1 This is our judgment on costs following our dismissal of an appeal against a decision of the High Court which rejected an application by Mr Deepak Sharma (“the Appellant”) for judicial review. The Appellant had sought a quashing order against a decision of a review committee of the Law Society of Singapore (“the Respondent”) on the basis that the review committee’s decision to dismiss the Appellant’s complaint against two lawyers contained errors of law and other deficiencies. The Attorney-General (“the AG”) intervened from the genesis of the proceedings after the Appellant served his ex parte originating summons and supporting documents on the AG as required under O 53 r 1(3) of the Rules of Court (Cap 322, R 5, 2014 Rev Ed) (“the Rules of Court”). The AG continued to participate in the proceedings, making both written and oral submissions on various matters, until the dismissal of the appeal. Having arrived at the close of the proceedings, the AG now seeks his costs both here and below from the Appellant.

2 The AG’s claim against the Appellant for costs raises a novel and difficult question of principle. When the AG intervenes in an application for judicial review that does not involve the Government and/or does not seek to challenge any governmental action or decision (referred to hereafter as “private judicial review”), can the AG subsequently seek an award of costs in his favour as a matter of principle, and if so, in what circumstances and against whom? The answer to this question requires an examination of the role and function of the AG in proceedings of this nature, considered in the light of the general principles and justifications for making an award of party-and-party costs as well as the relevant statutory background.

3 In our judgment, the principled way forward is to apply a calibrated approach that takes into account the nature and importance of the public function that the AG discharges when he intervenes to make submissions in private judicial review proceedings, but ensures at the same time that liability to pay costs is attributed and apportioned in a fair and controlled manner. It should not be forgotten that the AG’s intervention in the public interest takes place in the context of a private dispute that is personal to the parties, where the mounting costs of litigation are never far from mind.

4 The facts of the dispute between the Appellant and the Respondent are set out comprehensively in the judgment of the High Court (Deepak Sharma v Law Society of Singapore [2016] 4 SLR 192 (“the Judgment (HC)”)) and our decision on appeal (Deepak Sharma v Law Society of Singapore [2017] 1 SLR 862 (“the Judgment (CA)”)). We do not propose to recount the background to the dispute in full, but will describe only those facts and findings that are necessary to provide context to this judgment on costs, which is to be read together with those earlier judgments. For consistency, we will use the same short-form references that we used in the Judgment (CA).

5 In 2010, a disciplinary committee of the Singapore Medical Council (“the SMC”) commenced disciplinary proceedings against Dr Lim Mey Lee Susan (“Dr Lim”), who is the Appellant’s wife. The disciplinary committee recused itself upon an unopposed application by Dr Lim’s counsel, and the SMC subsequently decided to appoint a second disciplinary committee. Dr Lim filed an application for leave to apply for a quashing order against the SMC’s decision, as well as an application for a declaration that the SMC had no legal right to adduce certain confidential medical records in evidence. The second application was eventually withdrawn and the first was dismissed by the High Court. Dr Lim appealed against the dismissal of the first application, but her appeal was unsuccessful.

6 The SMC’s lawyers, WongPartnership LLP (“WP”), then sought costs against Dr Lim for her two failed applications and the single unsuccessful appeal. The SMC’s bills of costs were taxed down by an assistant registrar and the SMC applied for a taxation review. During the taxation review, the SMC’s lawyers reduced the SMC’s claim for costs by a substantial amount, explaining that the reduction was necessary to take into account an overlap between lawyers and re-getting up by new lawyers who later joined the team. In the result, the High Court judge (“the Judge”) increased the costs awarded to the SMC from the amount allowed by the assistant registrar.

