1. Status quo in Singapore
          2. Possibilities from Countries A, B, C…
          3. Can/should Singapore change?

        The advantage of the latter scheme is that, by going feature-by-feature rather than country-by-country, it allows for an argument that Singapore should pick some features from country A, some from country C, etc. to be made more clearly and in fewer words.

        Had time permitted, it could have been useful to distinguish between different types of judge. While all judges should have certain characteristics, e.g. independence, different types of judge may face different threats to these characteristics, or the level of importance of these characteristics may differ depending on the type of judge. For example, one argument for judicial security of tenure is that, otherwise, if X has the power to appoint judges then X could threaten a judge into ruling in X’s favour by threatening not to renew X’s appointment. But this threat would not work if that judge does not hear cases involving X in the first place.

        It may also be useful to consider relationships between different types of judge. If judge A faces certain threats to his independence, but there is a right of appeal against judge A’s decisions to judge B, and judge B does not face such threats, then on the whole the judicial system does not suffer from such threats.

        Some links to sources we mentioned in passing, for those who are interested:

        1. The UK Judiciary’s YouTube channel: https://www.youtube.com/channel/UCi3XytDJY8a3I9_vL7A_5SQ – consider how the judiciary presents itself to the public.
        2. Gary Low, ‘Emphatic Plea for the Empathic Judge’ (2018) 30 Singapore Academy of Law Journal 97 <https://journalsonline.academypublishing.org.sg/Journals/Singapore-Academy-of-Law-Journal/e-Archive/ctl/eFirstSALPDFJournalView/mid/495/ArticleId/1260/Citation/JournalsOnlinePDF>.
      24. Feedback on students’ presentations/essays: Should Singapore’s Parliament have a second chamber?

        Feedback on essays (in one year)

        The question was ambiguous: it could have meant one of the following:

        (a) Would Singapore be better off with a bicameral Parliament that is similar to the UK’s? 

        (b) Would Singapore be better off with a bicameral Parliament? By the way, the UK has a bicameral Parliament. 

        Neither interpretation was necessarily correct to the exclusion of the other. However, either interpretation would have called for a discussion of the possibility that Singapore should adopt something other than a carbon copy of the House of Lords. (Interpretation (a) must have done so because it would have called for a discussion of what characteristics would make something similar to the UK’s Parliament. In other words, a person who says “a bicameral Parliament like the UK’s” is likely to be referring to a bicameral Parliament that shares certain features in common with the UK’s — which raises the question of what these features are.)

        Weaker arguments appeared to be premised on the assumption that the person asking the question was proposing to have a carbon copy of the House of Lords in Singapore. For example, some argued that Singapore should not have a Parliament like the UK’s because Singapore is too small to have hundreds of Lords. This argument appeared to be attacking a straw man: is it not more likely that the person asking the question is proposing that Singapore’s Parliament have, say, a second chamber whose functions are similar to that of the House of Lords, or a second chamber the ratio of whose size to Singapore’s population is the same as the ratio of the size of the House of Lords to the UK’s population?  

        Similarly, some strenuously argued that there is no space in Singapore for a building for a second chamber of Parliament to meet. Putting aside the question of whether this assertion is true, one wonders whether something as banal as that should be relevant. (What’s next: will someone argue that Singapore should not have a second chamber like the House of Lords because the architectural style wouldn’t look good in central Singapore?)

        It would have been useful to distinguish between the costs and benefits of having a bicameral Parliament on the one hand, and the costs and benefits of switching to having a bicameral Parliament on the other.

        It was necessary to distinguish between arguments for bicameralism, and arguments in favour of a particular form of second chamber. Some essays confused the two, and fallaciously argued that, because the House of Lords has some feature that would not work in Singapore, therefore bicameralism can never work in Singapore.

        It is not true that members of the House of Lords are not subject to party whips.

        Some wrote about the possibility of “deadlock” or “gridlock”. It was necessary to distinguish between a “deadlock” or “gridlock” on the one hand, and a delay on the other. The words “deadlock” and “gridlock” connote something that is intractable or, at least, cannot be straightforwardly resolved. So it may be an exaggeration to say that bicameralism in the UK suffers from gridlock. It is true that there are delays(which, by the way, are not necessarily a bad thing), but the effect of the Parliament Act 1911 (as modified by the Parliament Act 1949) is that, if the House of Lords does not approve a Bill, the Bill can still become law if the House of Commons approves it; all the House of Lords can do is to delay the passage of the Bill.

        When speaking of ‘checks’ or ‘checks and balances’, it is necessary to consider what risk is being checked against. Some essays did not do so, and hence argued that a second chamber is unnecessary because judicial review already exists as a ‘check’. The problem with such an argument is that judicial review is a check against unlawfulness (including unconstitutionality), whereas unlawfulness is not the only ground on which a second chamber might refuse to approve a bill.

        It was useful to consider the potential impact of publicity attending the debates of a Bill in two Houses and the ‘ping-pong’ between the Houses.

        Something might be said about the role of legislative debates, including debates involving opposition MPs, in statutory interpretation.

        It is possible for both chambers of a bicameral Parliament to be popularly elected, and yet for different electoral procedures to apply to each of the two chambers, or for there to be different pools of candidates for each of the two chambers. 

        Feedback on presentations (in another year)

        One would think that, for a bicameral arrangement to make sense, there must be some difference between the two chambers in terms of who can be a member; how/when members are selected; what powers they have; etc. Otherwise, there will just be two bodies which are, for all intents and purposes, copies of each other.

        This question calls for a discussion of what a second chamber could add to a unicameral Parliament. This, in turn, depends on (a) what shortcomings the existing Parliament has; (b) what changes a second chamber could make to address those shortcomings; (c) how successfully those shortcomings could be addressed.

        As for (a), ideally one would avoid generalisations like ‘Parliament can pass laws too easily’. This is somewhat vague: Is the problem that it should take a minimum amount of time for laws to be passed (and, if so, why, and how much time?) Or is it that Parliament may act as a mere rubber-stamp? Is the argument against certain features that Singapore’s Parliament intrinsically has, or against certain features that Singapore’s Parliament happens to have now?

        While there was discussion of possible problems with the design of Singapore’s Parliament, a bit more could have been said on whether those problems arise from the fact that there is one chamber or from something else.

        In general, when comparing different legal systems, it is useful to think in terms of functions rather than institutions. For example, if you were writing about Singapore’s Presidency, it would probably not be right to draw comparisons with the USA’s Presidency. While the two offices have the same name, their functions are very different. Conversely, when thinking about what a second chamber could do, the group did well in comparing second chambers in other countries’ legislatures to things like the PCMR and NMP scheme. They have different names, but arguably serve similar functions.

        In general, when considering the benefits and drawbacks of something, it is useful to interrogate certain assumptions that we may have. Suppose we say that bicameralism leads to inefficiency. But it is worth asking: (a) Must it lead to inefficiency? (b) Is efficiency an unqualified good?

      25. Why did the Court of Appeal need to issue the Practice Statement (Judicial Precedent) 1994?

        In 1994, Chief Justice Yong Pung How issued the Practice Statement (Judicial Precedent) [1994] 3 SLR(R) xcvii, which states:

        “With the abolition of all appeals to the Judicial Committee of the Privy Council and the establishment of the Court of Appeal as the final appellate court in Singapore, we consider it desirable that we state the use of precedent in the Court of Appeal in future.

        We recognise the vital role that the doctrine of stare decisis plays in giving certainty to the law and predictability in its application to similar cases. However, we also recognise that the political, social and economic circumstances of Singapore have changed enormously since Singapore became an independent and sovereign republic. The development of our law should reflect these changes and the fundamental values of Singapore society.

        Accordingly, it is proper that the Court of Appeal should not hold itself bound by any previous decisions of its own or of the Privy Council, which by the rules of precedent prevailing prior to 8 April 1994 were binding on it, in any case where adherence to such prior decisions would cause injustice in a particular case or constrain the development of the law in conformity with the circumstances of Singapore.

        Therefore, whilst this court will continue to treat such prior decisions as normally binding, this court will, whenever it appears right to do so, depart from such prior decisions. Bearing in mind the danger of retrospectively disturbing contractual, proprietary and other legal rights, this power will be exercised sparingly.

        This statement is not intended to affect the use of precedent in the High Court or in any subordinate courts.”

        One of you asked: Why? Wouldn’t Privy Council cases naturally not bind any Singapore court any more, now that appeals to the Privy Council had been abolished? And wouldn’t the Court of Appeal naturally have power to overrule its own previous decisions, simply because it was the highest court? Besides, even when the Privy Council was the highest court for Singapore, why wouldn’t the Court of Appeal have been free not to follow its own previous decisions?

        Believe it or not, the law used to be (or at least appear to be) that the Court of Appeal of Singapore was generally bound by its own previous decisions: see page 409 of this article. This followed the position in Young v Bristol Aeroplane, an English case which held that the Court of Appeal of England and Wales is generally bound by its own previous decisions.

        In the UK, it was the Judicial Committee of the House of Lords (now the Supreme Court) – the highest court of that land which, since 1966, has not been bound by its previous decisions, as confirmed in the House’s 1966 Practice Statement. In Singapore, before appeals to the Privy Council were abolished, the Privy Council, too – the highest court in Singapore’s hierarchy of courts – was not bound by its previous decisions (see this case at [38]).

        In short, immediately before appeals to the Privy Council were totally abolished in Singapore (i.e., when the Privy Council was still the highest court of Singapore), the position in Singapore was the same as in England and Wales: the highest court was not bound by its own previous decisions, but the next-highest court (the Court of Appeal) was.

        So when appeals to the Privy Council were abolished in 1994, it made sense for the court to declare that the Court of Appeal – now Singapore’s highest court – was not bound by its previous decisions. You might think that that obviously followed from the simple fact that it was now the highest court. But note that in the UK, for a few decades before 1966, it was thought that the House of Lords – the highest court – was bound by its own decisions. So there may have been some reason to doubt that Singapore’s Court of Appeal could overrule its own previous decisions, and the 1994 Practice Statement put beyond doubt that it could. Besides, one could imagine that there may have been lingering doubts about whether the Court of Appeal would continue to be bound by previous Privy Council decisions; if so, the Practice Statement must also have removed the doubt.

        It is also worth pointing out a contemporaneous development: the passage of the Application of English Law Act 1993, and its coming into force on 12 November 1993. Section 3 stated that the common law of England up to that date would “continu[e] to be part of the law of Singapore”, but only “so far as it is applicable to the circumstances of Singapore and its inhabitants and subject to such modifications as those circumstances may require.” We can see the 1994 Practice Statement as a continuation of this trend of putting beyond doubt that the courts of independent Singapore, while building on English legal foundations, were free to develop the law for Singapore as they saw fit.

        Note also that, before 1993, the Court of Appeal only heard civil cases; there was a separate Court of Criminal Appeal that dealt with appeals in criminal cases. The latter was folded into the former only in 1993. To the extent that this may have created confusion about precedent, the 1994 Practice Statement must also have removed the confusion.

      26. On “CRuPAC”, “IRAC”, etc.

        “CRuPAC”, “IRAC”, etc. are useful guides to thought, in that they remind us that law is a mode of doing justice by reference to rules: “the only justice that can be attained by mortals, who are fallible and are not omniscient, is justice according to law; the justice which flows from the application of sure and settled principles to proved or admitted facts” (Cowcher v Cowcher [1972] 1 WLR 425 (EWHC) 430). 

        But they are not necessarily useful guides to writing, in that, if unthinkingly followed in a formulaic manner, they can lead to writing that is tedious, clunky, and simply not space-efficient.

        While we need to be clear about what separate issues exist (and how to divide a problem into issues can itself be debatable), what the relevant rule is (which can itself be debatable), and what the rules have to say about the relevant facts (which can itself be debatable), there is no need to insist on devoting (say) X sentences or Y paragraphs to each of them. In fact, it may be possible to say all one needs to say about them in just one paragraph.

        If one is not misled by red herrings (if any), avoids being unnecessarily verbose, avoids spending too much space discussing trite/undisputed issues, and instead focuses on the more contentious issues, one will be able to produce a meaningful answer to the questions that will be asked.

        I propose to illustrate all this with some examples from my old business law course materials. (I have chosen to ensure that I am providing suggestions as to style, not content. Of course, a deeper level of detail may be warranted.)

        Example 1

        “There is a contract between A and B: there have clearly been offer, acceptance, consideration, and intention to create legal relations. Has it – in particular, clause 3 – been breached? The problem is that the word “X” in clause 3 is ambiguous. Hence, we must interpret clause 3, in light of…, with a view to determining the parties’ intentions at the time the contract was made (Zurich v B-Gold). The question is whether the word “X” mean P or Q. Given the conversation between the parties, it would appear that both parties proceeded on the basis that the word “X” mean P.

        If this is correct, then the contract has been breached as what was delivered is clearly Q. On the other hand, if interpretation Q is correct, then the contract has been performed exactly as promised, and B may not sue A for breach of contract although B feels disappointed.

        If the contract has been breached, then what are the consequences? Damages are payable, calculated according to the rule that… If the contract had been performed… Now, however… Therefore, the sum of damages payable is…

        As there are no outstanding obligations under the contract, it does not matter whether or not B has a right to terminate the contract.”

        This example identifies one issue: whether a contract exists. But it swiftly states the rule (in few words), applies it, and states the conclusion. It then segues into a second issue: whether the contract has been breached. To address that issue, we must first address a third issue: the interpretation of clause X. We have stated a rule on interpretation, applied it, and posited two possible conclusions. Those two conclusions then lead to two possible answers to the second issue. We then move to discuss a fourth issue (the consequences) of breach; but we are careful to state clearly that this issue only arises if the contract has been breached. This may be divided into two parts: damages, and termination. We address the first part, again using our classic “IRAC” model. As for the second part, we have explained why it is actually not an issue, and hence no further discussion is required.

        Example 2

        Consider the following example. It uses the “IRAC” structure, but does not rigidly write “The issue is… The rule is… Applying the rule…” Instead, it addresses all 4 points – not necessarily in order, but still smoothly – and does so in a way that uses relatively few words, while discussing multiple issues and the relation between them:

        “[Issue:] If A is liable for breach of contract, he will claim that his liability has been excluded by an exemption clause contained on a sign on the wall. [Conclusion:] It is unlikely that his liability has been excluded as [Rule:] the sign could not reasonably have come to the notice of B, [Application:] having been put in an obscure location and in small print. [Issue:] Even if it had been excluded, [Conclusion:] it is unlikely to apply in favour of A: [Rule + Application:] the word “consequences” on the sign is ambiguous (does it include breach of contract?) and the [Rule + Application:] contra proferentem rule suggests that it will be interpreted in B’s favour.”

        Example 3

        “In any event, the exclusion clause, even if incorporated into the contract, probably falls foul of section 3 of the UCTA. As mentioned above, the requirements in sections 1 and 3(1) are met. However, it is doubtful whether the exclusion clause meets the ‘requirement of reasonableness’ because…”

      27. The taxonomy of grounds of judicial review of executive action

        I wrote these notes in response to questions from students on an essay question that called for students to evaluate the “traditional taxonomy of grounds of judicial review” set out in the GCHQ case.


        Principles such as the rule of law, alleged legality/merits distinction, etc. are not themselves the taxonomy; they are principles that (according to Jhaveri) are supposedly “reinforce[d]” by the GCHQ grounds of review. Examples of arguments about taxonomy could be:

        • It makes no sense that some of the categories within the taxonomy are defined in terms of the type of conduct that is unlawful (e.g. illegality, procedural impropriety), while others are defined in terms not only of the type of conduct but also the level of egregiousness (irrationality).
        • One might say that, if proportionality is recognised as a ground of review, it ought to sit in the same category as irrationality, for both are grounds of review relating to the substance of a decision. However, that assumes that the onlydefining characteristic of irrationality as a conceptual category is that it centres on the substance of the decision. Irrationality has at least one other defining characteristic: namely, that it functions as a ‘safety net’ which courts reserve to themselves as a legal basis to strike down decisions which are unlawful but the nature of whose unlawfulness is not easy to articulate clearly. In other words, irrationality exists precisely to be deliberately vague, and so any new ground that purports to improve on this vagueness should exist alongside, not instead of, illegality.
        • One would think that a taxonomy of grounds ought to distinguish between (a) grounds relating to the procedure applied by a decision-maker and (b) grounds relating to the decision itself. Given that procedural impropriety and illegality both fall into the former category, why should they be two separate categories?
        • Diplock never sought to create a taxonomy; he only sought to create a list of grounds of review. Therefore, not only is it wrong to attempt to treat the GCHQ grounds of review as set in stone; it is not even necessary that new grounds be developed only by analogy with the existing ones.
          • Counter-argument: if he did not mean to create a taxonomy, why did he take pains at 411A-411B to explain that “procedural rules that are expressly laid down in [a] legislative instrument” are rules of “procedural impropriety” and not “illegality”? After all, such rules could just as well be said to relate to “illegality”, as such rules form part of “the law that regulates [the decision-maker’s] decision-making power”. So Lord Diplock was aiming to do more than just create an exhaustive laundry-list of grounds of review that existed at the time…
        • It is ironic that “illegality, irrationality and procedural impropriety” has become a stock phrase, when it was never intended to be anything more than shorthand for…
        • It is impossible to understand Diplock’s taxonomy without considering it alongside the distinction he drew between rights and legitimate expectations…
        • Why did Diplock even attempt to develop any sort of taxonomy?
        • It is wrong to think of Diplock’s taxonomy as prescriptive, for he meant it to be descriptive only. Therefore, even if Diplock’s taxonomy is outmoded, it is outmoded as a description; but that does not make it confusing as a guide for the future, for it was never intended as a guide for the future to begin with.

        These points may or may not be correct, but the point is that they purport to be points relating to how grounds of review are taxonomised.