7 Following the taxation review, the Appellant sent a letter of complaint to the chairman of the Complaints Panel of the Respondent (ie, the Law Society of Singapore). The Appellant alleged that two of the SMC’s lawyers, Mr Yeo Khirn Hai Alvin SC (“Mr Yeo”) and Ms Ho Pei Shien Melanie (“Ms Ho”), were guilty of “gross overcharging” by submitting bills of costs which were “clearly exorbitant and which … would amount to grossly improper conduct and/or conduct unbecoming as members of an honourable profession”. A review committee (“the RC”) was constituted to consider the complaint. On 10 April 2014, the Respondent sent a letter (“the Decision Letter”) to the Appellant informing him of the RC’s decision. In summary, the RC dismissed the complaint against Mr Yeo in its entirety, but decided that part of the complaint against Ms Ho should be referred to the chairman of the Inquiry Panel to constitute an Inquiry Committee for further inquiry. The RC made two primary findings. First, it considered that “a reduction by the taxing master (Registrar or Judge of the Supreme Court) of the costs claimed[,] even if significant, would not amount to misconduct in the absence of improper or fraudulent claims”. As there was no improper or fraudulent conduct, there was no substance to the complaint that the excessive claim for costs in itself warranted a finding of misconduct. Second, the RC decided that the effective hourly rate reflected in the bills of costs was not excessive because “the amounts in the Bills of Costs reflect[ed] the work of all the solicitors involved”. It found that Mr Yeo was not involved in the preparation of the bills of costs, but that this part of the complaint against Ms Ho was not frivolous, vexatious, misconceived or lacking in substance, and that it would therefore refer this part of the complaint against Ms Ho to the chairman of the Inquiry Panel.

8 The Appellant filed an application for leave to commence judicial review proceedings, seeking to quash the RC’s decision and have his complaint considered by a freshly-constituted review committee. The Appellant raised three grounds in support of his application. First, the RC erred in law in concluding that professional misconduct through “gross overcharging” could not be established “in the absence of other impropriety”. Second, it erred in law in deciding that WP’s pursuit of the amounts in the bills of costs could not constitute misconduct given that they reflected the work of all the lawyers involved. Third, the RC should not have relied on WP’s purported clarification to find that Mr Yeo was not involved in drawing up the bills of costs or the taxation.

9 The AG intervened and participated in the hearing before the Judge (who also heard the taxation review (see [6] above)). The Judge decided two preliminary issues in favour of the Appellant. First, contrary to the Respondent’s argument in this regard, the RC’s decision was susceptible to judicial review. Second, the Appellant did not have to demonstrate that he had locus standi to make a complaint to the Respondent as there was no such requirement. The AG had argued that the Appellant must show that he had locus standi, which he did not possess: see the Judgment (HC) at [54]. The AG took the position that since the Appellant’s complaint concerned WP’s conduct in the taxation proceedings and the Appellant was a “stranger” to those proceedings (unlike Dr Lim herself), he was not in a position to make the complaint.

10 Notably, the Respondent was in agreement with the Appellant that no locus standi requirement existed: see the Judgment (HC) at [53]. This meant that the only party who took an objection to the standing of the Appellant was the AG. After considering the relevant provisions of the Legal Profession Act (Cap 161, 2009 Rev Ed) (“the LPA”) and its legislative history, the Judge found that Parliament had intended that any person could make a complaint against a lawyer to the Respondent, and thus rejected the AG’s argument. We observe that the parties’ arguments on these two preliminary issues took up a considerable portion of the hearing before the Judge; indeed, the Judge devoted no fewer than 60 paragraphs of his written grounds to these matters: see the Judgment (HC) at [22]–[81].

11 In so far as the merits of the Appellant’s allegations against the RC’s decision were concerned, the Judge held that a significant reduction in costs on taxation, without more, would not ordinarily mean that there was gross over-claiming amounting to misconduct, although he accepted that there were certain cases where gross over-claiming alone could constitute misconduct: see the Judgment (HC) at [101]–[104]. He found at [110] that on a proper interpretation of the Decision Letter, the RC had not misstated the correct position in law. On the Appellant’s argument regarding the effective hourly rate of the SMC’s lawyers, the Judge held at [124] that what the RC had meant was that the amounts in WP’s bills of costs reflected not only the hours spent by Mr Yeo and Ms Ho, but also those expended by other lawyers involved in the matter. He explained at [125] that it was not a breach of O 59 r 19 of the Rules of Court (“O 59 r 19”) for a litigant’s bill of costs to contain the number of hours spent by each lawyer who worked on the matter even though there were more than two lawyers involved, although care had to be taken to ascertain “overlapping work and the extent of the overlap”. The RC had referred, albeit in shorthand, to this principle, and it therefore had not erred in law. Finally, on the Appellant’s third ground, the Judge found that although the RC should not have relied on WP’s purported clarification that Mr Yeo was not involved in the taxation, there was nevertheless no reason for the court to refer this part of the complaint back to the RC because the RC had found that there was no misconduct on Mr Yeo’s part: see the Judgment (HC) at [145] and [173].