        To build an essay around the ‘legality-merits’ divide and the GCHQ‘taxonomy’, one could begin by attempting to render an account of ‘legality’ and ‘merits’ as conceptual categories, then explain why the GCHQ taxonomy of grounds is meant to cover the entirety of the former and none of the latter. But even if one could, and even if one could show that the ‘legality-merits’ distinction is outmoded/confusing, this would not directly address the question of whether the GCHQ taxonomy is outmoded/confusing unless one could show that the GCHQ taxonomy serves no other purpose other than to give effect to the ‘legality-merits’ divide – does it? Unless one can show a direct, inextricable link between GCHQ and the ‘legality-merits’ divide, a commentary on the latter is not necessarily a commentary on the former, and it would be better for an essay on the topic assigned to focus on GCHQ itself rather than to assume that the ‘legality-merits’ divide is a perfect proxy for GCHQ.

      28. On the concept of “arbitrariness” in case law relating to Article 12

        Ong Ah Chuan and subsequent cases on Art 12(1) appear to use “arbitrary” to mean that there is no rational nexus between the differentia and the object of the statute. (In Ong Ah Chuan at [37], the Privy Council said that the differentia based on the mass of heroin trafficked in was “not purely arbitrary but bears a reasonable relation to the social object of the law”.)  (Prabagaran, at [93], added the suggestion that, for a law not to be “arbitrary”, it must have a “clear purpose”.) 

        Chng Suan Tze extended what was said in Ong Ah Chuan by saying that a statute that creates discretion that could be used in a way that has no relation with the purposes of the law creates “arbitrary powers of detention”. 

        But arguably “arbitrary” can mean more than just exercising powers in a manner other than the purpose for which they were conferred.  In Chng Suan Tze at [82], the Court of Appeal agreed that “if the discretion… is subjective, that would allow arbitrary detention…” What does “subjective” mean? 

        Earlier in the judgment, the Court of Appeal uses that word to refer to a type of discretionary power which can be exercised in any way, regardless of “[w]hether or not the facts on which the order of detention is to be based are sufficient or relevant”, as long as the “detaining authority” has a particular “state of mind” (namely, ‘being satisfied’): at [52], quoting Karam Singh. In the case of a “subjective” discretion, there can be no “judicial inquiry into the sufficiency of the grounds to justify the detention”: ibid.  Later, at [66], the Court of Appeal suggested that the “subjective test” means that “the court cannot require [a] police officer [making an arrest] to prove to the court the sufficiency of the reason for his belief” – which suggests that a “subjective” discretion to do something means the power to do that thing even without any supporting evidence. And at [86], in rejecting the “subjective” test, the court said that it must be the judge of the “lawfulness” of executive action (citing GCHQ). 

        From all this, it seems to me that “subjective” doesn’t just mean “exercised for a purpose other than that conferred by the statute”; more broadly, it refers to a power not subject to legal limits (other than someone’s claiming that he/she is “satisfied” of something). And the court appears to have used “subjective” and “arbitrary” interchangeably: see [82], and also [100] (where Karam Singh – previously described as allowing for “subjective” discretion – is now said to give rise to “arbitrary” power). 

      29. For the purposes of Singapore’s law of standing, what does “right” refer to in cases involving judicial review of administrative action?

        [I’m referring to Tan Eng Hong, Vellama, and Kenneth Jeyaretnam below.]

        I think our best bet is that it could refer to a ‘right’ to have a licence, a ‘right’ to a hearing, a ‘right’ to not be dealt with on the basis of irrelevant considerations, a ‘right’ to have legitimate expectations duly considered, etc.

        One reason why courts have not told us is that it rarely matters. In many cases involving administrative law, the decision being impugned will be one that targets the applicant in particular (e.g. a decision to not make a grant; a decision to revoke a licence; etc.) According to Vellama at [33], the law on standing aims to prevent a situation where “the courts will be inundated by a multiplicity of actions, some raised by mere busybodies and social gadflies, to the detriment of good public administration”; according to Tan Eng Hong at [82], “‘mere busybodies’ whose rights are not affected” cannot have standing. If only one person is affected, the purposes of the law on standing are met by simply granting standing to that one person and to nobody else, for there is no definition of “busybody” or “gadfl[y]” according to which that one person would be a busybody/gadfly.

        Perhaps the courts ought to take a step back from formulaic references to “private right” etc., and simply ask themselves why the law on standing exists. If the purpose is simply to ensure that the person before the court has the best arguments and evidence to make the application, then clearly the one person affected by the decision would – and should – be the one to have standing. Consider the pragmatic approach in Comptroller of Income Tax v ACC [2010] 2 SLR 1189 (CA): the Comptroller had concluded that ACC was personally liable to pay withholding tax; therefore, “if [ACC] did not have locus standi to commence proceedings for a review of the Comptroller’s determination [that ACC had to pay withholding tax], we were unable to see who else would have locus standi to do so.”

      30. What are ouster clauses?

        An ouster clause is a statutory provision that purports to prevent the courts from performing judicial review — or, as one might say, to oust the courts’ power of judicial review. Here are some examples of ouster clauses:

        • “any decision of a claim or question [by some executive officer] shall be final”
        • “The determination by [some executive officer] of any application made to them under this Act shall not be called in question in any court of law.”
        • “determinations, awards and other decisions of the Tribunal… shall not be subject to appeal or be liable to be questioned in any court”
        • “… the decision of the Minister shall be final and not open to review or challenge on any ground whatsoever.”

        We have discussed at length the importance of judicial review of executive action. Therefore, the courts often omit to give effect to ouster clauses, and therefore hold that they can perform judicial review of the executive action in question. How, you ask, can they do so without being said to be acting in defiance of the law? There are, broadly speaking, two techniques:

        1. A court can hold that the provision, properly interpreted, doesn’t actually have the effect of ousting judicial review. For example, a court may hold that the word “decision” only refers to lawfully-made decisions and not unlawful ones. 
        2. A court can hold that the provision is unconstitutional. The Singapore courts have not yet explicitly said such a thing, but it is likely that they will. In Nagaenthran a/l K Dharmalingam v Public Prosecutor [2019] 2 SLR 216 at [74], the Court of Appeal described an ouster clause as being “constitutionally suspect for being in violation of Art 93 of the Singapore Constitution as well as the principle of the separation of powers”. I attempted to unpack this argument in Benjamin Joshua Ong, ‘The constitutionality of ouster clauses’ (2019) 19 Oxford University Commonwealth Law Journal 157 <https://ink.library.smu.edu.sg/sol_research/2915/>. In short:
          1. Article 93 says that the Supreme Court (part of which is the High Court) have judicial power. The power to perform judicial review is, by definition (through conventional usage), part of the High Court’s judicial power. (Tan Seet Eng at [1] says so: “Judges are entrusted with the task of ensuring that any exercise of state power is done within legal limits.”) So an ouster clause would stop the High Court from exercising the judicial power that the High Court, by definition, has.
          2. If some authority’s decisions are immune from judicial review, then they have the final power to determine what their legal limits are. If someone has the final power to determine what their legal limits are, then they, in effect, exercise judicial power. In other words, if a statutory provision prevents judicial review of some executive authority’s decision, then its effect is to confer judicial power on that executive authority. That would run contrary to Article 93, which, properly interpreted, vests judicial power exclusively in the courts — for an executive authority is not a court.

        Now, it may be that a statutory provision makes the legal limits to some executive body’s powers very wide. But, however wide the powers are, it is the court’s job to ensure that their limits are not transgressed.

      31. Miscellaneous notes about democracy

        I cannot at all remember the substance of the discussion in class in response to which I wrote the following.


        Should the elderly be disenfranchised?

        One argument in favour of disenfranchising the elderly is that they will vote for the incumbent party out of ‘gratitude’ rather than out of a genuine belief that such a vote would be best for the future of the country. One response was that the argument that the elderly will vote on this basis is not necessarily true. I would also like to call into question the assumptions that it is wrong to vote out of ‘gratitude’; that the only basis for voting ought to be a consideration of the future of the country; and, more generally, that it is possible and desirable to formulate rules setting out the permissible bases on which to decide whom to vote for.

        Speaking for myself, I do not think that it is necessarily wrong to vote for somebody merely because they have delivered good things in the past. Indeed, one feature of a system of representative democracy is that a party can be incentivised to do well by the possibility of being re-elected. If a ‘gratitude’ vote is a reward for past performance, then we can expect a party to aim to perform well in the present with a view to receiving ‘gratitude’ votes in future, and hence we can expect a party to aim to perform well in the future with a view to receiving ‘gratitude’ votes in the more distant future.

        It is also possible that the elderly, in deciding whom to vote for, do not consider the interests of future generations. To the extent that this is merely an instance of the general phenomenon of self-interested voting, it is an inevitable feature of democracy; the solution, to my mind, is a robust system by which the rights of people other than majority voters are protected, and not necessarily a restriction of the right to vote. To the extent that the issue is that nobody should be entitled to vote in a manner that neglects the interests of future generations, the issue is this: can, and should, there be a rule (express or implied) stating what considerations may and may not be taken into account by voters? Short of systems in which people vote on individual issues, I do not know of systems with such a rule.

        Are we repressed?

        While I was indeed once punched in the face (without my consent) while I was in the UK, the punching took place at the hands of a private individual and not at the hands of the state. Accordingly, the only basis on which I can fault the state is for having failed to protect me from being punched; but I do not attribute especial blame to the state for this, because (a) no state — not even Singapore — succeeds perfectly in preventing crime, and a single anecdote does not constitute data; (b) neither the UK nor England positively encourages such crime. Accordingly, the experience of having been punched in the face plays no part in any comparison I may seek to draw between the experience of living in the UK and that of living in Singapore. In other words, I do not consider that Singapore is necessarily better at protecting the right not to be punched in the face than is the UK.

        This incident was brought up in the context of a discussion about whether we felt repressed in Singapore. I agree with [student’s] point that (if I may paraphrase it) to the extent that this is a question of how we feel, we cannot say what it is like being inside or outside Plato’s cave without, respectively, having experienced life outside or inside it. At the same time, I do not deny that it can be possible to formulate an objective basis on which one may seek to compare life in Singapore with life overseas even without having lived overseas. At the same time, I do not deny that such an objective basis will not paint a complete picture, for — and this is a point that lawyers often forget — law is not only a set of rules and principles, but also an experience lived by individuals. (There is a school of thought according to which the effect — if not the purpose — of legal education is to brainwash students into forgetting the latter.)

        But even without going so far as to speak of lived experience, I would like to suggest this: that, if we want to paint a complete picture of what life is like under one legal system, we should consider not only what liberties its subjects have, but rather what benefits, more broadly conceived, that its subjects enjoy. For a liberty (according to at least one possible definition) is but an absence of an impediment to doing something (which has sometimes been termed a negative right), or a prohibition of creating such an impediment; whereas a benefit is something positive, and includes a positive right (namely, a duty on the part of the state to do or facilitate something). In other words, I agree with the point alluded to by [students], namely (if I may paraphrase it), that we should be concerned not only with non-repression but also with flourishing.

        In this regard, what stands out to me about political life in the UK is not the rights set out in (say) the European Convention on Human Rights, but rather avenues which facilitate democratic participation (whether directly or indirectly), such as access to information through the Freedom of Information Act; accessible avenues of redress (even for perceived wrongs that fall short of illegality) through ombudsmen; and various formal opportunities to make one’s views known on municipal matters (such as one’s right to go to what is literally a town-hall meeting to state one’s views on someone’s proposal to open a pub: see https://www.gov.uk/guidance/alcohol-licensing#community-involvement-in-licensing). One could call these rights, but they are, more precisely, avenues of democratic participation, whose very availability encourages democratic participation. Without commenting on the benefits and/or drawbacks of the availability of such avenues (consider, for instance, the possibility of frivolous Freedom of Information requests), I would like to suggest that such things — and not only the rights enumerated in bills of rights — should be considered to be defining features of a community.

        The power of the law over us

        [Student] asked about the idea of citizenship and by what rules we decide who is a citizen. The answer I gave is that citizenship is a legal concept and the rules on who is a citizen of Singapore are found in Part X of the Constitution. But the context of this discuission was the issue of democratic participation; so I suppose the question is not who is a citizen, but rather who is part of the demos.

        I suppose there are actually two related issues at stake. First, if, as Hart tells us, in the final analysis, a legal system rests on the social fact of the acceptance of some secondary rules by a community of people, what constitutes the community and who is part of it? Second, if this legal system is democratic, i.e. ruled by the people, which people play what role in ruling?

        The first question may be one for a sociologist. As to the latter, all I will say is that it is not necessarily the case that one either participates (through voting) or does not. Consider, for example, the case of the student who writes a letter to his/her MP: he/she thereby participates in democracy, even though he/she does not vote. Moreover, it is possible that one’s right to participation is exercised one one’s behalf: consider the adult who, when voting or advocating that the government take a certain policy position, does so with a view to the interests of his/her baby — is that baby not thereby made a member of the demos?

        Another, potentially more difficult, issue is this: what justifies the coercive power of law? That is a question that is of particular interest to a non-democracy, but is also relevant to a democracy, for, even though (some of) the people have a say in governance in a democracy, at least one thing is foisted upon people, namely, the very democratic system. One answer may be that, because any legal system rests on the social fact of popular acceptance, any legal system may be changed simply by an alteration of this social fact. But, perhaps, it is more accurate to say that any legal system rests on the social fact of popular acquiescence, for a state, by definition, has a monopoly on the use of violence, which can make it practically impossible for the people to effect regime change. It may be that a legal system contains rules allowing for systemic change (as, for example, the rules allowing Singapore’s Constitution to be amended). But what about the person who does not agree with a particular change: is that person not thereby coerced against his will? 

        I cannot pretend to have an easy answer to that question. One tempting answer might be the notion of a social contract. But a contract is, by definition, entered into with the consent of the parties thereto; what, then, of the person who has not had a chance to express a choice as to whether or not to accept the system? What of the person who would not accept the system even if asked? And what of the person who has positively chosen to reject the system? Perhaps a more promising approach is simply to accept that law is, by its nature, something foisted upon us. And, given that it is, perhaps the real question is: under what conditions is it justified for something to be foisted upon us? For example, is the fact that we are thereby saved from a life of precariously wandering through the wilderness a sufficient justification?

        One last thing: Agreement is not the only thing that can justify one’s coming under an obligation. Another can be the very act of receiving a benefit. That may explain the arguments for why a male Singapore citizen is not allowed to renounce his citizenship before completing his National Service obligations, namely, that he has already received the benefits of Singapore citizenship, so the state is justified in imposing the burdens upon him. (There are other problems with this argument but I shall not go into them here.)

      32. Sections 16-17 of the SCJA and the jurisdiction of the General Division of the High Court

        A student asked about ss 16-17 of the Supreme Court of Judicature Act 1969, which state what the civil jurisdiction of the General Division of the High Court encompasses. But doesn’t Hinds v R mean that that court, by definition, has “unlimited original jurisdiction” in “substantial” civil cases (which, it is submitted, must mean all civil cases not covered by the jurisdiction of a subordinate court)?


        I understand section 16 of the SCJA to relate merely to the manner in which the jurisdiction of the High Court is to be invoked, rather than the subject-matter of the jurisdiction. (Similarly, I understand the purpose of s 15 to be to expand the range of situations in which the High Court has criminal jurisdiction beyond what is provided for at common law, as, for example, in the case of limb (e).)

        As for section 17, I think one possibility is that it expands the jurisdiction of the High Court beyond what it has at common law, or at least clarifies doubts about the extent of its jurisdiction. (For an example of such doubts: The “Ocean Jade” [1991] 1 SLR(R) 354 (HC); Leong Wai Kum, ‘The High Court’s Inherent Power to Grant Declarations of Marital Status’ [1991] SJLS 13.) 

        Certain limbs of s 17 may also be necessary to expand the jurisdiction of the High Court beyond what an originalist reading of Art 93 would give the High Court (see e.g. Emilia Shipping Inc v State Enterprise for Pulp and Paper Industries [1991] 1 SLR(R) 411 (HC)).

        While I am not entirely well-versed in the history of our courts, I note that courts in the Straits Settlements were created by the Second Charter of Justice in 1826 and restructured by the Straits Settlements Courts Ordinance in 1866 (an extract of which is reproduced on page 32 of Leong’s article). I understand that the latter gave the Supreme Court of the Straits Settlements the jurisdiction of the Court of Queen’s Bench, the Court of Chancery, the Court of Common Pleas, and the Court of Exchequer. This was before those four courts and others (e.g. the High Court of Admiralty) were combined into the High Court of England and Wales by the Judicature Acts, 1873-1875. Given this vintage, it is therefore possible that Malaysia’s Courts of Judicature Act 1964 (from which our SCJA was copied) was enacted in order to update the jurisdiction of the superior courts from what they had been fixed at as of 1866.

        But all this is my tentative view only; I will need to do more research to verify its accuracy.

        All that said, I must stress that, though the legislature has the power to pass laws conferring jurisdiction/power on the Supreme Court, the separation of powers still has the important consequence that the court exercises this jurisdiction/power without interference of the other branches of the state.  Moreover, I do not think that it is open to the legislature to pass a law reducing the jurisdiction/powers of the Supreme Court/its components below what is required by the very labels “Supreme Court”, “High Court”, etc. in the Constitution [edit: this minimum ‘level’ includes the power to engage in judicial review of legislation and executive action]. For a Malaysian case where the legislature attempted to do this (by constitutional amendment no less) and failed, see the judgment of the Federal Court in Semenyih Jaya.

      33. Thoughts on Dow Jones v AG [1989] 1 SLR(R) 637 (CA)

        For the purposes of the law on procedural fairness, when does one have a right to a hearing? In Dow Jones, at [56], the Court of Appeal cited a passage from Lloyd v McMahon which used the phrase “what the requirements of fairness demand when any body… has to make a decision which will affect the rights of individuals” (emphasis added). At [59], the court then said, “The appellants have no right but simply a privilege to sell the AWSJ in Singapore”, and therefore they had no right to a hearing.