12 In the circumstances, the Judge dismissed the Appellant’s application for judicial review.

13 Dissatisfied with the outcome, the Appellant appealed against the Judge’s decision. As mentioned at [1] above, the AG participated in the appeal as well, advancing both written as well as oral submissions before this court.

14 The issues on appeal were somewhat different from those before the Judge. First – and notably – the two preliminary issues were not pursued on appeal by any of the parties. This included the issue of locus standi, which the AG had argued with such vigour in the court below. Second, the Appellant no longer took issue with the RC’s conclusion on the hourly rate of WP’s solicitors. Rather, the focus of his argument regarding O 59 r 19 was that the RC had erred or misdirected itself in failing to take into account WP’s acknowledgment at the taxation review that it had included a substantial amount of duplicated costs in its claim, which was a breach of O 59 r 19. Apart from this, the Appellant’s two other primary submissions on appeal were similar to those which he made before the Judge.

15 The Appellant failed on all three submissions before us. We held at [47] of the Judgment (CA) that given the adversarial nature of the taxation of party-and-party costs, an excessive claim for costs would not, in and of itself, generally constitute professional misconduct, absent proof that the lawyer had put forward either an improper or a fraudulent claim. Having said that, a lawyer would be found to have made an improper claim where the quantum of costs sought was so astronomical as to be so disproportionate and unjustifiable that the making of the claim would itself constitute professional misconduct – although such a claim would, in practice, be extremely rare. The Judge (and the RC) had applied precisely the same test as that just stated, and therefore had not made any error in law. We also rejected the Appellant’s claim that the RC had failed to take into account WP’s voluntary reduction of its claim for costs. There was simply no evidence to support his assertion that the RC had ignored or disregarded that fact; in fact, the RC had concluded that WP had not made any improper claims, and in coming to this conclusion it must have decided that there was no impropriety in respect of WP’s voluntary reduction: see the Judgment (CA) at [58]–[64]. Given the Appellant’s failure to show any deficiency in the RC’s decision that Mr Yeo was not guilty of improper conduct, there was no basis for the Appellant’s third and final argument regarding the RC’s erroneous reliance on WP’s clarification that Mr Yeo was uninvolved in the taxation. Thus, that argument also fell away.

16 Having dismissed the appeal, we directed the parties to file written submissions on costs. The AG sought full costs and disbursements from the Appellant for his involvement both in the High Court and on appeal. The Appellant denied the AG’s entitlement to any costs, arguing that he should not be made to bear the AG’s costs “as a matter of principle”. We then directed the parties to make further written and oral submissions on the issue of whether the AG is entitled to costs when he intervenes in private judicial review proceedings after being served with an ex parte originating summons pursuant to O 53 r 1(3) of the Rules of Court (“O 53 r 1(3)”). We will briefly summarise those submissions.

17 The AG takes the position that where he participates in private judicial review proceedings as the “guardian of the public interest” and the court agrees with his position on whether the application for judicial review should be allowed or dismissed, he is entitled to costs. If the AG’s participation in such proceedings is, however, not aimed at securing the public interest, then he should not be entitled to costs. But any claim that the AG’s participation in the proceedings is not to secure the public interest should be made promptly. If the parties agree or do not raise any objections to the AG’s participation, or an objection is raised but is dismissed by the court, this should be taken as a strong indication that the AG’s participation in the proceedings is warranted to secure the public interest.

18 Where the AG participates in private judicial review proceedings to secure the public interest, but makes one or more submissions: (a) on issues that do not clearly pertain to the public interest; or (b) which are partly or wholly unsuccessful, the fact that he made such submissions should not bar him from obtaining all his costs in the action where the court agrees with his ultimate position on whether the judicial review application should be allowed or dismissed. The AG should be deprived of part of his costs only where he has “unnecessarily or unreasonably protracted or added to the costs or complexity of the proceedings” (citing O 59 r 6A of the Rules of Court). Finally, where the AG participates in private judicial review proceedings as the “guardian of the public interest” and “ends up being aligned with one party or another”, and that party is substantially or wholly unsuccessful in the proceedings, then unless the AG has unnecessarily or unreasonably protracted or added to the costs or complexity of the proceedings, the AG should not be required to pay costs to the successful party.