        Of course, the Dow Jones company indeed did not have a right under Article 14(1)(a). But is it really true that one only has a right to a hearing if the thing that the respondent is threatening to take away is a “right”, but not if the respondent is threatening to take away a “privilege”? Bank Mellat says: no; it is enough if the applicant would be “significantly detrimentally affected”. Kioa v West says: no; it is enough if the respondent would be affecting the applicant’s “rights, interests [or] legitimate expectations”. And Kay Swee Pin did not insist that the applicant have a strict “right” at stake; at [7] of that case, the court spoke of the applicant being “directly affected by the outcome of the disciplinary proceedings”. In none of these cases did the courts say that one does not have a right to a hearing if all that is at stake is a mere “privilege”. Moreover, Dow Jones does not itself explain what a “right” means.

        It is true that, at [58], the Court of Appeal in Dow Jones suggested that one has a right to be heard if one has had a legitimate expectation that one would be heard. But the Court of Appeal suggested this would only take place when one had been promised that one would be heard. This is much narrower than the approach taken in GCHQ, in which the House of Lords said that one can have a “legitimate expectation that [the respondent] would consult [one]” that arises, not from a promise that one would be heard, but rather from one’s “legitimate expectation of receiving [a] benefit or privilege” — and that is “even where [the] person claiming [the] benefit or privilege has no legal right to it”.

        Dow Jones is therefore an outlier in claiming that one has no right to a hearing before being deprived of something that is not a “right” but instead “simply a privilege”. I would suggest that Dow Jones is wrong on this point (or, at least, should not be followed now), not only because it goes against the trend of authority, but also, taken to its logical conclusion, would lead to the result that hardly anybody would ever have a right to a hearing.

        [Edit: Now, see Lim Siew Mui v SkillsFuture Singapore Agency [2025] SGHC 235.]

      34. ‘Judicial review’ of decisions of social clubs, etc.

        I think the neatest analysis is that we shouldn’t call Kay Swee Pin v Singapore Island Country Club a judicial review case, but rather a claim for breach of contract. (This is notwithstanding the Court of Appeal’s use of the phrase “judicial review” at [44] and the headnote in the SLR.)

        This is the analysis adopted by Teo Guan Siew AR, who had to assess the damages which the Court of Appeal had ordered to be paid (note this: damages are not actually a remedy available in judicial review proceedings).

        At [2008] SGHC 143, [31], Teo AR explained that “this case is not about judicial review in the sense of an application under O 53 for the review of a public body’s decision and to obtain public law remedies like certiorari and mandamus”; rather, “Mdm Kay’s claim is clearly in the form of a private law action, which in the circumstances must quite inevitably be characterised as one for breach of contract although this was not expressly stated in the originating summons”.

        At [33], he went on: “The principles of administrative law that were applied by the Court of Appeal, such as the audi alteram partem principle and those relating to the irrationality of the decision-making process, can therefore be seen as being applied by analogy to a private law context for the purpose of deciphering whether there had indeed been a breach of contract. Properly conceived, the claim for damages hence arises from a contractual breach, and it is therefore to the contractual principles of damages that recourse should be had in assessing the quantum of damages that is payable in this case.”

        Subsequently, Navin Anand AR applied the same analysis in Shepherdson, Terence Christopher v Singapore Recreation Club [2019] SGHCR 5, [7].

        But as Calvin Liang rightly points out, this analysis only works if there is a contract between the parties in the first place. What about if, say, the plaintiff’s complaint is that the club has refused to admit him as a member in the first place?

        One solution, says Liang, is that the courts can find some kind of implied contract. (E.g., if there is an application fee, then the court may say that, in consideration of the applicant’s paying the fee, the club has agreed to process the application for membership in accordance with the rules of natural justice, etc.)

        But if that doesn’t work, then Liang points to Nagle v Feilden, a Lord Denning case which asserts that the court has the power to assert “supervisory jurisdiction” over the conduct of the Jockey Club. As Timothy Endicott, Administrative Law puts it (at 606), “Lord Denning excavated the ancient law on the right to work”, e.g. Tailors of Ipswich (1614) 11 Co Rep 53a, 53b, which said that “the law abhors idleness” and “the common law abhors all monopolies, which prohibit any from working in any lawful trade”. The Singapore courts have not yet explored this line of thought. Personally, I would regard it with a great deal of caution: if it is accepted, what limits will there be on the courts’ power to poke their noses into private affairs — contrary to cases like Datafin and Yeap Wai Kong, which, as you say, demand that the applicant show that the respondent’s power is of a public nature?

      35. Notes in response to Chan Sek Keong SC’s ‘Equal Justice Under the Constitution and Section 377A of the Penal Code: The Roads Not Taken’

        I wrote this note in response to a student who had some questions about Chan Sek Keong SC, ‘Equal Justice Under the Constitution and Section 377A of the Penal Code: The Roads Not Taken‘ (2019) 31 SAcLJ 773.


        ‘Legitimate state interest’

        You rightly note that Chan has applied a “legitimate state interest” test. At [49(e)] and [76(d)], Chan asserts that a “legislative classification (or differentia)” must be “reasonable in that it serves or advances a legitimate state interest”. 

        To my mind, this test is similar to the ‘legitimacy of purpose’ test that Quentin Loh J attempted to introduce in the High Court’s decision in Lim Meng Suang. I note that Chan appears not to have engaged either with this aspect of the High Court’s judgment nor the reasons why the Court of Appeal rejected it. Therefore, with respect, Chan has not defended the assumption that the court is well-placed to judge the legitimacy of a state interest. Chan purports to draw support for the ‘legitimate state interest’ test from the Court of Appeal’s expanded “intelligible differentia” test. So either Chan thinks that the ‘legitimate state interest’ test is in substance the same as the expanded ‘intelligible differentia’ test, or Chan has not engaged sufficiently with the Court of Appeal’s arguments. 

        (By the way, at [76(d)], it looks as though Chan has overlooked the fact that Loh J attempted to introduce the ‘legitimacy of purpose’ test. But this does not matter to Chan’s thesis. Chan’s criticism at [76(d)] appears to be aimed primarily at Loh J’s findings as to what the purpose of s 377A was and the reasonableness of that purpose.)

        At [49(f)], Chan provides examples of what, in his view, would not be a legitimate state interest to pursue.  At [84], as I understand it, Chan is saying that ‘to prevent public health problems’ is a legitimate state interest, but ‘to prevent unseemly/undignified behaviour’ is not. I note that Chan has not purported to provide a methodology by which to determine what state interests are legitimate.

        At [83] and [87], Chan moves to making another point. This point is that there is a presumption that there is no ‘legitimate state interest’ in, by way of a criminal law, only persons of one sex from doing something, when in fact persons of any sex are physically capable of doing it. We may read this as either saying that such a law would presumably never serve a ‘legitimate state interest’, or that the differentia behind such a law would be ‘unintelligible’ in the expanded sense. Either way, Chan has not sought to justify this assertion.

        ‘Equality before the law’ vs. ‘equal protection of the law’

        As I understand it, in saying that “males and females” have “equality before the law” [105]-[106], Chan is asserting that there can be no legitimate state interest for criminalising an act when it is done by a man but not when it is done by a woman. At [106], Chan appears to raise the question of whether there can ever be a legitimate state interest for criminalising an act when it is done by a man with a certain sexual orientation, but not when it is done by others. But Chan does not answer this question.

        Other issues

        I agree with Chan that Art 12(1) cannot be treated as merely “aspirational”.  To my mind, if it were possible for any constitutional provisions (at least in Part IV) to be merely “aspirational”, then it would be wholly unclear which provisions are “aspirational” and which are more (and why).

        At [79], Chan suggests that a statute can lose a purpose over time: “if the social conditions that caused [the] enactment [of s 377A of the Penal Code] in 1938 no longer subsisted in 2013… there would no longer exist a purpose…” This is interesting, but, on its face, it would appear to contradict the point at [50(a)] that “[t]he purpose of any law is determined and fixed at the time of enactment of the law… and… will remain the same forever”. If Chan wishes to say that s 377A has no more purpose, then I respectfully think that the stronger arguments to this effect are those set out at [50(e)].

        The other comments which are to be found in the various sub-paragraphs of [49], [50], [55], and [76] are interesting, but their content is largely covered in either the other materials on the reading list or in what I was planning to say in class. The same applies to the remarks on the presumption of constitutionality at [112]-[125], as well as the remarks on the scope of Art 162 at [126]-[132].

        Conclusion

        In noting that Chan has made various assumptions which he has not sought to justify in detail, I am not saying that I disagree with these assumptions. There might even be moral arguments for saying that these assumptions are self-evident. But given that the Court of Appeal has very much rejected these assumptions (or similar ones), I think something more than Chan’s reliance on these assumptions would be needed to persuade the Court of Appeal that its approach in Lim Meng Suang is wrong. Perhaps more persuasive arguments are those set out at [49], [50], [55], and [76], as well as his arguments at [8]-[44] as to the true purpose of s 377A.

      36. What role do rights provisions in the Constitution play in judicial reasoning?

        First, the court has a “duty to declare invalid any exercise of power, legislative and executive, which exceeds the limits of the power conferred by the Constitution, or which contravenes any prohibition which the Constitution provides”: Chan Hiang Leng Colin v Public Prosecutor [1994] 3 SLR(R) 209 at [50].

        Second, the Constitution can serve as a source of norms used by the courts in deciding how to develop the common law. In Review Publishing Co Ltd v Lee Hsien Loong [2012] 1 SLR 52, one question was whether the constitutional right to freedom of expression was an impetus for the common law of defamation to recognise the ‘responsible journalism’ defence, also known as “Reynolds privilege”. (The court did not provide a settled answer to the question because it was not necessary to decide the case: the defendants in that case were not Singapore citizens, and hence had no constitutional right to the freedom of expression.)

        Third, the Constitution can serve as a source of norms used by the courts in deciding how to interpret a statute. Consider Public Prosecutor v Yue Mun Yew Gary [2013] 1 SLR 39, which involved section 267C of the Penal Code:

        “Whoever —

        (a) makes, prints, possesses, posts, distributes or has under his control any document; or

        (b) makes or communicates any electronic record,

        containing any incitement to violence or counselling disobedience to the law or to any lawful order of a public servant or likely to lead to any breach of the peace shall be punished with imprisonment for a term which may extend to 5 years, or with fine, or with both.”

        The High Court interpreted section 267C as containing an implied rule that a person could only be convicted if he intended to incite violence (etc.) The reason for this — refusing to hold that a person could be convicted regardless of his intentions — was “the need to give sufficient weight to the value of free expression” (at [38]). 

        Note that the judgment in Gary Yue spoke of the freedom of expression, but did not mention Article 14. This illustrates the following broader point: Independent of the Constitution, the common law does recognise certain rights — such as the right to freedom of expression and the right of access to courts. These common-law constitutional rights do not form a basis on which the court can strike down legislation. However, they can play the following roles (which are especially important in the UK, where the doctrine of parliamentary sovereignty means that the court cannot declare an Act of Parliament void on any ground or otherwise disapply an Act):

        1. When a statute is ambiguous and capable of being interpreted in one of two ways, the court will choose the interpretation that better gives effect to these rights.
        2. The court will be slow to hold that a statute confers on an executive body the power to derogate from these rights, unless the statute clearly and explicitly confers such a power. The court will be slow to hold that such a power is merely implied by statute. Hence, in R v Lord Chancellor, ex parte Witham [1998] 1 QB 575, the High Court of England and Wales held that a statutory power allowing the Lord Chancellor to set court fees did not empower him, as Lord Justice Rose put it, “totally to preclude the poor from access to the courts”. (If you’re interested in a more recent English case on the same issue, consider this blog post: https://publiclawforeveryone.com/2017/07/26/unison-in-the-supreme-court-employment-fees-constitutional-rights-and-the-rule-of-law/)
        3. One ground of judicial review of executive action is that an executive officer, when making a discretionary decision, failed to take into account a relevant matter (the term more frequently heard is “relevant consideration”) which he/she was bound to take into account in the course of the decision-making process. The court may hold that a common-law constitutional right is such a consideration. 

        These three points apply a fortiori to rights explicitly set out in the Constitution. For an example of the third, consider the Vijaya Kumar case: at [38], the High Court (implicitly) stated that the police, when making licensing decisions, are bound to have regard to the applicant’s right to freedom of religion.

      37. Crime – miscellaneous notes

        Here are some miscellaneous notes I wrote in response to a class on crime; I can’t remember precisely what we discussed.


        Criminal law involves state authorities peering intimately into others’ lives. Witnesses are questioned about whether they can trust their memories of what they think they saw. Victims are asked to narrate, in excruciating detail, traumatic experiences. And offenders’ lives (and those of their families) are put under a microscope as courts decide how they will respond to rehabilitative efforts. None of this is pleasant business.

        One might retort that such intrusion is justified on the ground that the prevention of crime is a good thing. But why treat ‘crime’ as a single conceptual category? Don’t many people routinely cross the road at a spot 50 metres of a designated pedestrian crossing, (contrary to rule 3(1) of the Road Traffic (Pedestrian Crossings) Rules? Did you know that it is a criminal offence to bring beer or wine into a public hospital (section 23 of the Miscellaneous Offences (Public Order and Nuisance) Act)? How many of you have, in your youth, engaged in “disorderly… behaviour in any public road or in any public place”, contrary to section 20 of the Miscellaneous Offences (Public Order and Nuisance) Act? Can it be said that the very act of smoking a cigarette is an offence under section 278 of the Penal Code? Should the criminal process be set in motion in each of these cases?

        One might sometimes get the impression that the law treats crime as though it were a single phenomenon which warrants a single stock set of responses — arrest, investigations, charge, trial, conviction, sentence. But would that be right? The Public Prosecutor himself does not think so, which is why he does not direct that every person who commits an offence under s 377A of the Penal Code be prosecuted despite the fact that he would, in the eyes of the law, be perfectly justified in doing so. 

        To the contrary, one may ask: if the law has singled out a particular type of conduct for condemnation, then are we not justified in condemning it? But that would be to treat the law as a perfect moral leader of society, ignoring the fact that society (or parts of society), too, shapes the law. Besides, condemnation is only one end which criminal law purports to pursue. The law aims both to condemn and prevent wrongdoing that harms others and to help wrongdoers to stop such wrongdoing. But, putting aside how simplistic it would be to think that there exists a single notion of ‘wrongdoing’, the cases we have studied show that these two aims are often in mutual tension. 

        This legal difficulty reflects a deeper moral difficulty. Why did some of us in class today respond with much more revulsion to the offences committed by Lim Choon Hong and Chong Sui Foon compared to those of Abdul Qayyum bin Abdul Razak? (This is not to say that we would not have been justified in doing so, but it is worth reflecting on our motives and underlying assumptions.) Why would we be more concerned about procedural fairness (in terms of not being dragged through the criminal process for too long) to, say, a poor individual accused of being a drug mule (which was the hypothetical example I raised) than about fairness to Lim and Chong? (It may be that it is difficult to sympathise with Lim and Chong given that we know what they are guilty of. But what if such concerns of procedural fairness arise at a stage when the court cannot yet tell conclusively whether the accused person is guilty or innocent?)

        One reason might be that we have certain notions of what conduct is more deserving of condemnation than other conduct. Another possibility is that we have certain notions of which people are more deserving of condemnation than other people. (I hope this latter possibility is not true; but, by its nature, bias against persons operates quietly and insidiously.) Of course, those in power ought to focus on the former, not the latter; and what a challenge that must be! One function of the law — from step-by-step frameworks such as that set out in Boaz Koh, to tables and charts such as those we see in Takaaki Masui and Kunasekaran (the latter of which was applied in Terence Siow) — is to enable those in power to exercise their power in an even-handed, dispassionate manner. if this is true, then in what sense can it can be said that the law reflects/should reflect the views of society, which are often neither even-handed nor dispassionate?

      38. Juries

        If you’re interested in why Singapore doesn’t have a jury system in Singapore, consider:

        1. Andrew Phang Boon Leong, ‘Jury Trial in Singapore and Malaysia: The Unmaking of a Legal Institution’ (1983) 25 Malaya Law Review 50. (The writer, who is now a judge, used to be an academic specialising in contract law as well as Singapore’s legal history.)

        2. Debates on Parliament on the Criminal Procedure Code (Amendment) Bill 1969, which abolished the jury system.

        3. The report of the Select Committee on the Criminal Procedure Code (Amendment) Bill 1969. In particular, the transcripts in section B make very interesting reading.

      39. A few extracts from cases on sexual crimes generally

        I find the following passages from GHI v Public Prosecutor [2024] SGHC 220 noteworthy. (I must warn you that the full judgment necessarily goes into some detail about the facts of the case, in which a child had been molested.)

        • At [35]: “Victims of sexual crimes cannot be straitjacketed in the expectation that they must act or react in a certain manner (GBR v Public Prosecutor and another appeal [2018] 3 SLR 1048 (“GBR”) at [20]). There is no general rule requiring victims of sexual offences to report such offences immediately or in a timely fashion. In particular, young victims of sexual assault may not report offences in a timely manner for various reasons including feelings of shame and fear. Delay in reporting is not, on its own, a reason to disbelieve a victim. The court should consider the reasons for the delay in reporting the offences to the police or to anyone else (Public Prosecutor v Mohd Ariffan bin Mohd Hassan [2019] 2 SLR 490 at [65]–[68]).”
        • At [69], on questioning witnesses: “…  I noted that the victim in the present case was questioned on her attire at the time of the molest. The practice of asking victims about their clothing at the time of the alleged offence is acceptable if this sheds light on how the offence was committed. For example, the line of questioning may confirm whether the touching was above or under clothes, whether there was skin-on-skin contact and/or how there came to be such skin-on-skin contact. Such questions are necessary to provide the court with the proper context in which the offence was committed. However, the enquiry becomes objectionable when it is premised on, or leads to, the submission that the victim’s attire had in some way, invited the sexual assault. For example, any line of questioning which invites the implication that the victim had encouraged unwanted attention because she was dressed provocatively, must be rejected.”
        • And at [71]: “… I emphasise that the court will not tolerate a line of inquiry regarding the victim’s attire when its implication is such that the victim invited the sexual assault.”
        • And again at [73]: “It should be borne in mind that the purpose of cross-examination is not to cause unnecessary discomfort to, harass or abuse a witness. In cases of sexual offences especially, unwarranted questioning of the victim’s credibility, delving into irrelevant personal history or insinuating blame can not only re-traumatise the victim but also perpetuate harmful stereotypes about sexual violence. This approach can dissuade other victims from coming forward for fear of being subjected to a similar ordeal.”