19 We distil from the AG’s submissions several key reasons for his position that he is entitled, as a matter of principle, to costs when he participates in private judicial review proceedings. First, when the AG appears in such proceedings, he does so as the “guardian of the public interest”. In proceedings of this nature, the AG is “objective and non-partisan” in deciding whether to appear and participate in the public interest. Whether the AG ends up being aligned with one side or the other is simply a factual result based on the AG’s assessment of where the public interest lies. When acting as the “guardian of the public interest”, the AG should not be equated with “an ordinary litigant who litigates to enforce or protect his private interests”, who should bear costs in the event that his claim or defence is unsuccessful (citing Law Society of Singapore v Top Ten Entertainment Pte Ltd [2011] 2 SLR 1279 (“Top Ten”) at [24]). Second, where the AG “succeeds” in the judicial review application, costs should follow the event, and thus, the AG’s costs should be borne by the unsuccessful party in the litigation. Third, the AG notes that there have been previous cases involving the AG’s intervention in private judicial review proceedings where the unsuccessful party was ordered to pay the AG’s costs. In this regard, several such cases were cited to us (see also below at [23] and [28]).

20 The Appellant seeks to draw a distinction between a hearing on the granting of leave to commence private judicial review proceedings and a hearing on the substantive merits of the judicial review application. His position is that at the leave stage, the AG is entitled to seek costs against the applicant if he participates in the leave proceedings and successfully opposes the applicant’s application for leave. The general rule that costs follow the event should apply. At the merits stage, however, the AG should not be entitled to costs even if the applicant ultimately loses at this stage; conversely, if the AG makes submissions on the public interest which are not accepted by the court, no costs orders should be made against him. The agreement or objection of a party to the AG’s participation, whether at the leave stage or the merits stage, is irrelevant to the question of whether the AG is entitled to costs.

21 The Appellant explains his approach as follows. At the leave stage, if the AG wishes to oppose the applicant’s application for leave and the court permits him to do so, he will “become a statutory party and the usual cost[s] consequences should apply depending on the outcome of the leave stage (whether in favour [of] or against the Attorney-General)” [emphasis in original omitted]. This is “akin to a situation where a judicial review is directed at the Government and the Attorney-General attends and opposes the judicial review”, where it is established that the AG “is a substantive party to the proceedings and can make the usual cost[s] applications” [emphasis in original omitted] (citing Jeyaretnam Kenneth Andrew v Attorney-General [2013] 1 SLR 619 for the lattermost proposition). At the merits stage, however, the AG has “no statutory right or obligation to participate in the proceedings”. In private judicial review proceedings, the AG acts only as an intervener if he wishes to oppose the application on the merits. His position is “similar to that of an amicus curiae – he is merely there to assist the Court on issues of public interest, if any”. As the AG “does not have any substantive interest in the outcome of the application and is seeking to be heard only on matters of public interest”, the principle that costs follow the event does not apply to the AG, who is an intervener and not a “winning party”, and this is the case even if the applicant ultimately loses at the merits stage.

22 The Appellant takes the view that it is intrinsically wrong that a private litigant should be ordered to fund the AG’s efforts to raise public interest considerations. The AG is discharging a public service and thus should not expect to be awarded costs. There are questions as to whether it is “fundamentally fair” to make private litigants liable to pay costs for the AG’s contributions as an intervener. In addition, there are implications on access to justice that would arise from making private litigants liable to pay costs in this manner.

23 Finally, the Appellant observes that in the Singapore High Court’s decisions in Chiu Teng @ Kallang Pte Ltd v Singapore Land Authority [2014] 1 SLR 1047 and AXY & others v Comptroller of Income Tax (Attorney-General, intervener) [2017] SGHC 42, no rationale or principle was articulated by the High Court in coming to its decision to award costs to the AG, who had intervened in the private judicial review proceedings in both cases.