        The court also quoted rule 3 of the Evidence (Restrictions on Questions and Evidence in Criminal Proceedings) Rules 2018:

        “Restrictions on questions and evidence in criminal proceedings involving sexual offence or child abuse offence

        3. In criminal proceedings where the accused is charged with committing a sexual offence or child abuse offence, the following apply:

        (a) except with the permission of the court, no question may be asked of the alleged victim of the offence, during cross‑examination by or on behalf of the accused, about the alleged victim’s sexual behaviour or physical appearance;

        (b) except with the permission of the court, no evidence may be adduced by or on behalf of the accused about the alleged victim’s sexual behaviour or physical appearance.”

        In Public Prosecutor v Mohd Ariffan bin Mohd Hassan, the Court of Appeal (at [63]) accepted the expert evidence of Ms Ng Pei Yu, Vivienne, the Chief Psychologist at the Ministry of Social and Family Development, and decided as follows. (Again, the full judgment sets out the details of a case in which a young person was sexually assaulted.)

        “63 In relation to the issue of delay in reporting, we accept the expert opinion of Ms Ng, which was undisputed by the respondent, that only a small proportion of victims of sexual offences report the offences in a timely manner. Ms Ng stated in her expert report that:

        5.3.4 Very few victims report immediately to law enforcement, but if they do report to law enforcement, it is often after a delay of days, weeks, months, or even years (please see Section 7 on difficulties in disclosure and delay in reporting).

        …

        7.4 The shock, shame, and stigma attached to being a victim of crime make it difficult for even adults to report victimization. It is not surprising that sexual assault has been found to be the most under-reported violent crime to any authority by adults and teenagers (Ciarlante, 2007). Victims often delay in reporting sexual victimization and face negative social reactions to disclosure (Dworkin et al., 2017; Fanflik, 2007). For children, there are additional factors that affect their willingness to disclose sexual abuse. …

        [emphasis in original]

        64 Ms Ng also explained in her report that the likelihood of delayed disclosure is greater in cases where the sexual assault was committed by an individual known to the victim. There could be a number of reasons for delayed disclosure by such victims, as Ms Ng explained:

        7.7 Fear of vengeance and feelings of guilt and shame are the most important causes for suspending disclosure (Sauzier, 1989). Victims of intra-familial rape may have to continue to live in proximity with the perpetrator and hence fear punishment or retaliation if they disclose the assault. In the majority of cases, the absolute dominance of the offender upon the victim and the obedience of the latter, via complete submission and passivity, generate in the victim a feeling of “conspiracy” and complicity. Because the perpetrator’s strategies are oriented towards the maintenance of the child’s compliance and silence, the child is often inhibited to disclose abuse. As the child’s relationship with the perpetrator is often an emotionally significant one, many victims report ambivalent feelings for their perpetrators and do not disclose the abuse. Defining oneself as the victim of a family member’s assault requires significant and painful alterations in the victim’s perceptions of the perpetrator and the role that he/she was expected to demonstrate …

        7.8 Victims of intra-familial assault often have a personal, financial, legal, or social relationship with their abusers that may make them dependent on their abusers. Family or social considerations such as loyalty to family members, wanting to protect the privacy of the family, fear of not being believed, fear of causing family disruption or bringing shame to the family, not [wanting] negative consequences for loved ones when they disclose etc. have a strong impact on the victim’s decision on whether to disclose the offence and how the disclosure is made. Some of these themes are consistent with local researchers’ views that when there is a relationship between the rape victim and the offender, the victim is more likely to experience a sense of betrayal and may come under suspicion or criticism about whether she had consented to having sexual intercourse with the perpetrator (Lim, Chan, Chan et al.,2002). …

        65 We accept that a victim of sexual assault, especially a youthful one assaulted in a familial context, may not report the offence in a timely manner as there are empirically-supported psychological reasons for delayed reporting, including feelings of shame and fear. With respect, we must reject the Judge’s suggestion (see [40]–[41] of the GD) that with the passage of time, a victim would have recovered from distress or embarrassment and would have no difficulty in disclosing the offences and recounting the abuse that he or she was subjected to.

        66 Therefore, that there is a delay in reporting by a complainant is not, on  its own, reason to disbelieve the complainant and his or her allegations against an accused person. In this regard, we affirm the legal principles set out in past cases on how delay in reporting by a complainant should be treated by a court. These principles, in our view, give due regard to the likely thought-processes and behaviour of sexual assault victims as highlighted above.

        67 In DT v Public Prosecutor [2001] 2 SLR(R) 583, the High Court stated (at [62]) that there is no general rule requiring victims of sexual offences to report the offences to the police immediately. The court explained that, instead, the explanations proffered by the complainant for his or her delay in reporting the offences to the police are to be considered by the court in determining the impact of the delay, if any, on the credibility of the complainant. We would add that the requirement of examining the reasons proffered by the complainant applies not only to the complainant’s delay in reporting the offences to the police, but also to any delay in disclosing the assault to anyone else, such as to his or her family members.

        68 The above principles on the significance of delayed reporting by victims of sexual offences were also adopted by the High Court in the recent case of Public Prosecutor v Yue Roger Jr [2019] 3 SLR 749 (“Roger Yue (HC)”), which was a decision that we affirmed on appeal in Yue Roger Jr v Public Prosecutor [2019] 1 SLR 829. In Roger Yue (HC), Aedit Abdullah J found that the complainant’s failure to report the offences until five years later did not make her evidence less credible as there were reasonable explanations provided for her conduct. Aedit Abdullah J stated (at [30]):

        I accepted that victims of sexual offences may not behave in a stereotypical way. Many victims report their sexual abuse early to a family member, friend, the police, or other person in authority. However, there is no general rule requiring victims of sexual offences to report the offences immediately or in a timely fashion. Instead, the explanation for any such delay in reporting is to be considered and assessed by the court on a case-by-case basis (see DT v PP [2001] 2 SLR(R) 583 at [62]; Tang Kin Seng v PP [1996] 3 SLR(R) 444 at [79]). While I accept that an omission to report the offence in a timely fashion, in the absence of other evidence, may in certain circumstances make it difficult to establish a case against the accused beyond reasonable doubt, I emphasise that the effect of any delay in reporting always falls to be assessed on the specific facts of each individual case.

        We endorse the learned judge’s remarks in Roger Yue (HC).

        69 Having assessed the evidence given by the complainant, we are satisfied that if we were to accept her primary allegations against the respondent, then there were plausible legitimate explanations for the complainant’s initial non-disclosure of the offences. Three reasons for the complainant’s initial non-disclosure of the offences can be discerned from her testimony.

        70  First, she was afraid of the respondent. She testified that she had been warned multiple times by the respondent not to speak to anyone about the acts. The respondent had also allegedly told her about how he had previously participated in a gang-rape, and the complainant was fearful that she would be subjected to the same treatment should she disobey the respondent’s instructions to remain silent.

        71 Second, the complainant explained that she had not disclosed the offences earlier to her family members as she did not want to spoil the relationship between the respondent and her mother, especially since the respondent (who she treated as a father figure) had promised to marry her mother. She also did not want to disclose the matter to the brother as she feared that he might do something untoward to the respondent as he was, according to her, a “very hot-tempered person”.

        72 Third, the complainant explained that she was reluctant to lodge a police report as she was concerned over what she might have to go through at the trial, if the offences were reported.

        73 We are satisfied that in this case, the delay in the complainant’s disclosure and her reluctance to report the matter to the police were based on reasons that were sound and credible from her perspective, and were in fact aligned with the expert evidence on the thought-processes and behaviour of many victims of sexual offences. Therefore, we find that the complainant’s delayed disclosure to her loved ones and her reluctance to make a police report, did not, taken in isolation, undermine her credibility.”

      40. “[I]nsult[ing] the modesty of any woman” etc. – language matters

        Section 509 of the Penal Code used to state: “Whoever, intending to insult the modesty of any woman, utters any word, makes any sound or gesture, or exhibits any object, intending that such word or sound shall be heard, or that such gesture or object shall be seen by such woman, or intrudes upon the privacy of such woman, shall be punished with imprisonment for a term which may extend to one year, or with fine, or with both.”

        The courts interpreted section 509 as including all manner of things, including indecent exposure (Ramakrishnan s/o Ramayan v PP [1998] 3 SLR(R) 161), peeping at someone who was showering (Ye Wei Gen v PP [1999] 2 SLR(R) 1074), filming people using the restroom or changing (PP v Tay Beng Guan Albert [2000] 2 SLR(R) 778), and taking ‘upskirt’ photographs (PP v Ng Jiak Teng [2007] SGDC 115).

        In 2019, the Criminal Law Reform Act replaced s 509 of the Penal Code with section 377BA. This was worded identically, except that “woman” was changed to “person”. Also, various provisions on sexual offences were added to the Penal Code, including ones specifically criminalising voyeurism (section 377BB-377BE), “sexual exposure” (including ‘cyber-flashing’) (section 377BF), and “child abuse material” (sections 377BG-377BL).

        Is it a crime to possess an “intimate image” etc. taken without the subject’s consent, if someone sent it to you without your having asked for it? What about if a child takes a nude picture of themselves? See sections 377BM-377BN.

         I think the 2019 changes can only be a good thing. Whatever “modesty” means, it would be wrong to assume that only women have it. And the new provisions comprehensively cover new forms of sexual offences, not all of which might have been covered by the old law.

        But let’s dig a bit deeper into the language of ‘insulting modesty’. I think this opinion piece by Avantika Tewari explains, much more eloquently than I did in class, what the problem is: https://feminisminindia.com/2017/08/28/modesty-outraged-patriarchy-law/

        I think the following extracts from the Indian Supreme Court case of Rupan Deol Bajaj v KPS Gill [1996] Cri LJ 81 show why the language is potentially problematic:

        • “Modesty” means “womanly propriety of behaviour; scrupulous chastity of thought, speech and conduct”; being “decorous in manner and conduct; not forward or lewd; shamefast”; “freedom from coarseness, indelicacy or indecency; a regard for propriety in dress, speech or conduct”; “womanly propriety of behaviour; scrupulous chastity of thought, speech and conduct (in man or woman); reserve or sense of shame proceeding from instinctive aversion to impure or coarse suggestions”.
        • This includes “any act done to or in the presence of a woman [that] is clearly suggestive of sex according to the common notions of mankind”.
        • The “ultimate test” for outrage of modesty is: whether “the action of the offender [is] such as could be perceived as one which is capable of shocking the sense of decency of a woman”.  It need not have “sexual overtones”: it was still an “affront to the normal sense of feminine decency” and to the “dignity of the lady”.

        In short, the problem is that the language of ‘insulting modesty’ doesn’t accurately reflect what is wrong with (say) molesting someone or taking an ‘upskirt’ photograph. What is wrong with doing so is not that it is ‘improper’, ‘coarse’, or ‘shameful’ for the victim (if anything, it is the offender who should be ashamed!), but rather, that it is an intrusion on personal privacy, bodily integrity, etc. This being so, making the former section 509 gender-neutral did not address all of the problems.

      41. A possible example of how to litigate an Art 12(1) case

        Suppose you had a statute that banned women from driving, and you wanted to challenge it under Art 12(1).

        You would want to persuade the court to adopt the following analysis:

        A

        1. The differentia is between men drivers and women drivers.
        2. This differentia is ‘unintelligible’ in the expanded sense.
        3. So the statute fails the ‘intelligible differentia’ test.

        But not:

        B

        1. The differentia is between men and women.
        2. This differentia is ‘unintelligible’ in the expanded sense.
        3. So the statute fails the ‘intelligible differentia’ test.

        The problem with B is that it will be much harder to show that the differentia between men and women is ‘unintelligible’ in that “no reasonable person would ever contemplate the differentia concerned as being functional as intelligible differentia” (Lim Meng Suang (CA) at [67]; emphasis in original). This is because one can think of plenty of reasons why a law might differentiate between men and women. By contrast, it is much harder to think of a reason why a law might differentiate between men drivers and women drivers, so A is more likely to succeed.

        Alternatively, you could argue:

        C

        1. The differentia is between men drivers and women drivers.
        2. The object is to promote social harmony.
        3. But there is no logical link between treating men drivers and women drivers differently, and promoting social harmony.
        4. So the statute fails the ‘rational nexus’ test.

        But you would want to resist any attempt by the state to argue:

        D

        1. The differentia is between men drivers and women drivers.
        2. The object is to prevent women from driving.
        3. There is a logical link between treating men drivers and women drivers differently, and preventing women from driving. (In fact, there is a complete coincidence between the differentia and the object.)
        4. So the statute passes the ‘rational nexus’ test.
      42. (Potentially) unconstitutional legislation: what does Tan Eng Hong say about Art 4 and Art 162 of the Constitution?

        The passage in Tan Eng Hong beginning at [59], which discusses Articles 4 and 162 of the Constitution, is somewhat confusing. One of you expressed confusion as to what the following sentences from [59] mean: “Article 4 provides for the unconstitutional portion of the law to be severed while retaining the remaining part of the law in the statute books. In other words, the court’s power to void laws for inconsistency with the Constitution under Art 4 can be interpreted to include the power to void laws which pre-date the Constitution.” 

        One possibility is that “[i]n other words” means that the court can (for example) use its power under Art 162 to “constru[e]” the entire Penal Code with the “modificatio[n], adaptatio[n], qualificatio[n] [or] exceptio[n]” that it does not include s 377A. But this would be stretching the meanings of the words “modificatio[n], adaptatio[n], qualificatio[n] [or] exceptio[n]”.

        I think there is a simpler way of looking at the matter.

         The way I understand it, Art 4 says two things:

        1. “This Constitution is the supreme law of the Republic of Singapore”
        2. “any law enacted by the Legislature after the commencement of this Constitution which is inconsistent with this Constitution shall, to the extent of the inconsistency, be void”

         All (b) does is to restate one of the consequences of (a), namely, that legislation contrary to the Constitution enacted after 9 August 1965 is void. But there are still other consequences of (a), e.g., that any legislation contrary to the Constitution can be void. If s 377A were unconstitutional, then striking it down would take place pursuant to (a), not (b). In fact, (b) would be irrelevant. So we wouldn’t be able to say that (b) is being contradicted or departed from.

        I draw support for this view from the part of para [60] which says that “Article 4 provides for one of the most important features of the Constitution, viz, that it is the supreme law of Singapore” – the court focuses on point (a); there is no mention of point (b). Further, at [63], the court uses the Wee Commission report in order to make the point that “the supremacy of the Constitution is necessary for the purposes of the Constitution to be protected” – again, there is no reference to (b).

        If we see things this way, then there’s really no need to refer to Art 162 at all. In short, we could simply say that, if a statutory provision is contrary to the Constitution, it is void simply because the Constitution is the “supreme law of the Republic of Singapore”, and “suprem[acy]” implies that anything contrary to that which is “supreme” is of no effect.

        One might then ask: what is the point of (b), if what it says is already encapsulated in (a)? My answer would be that (b) serves to make the point that, while legislation contrary to the Constitution enacted after 9 August 1965 shall be void, other legislation contrary to the Constitution might be void but, because of Art 162, is not necessarily so.


        Art 162 says that “all existing laws shall continue in force… but all such laws shall… be construed… with such modifications [etc.] as may be necessary to bring them into conformity with this Constitution”. To my mind, the Court of Appeal read this as meaning that all such laws must be brought into conformity with the Constitution (as opposed to merely meaning that the courts must try to bring all such laws into conformity with the Constitution). If this is correct, then Art 162 does not say that all such laws are constitutionally valid; it says that they are not, but they must be made so. This would leave a gap: what if they cannot be made so using the means set out in Art 162? The Court of Appeal’s answer was: they are void.

        I think that a more problematic aspect of the Court’s reasoning was its reading of “construed”. Yes, there are limits to construction: for example, the orthodoxy is that construction cannot be rewriting. But Art 162 does not merely say that certain laws must be “construed”; it says that they must be “construed… with… modifications” etc. — so, one would think, Art 162 does authorise (in fact, require) the court to make changes to the statute to make it constitutionally valid. On the other hand, this view is not totally unproblematic: modification is not re-writing. (Can it be said that the statute has a ‘basic structure’ which cannot be changed by way of “modifications”?)

        Perhaps the answer to this is similar to how the UK courts have, under s 3 of the Human Rights Act 1998, “read and given effect” to legislation “in a way which is compatible with the Convention rights”. This includes ‘reading in’ words into the statute / ‘reading down’ the words in the statue: see Ghaidan v Godin-Mendoza (which we will be examining in the last topic of the course). Chan Sek Keong, in his article on s 377A (2019) 31 SAcLJ 773, alludes (though very obliquely) to this. At [128], he says that, under Art 162, “the court cannot rewrite the text of the written law…”, but can “read the law or the words of the law in such a way that the law conforms to the Constitution”. His idea of how this may take place is set out at [132]: “the court must construe s 377A by reading it down, or up, as the case may be, to make it conform to Art 12(1)…”