24 As the Respondent correctly observes, the question of whether the AG is entitled to costs when he appears in private judicial review proceedings does not directly relate to it. It therefore suffices for us to briefly state its position. In short, the Respondent agrees with the AG that the latter is acting as the “guardian of the public interest” when he participates in proceedings of this nature. Where the AG’s submissions are wholly or partly unsuccessful but the AG ultimately succeeds in relation to the outcome of the proceedings, he should not be prevented from obtaining costs if his arguments are reasonable. Where the AG aligns himself with a party that is ultimately unsuccessful in the proceedings, the AG will be entitled to rely on the principle laid down in the English Court of Appeal decision of Baxendale-Walker v Law Society [2008] 1 WLR 426 (“Baxendale-Walker”), which states that public bodies performing regulatory functions should be protected from having to pay costs unless they are proved to have acted in bad faith or are guilty of gross dereliction. The AG is performing such a role when he participates in private judicial review proceedings.

25 From the facts and arguments set out above, it should be evident that the central question of principle in the present dispute as to costs concerns the circumstances in which the AG is entitled to his costs when he intervenes in private judicial review proceedings. The remaining issues concern the application of well-established principles on liability for and quantification of costs. These matters need not detain us presently and are dealt with in a later part of this judgment (in particular, the costs that ought to be awarded to the Respondent, on which issue we also received written submissions from the parties (see [83]–[86] below)). The focus of our analysis will therefore be on the sole question of conceptual difficulty before us.

26 In that regard, we will begin by identifying the juridical basis for making an award of costs in favour of the AG when he intervenes in private judicial review proceedings. This requires, first of all, a brief examination of the AG’s role and the purpose of his involvement in such proceedings. This will be followed by a review of the general reasons of principle and policy, as explained in the case law, for ordering a losing party to pay party-and-party costs to a winning party. The objective in doing so is to determine if these principles apply to the AG’s participation in private judicial review proceedings – and if they do not, whether there are other reasons why the AG should nevertheless be entitled to costs. We will then turn to consider the statutory regime under the Government Proceedings Act (Cap 121, 1985 Rev Ed) (“the GPA”) that specifically governs the AG’s entitlement to costs. In this last-mentioned regard, we will demonstrate that that statutory regime does furnish a juridical basis for the award of costs to the AG in proceedings of the present kind in appropriate cases.

27 Finally, we will set out the factors and guidelines that ought, in our judgment, to guide the court’s discretion in determining whether it is appropriate to make an award of costs in favour of the AG when he intervenes in private judicial review proceedings. In identifying and explaining these factors, we aim to establish a practical framework for analysis that may be applied in this and subsequent cases.

28 We should point out at the outset that although counsel for the AG cited a number of decisions in which costs were awarded in favour of the AG in situations where the AG was not originally a direct party to the proceedings concerned (see, for example, the Singapore High Court decisions of UDL Marine (Singapore) Pte Ltd v Jurong Town Corp [2011] 3 SLR 94 and Yip Man Hing Kevin v Gleneagles Hospital [2014] 2 SLR 515, as well as those cited by the Appellant (see [23] above)), it is clear that in none of these cases was the principle and rationale for the award of such costs raised directly for decision before the courts concerned. It is therefore necessary in the present proceedings to consider this particular issue from the perspective of first or general principles.

29 The starting point of the analysis is O 53 r 1(3), which requires an applicant seeking leave to commence judicial review proceedings to serve the originating documents on the AG at the same time. Order 53 rr 1(1)–1(3) state as follows:

Having been served these documents, the AG becomes aware of the existence of the application for leave, the nature of the applicant’s allegations and the type of relief sought. The AG may then decide to intervene in the proceedings by appearing at the hearing of the leave application to make submissions: see Singapore Civil Procedure 2017 vol 1 (Foo Chee Hock gen ed) (Sweet & Maxwell, 2017) at para 53/1/10.

30 In the Singapore High Court decision of Chan Hiang Leng Colin and others v Minister for Information and the Arts [1995] 2 SLR(R) 627, Judith Prakash J (as she then was) considered the AG’s entitlement to appear at the hearing of the application before her for leave to commence judicial review proceedings. Prakash J cited with approval the Malaysian High Court decision of George John v Goh Eng Wah Brothers Filem Sdn Bhd [1988] 1 MLJ 319, where the applicant had argued that the AG of Malaysia had no standing to appear in such proceedings. Lim Choon Beng J dismissed the objection. He reasoned (at 320) that the grant of leave to apply for a public law remedy would “invariably have the effect of placing the public officer or authority who made the impugned decision and the member of the public who would benefit from the decision so made in a state of uncertainty as to whether he or it should proceed to implement the administrative decision while proceedings for judicial review of it [were] pending even though misconceived”. Thus, the AG was allowed an opportunity to intervene “to remove this uncertainty if there [was] good ground for him to do so in the interest of the government in particular and the public in general” [emphasis added]. Prakash J found at [5] that Lim J’s reasoning was cogent, and that therefore, no objection could be taken to the AG’s attendance at the hearing before her.

31 In Kanawagi a/l Seperumaniam v Dato’ Abdul Hamid bin Mohamad [2004] 5 MLR 495 (“Kanawagi”), the Malaysian High Court elaborated on the reason for the requirement under O 53 r 3(3) of the Malaysian Rules of the High Court 1980 (“O 53 r 3(3)”), which is in pari materia with our O 53 r 1(3), that the applicant is to serve his cause papers on the AG. The court held at [4] that the purpose of the rule was to allow the AG “to ascertain if his participation [was] warranted”. It decided that as the AG was the “guardian of the public interest” and as public bodies “impliedly attract[ed] public interest”, O 53 r 3(3) “ma[de] the AG a nominal party in all judicial review applications”. In the court’s view, whether the AG elected to appear was a matter solely at his discretion. If the AG did appear, the court was “bound to give a hearing and thereafter decide the matter”; it had “no jurisdiction not to hear the AG”. However, if the AG did appear, he then had to “show the public element involved in the application and to either support or oppose the application”. If there was no public element involved, then there would be no need for the AG to appear.

32 In this regard, we note the Singapore High Court’s decision in Lim Mey Lee Susan v Singapore Medical Council [2011] SGHC 131 (a matter related to the present case), where the AG applied to intervene in the substantive hearing of a judicial review application as a party to represent the Ministry of Health. Philip Pillai J found that the court would be assisted if the AG were to make submissions on one proposed area, but did not see any public interest arising from submissions that the AG sought to make with regard to other proposed areas. He therefore did not grant the AG’s application to intervene in respect of those latter areas. Upon initial examination, it appears that Pillai J’s decision envisages a more restrictive approach to the AG’s intervention than that adopted by the Malaysian High Court in Kanawagi, although we observe that Pillai J’s decision concerned the AG’s intervention at the merits stage while Kanawagi pertained only to the AG’s intervention at the leave stage.

33 It is evident from the case law that there remain questions about whether restrictions can be placed on the scope of the AG’s involvement in private judicial review proceedings, depending on the stage at which the AG seeks to intervene. Having said that, it appears to us that if the purpose of the requirement under O 53 r 1(3) is to allow the AG to decide whether his participation in the proceedings is warranted, that suggests that (at least in the context of the hearing of the leave application) the question of whether it is appropriate for the AG to intervene is a matter exclusively for his decision. In any event, these questions do not arise in this case and we therefore leave them for future decision.

34 For present purposes, the crucial point that we draw from the case law is that when the AG intervenes in private judicial review proceedings, he does so in his role as the “guardian of the public interest”. More specifically, the purpose of his intervention is to make submissions to the court on issues of public interest that arise in the proceedings. This is the very reason why O 53 r 1(3) requires service of the cause papers on the AG when leave is sought to commence judicial review proceedings. It is to ensure that the AG is kept aware of the existence of judicial review proceedings so that he can decide whether to intervene to make submissions on issues of public interest that he deems necessary in fulfilment of his duty as the “guardian of the public interest”. In doing so, the AG also assists the court by drawing the court’s attention to public interest considerations that arise in the dispute and that may potentially be affected by the court’s decision.

35 This is a point that counsel for the AG, Mr Khoo Boo Jin, repeatedly emphasised in both his oral as well as written submissions. Indeed, we did not understand either the Appellant or the Respondent to have taken a contrary stance on this matter; nor would we have agreed with them if they had adopted a different position in this regard. The AG’s role as the “guardian of the public interest” is enshrined not only in its mission statement – which is to act as “Guardian of the Public Interest [and] Stewart of the Rule of Law” – but has also been widely recognised as a core function of the AG. In a speech by Minister for Law Mr K Shanmugam (“the Minister”) at the 4th Annual Event of the Attorney-General’s Chambers of Brunei Darussalam, Malaysia and Singapore on 16 April 2009, the Minister described the AG as the “guardian of the rule of law”. As the Minister explained, “[w]ith his dual roles as the government’s legal advisor and Public Prosecutor, the AG serves as a unique custodian of the rule of law, ensuring that our laws are constitutionally valid when enacted, transparent in application, and fair in enforcement”. In this capacity, the AG assists in maintaining “political and social stability … [and] sustain[ing] the public confidence of the local and international community [which] in turn, is fundamental to economic growth and allows society to flourish when the rule of law is sacrosanct”.

36 We make a further point that follows from these observations. When the AG intervenes in private judicial review proceedings in order to carry out his function as the “guardian of the public interest”, he does so not because he wishes to advance the interests of any of the named parties to the litigation or desires a particular outcome for any of the parties. Rather, he participates purely on a non-partisan basis. Put another way, the AG is (and ought to be) concerned only with drawing the court’s attention to issues of public interest that arise in the dispute and making the submissions that he deems necessary and appropriate on those issues. In doing so, the AG discharges his public function, and his role in the litigation ends there. His function does not (and should not) extend to the deliberate enablement of a specific outcome for one of the parties.

37 In our judgment, two consequences flow from this point. First, if the AG’s submissions on a particular issue of public interest have the effect of lending support to one party’s position in the litigation, that effect is purely incidental and forms no part of the AG’s intention. As the AG submits, “[i]n deciding whether to appear and participate in the public interest, AG is objective and non-partisan. Whether AG ends up being aligned with one side or the other is simply based on AG’s assessment of where the public interest lies” (see also [19] above). Indeed, it is conceivable that the AG may make submissions that do not have the effect of either supporting or opposing any party’s position; he may simply highlight an issue of public interest for the court’s consideration. Alternatively, the AG may make certain submissions that have the effect of lending support to certain aspects of one party’s claim or defence, and at the same time make arguments on other issues that undermine other aspects of that same party’s case. Either effect is (and ought to be) incidental and unintended. Second, because the AG is acting on a non-partisan basis and is neither an applicant nor a respondent in the judicial review proceedings, he ought not to be regarded either as a “winner” or “loser” in the litigation in the way that the applicant or the respondent might be a winning or a losing party, depending on the outcome. This is so regardless of whether the AG’s submissions are accepted or rejected, either in whole or in part, by the court. We note that both the AG and the Appellant have taken a common position on this matter.

38 We consider at a subsequent juncture the implications of this analysis on the narrower issue of whether the AG is entitled to costs when he participates in private judicial review proceedings. It is first necessary to describe the rationale for and the principles governing the making of an award of party-and-party costs before evaluating how the AG’s unique role and function in private judicial review proceedings impacts the analysis on costs.

39 In Then Khek Koon and another v Arjun Permanand Samtani and another and other suits [2014] 1 SLR 245 (“Then Khek Koon”), Vinodh Coomaraswamy J provided (at [153]–[160]) a useful exposition of the conceptual and practical underpinnings of an award of party-and-party costs. For present purposes, we propose only to summarise the principles identified.

40 At the heart of the regime on party-and-party costs is the so-called indemnity principle, which gives the winning party an indemnity in respect of his costs from the losing party. As Coomaraswamy J described at [155], one of the functions of the indemnity principle is a compensatory function. The indemnity principle “makes the vindicated winner whole for the costs of what he has shown, by the court’s judgment, to be unmeritorious litigation”. An award of party-and-party costs therefore “serves to compensate the winner”. The matter can be put more simply. The winner of the litigation, of course, can be either the plaintiff or the defendant. If the plaintiff is successful, he is compensated for the legal expenses that he incurred in having to bring his case before the court in order to obtain redress for the wrong that he suffered at the hands of the defendant. If the defendant is successful, he is made whole for the legal costs that he incurred in having to resist the baseless claim brought by the plaintiff against him.

41 While it is ostensibly the aim of an award of party-and-party costs to compensate the vindicated winner, the “ultimate policy” of the indemnity principle lies in enhancing access to justice: see Then Khek Koon at [165]. Shortly put, the indemnity principle enhances access to justice by ensuring that successful litigants are not put out of pocket by having to seek recourse through the legal system in order to vindicate or defend their rights, as the case may be. A litigant with a meritorious cause may be deterred from participating in litigation altogether if he knows that he will not be able to recover his legal expenses even if he prevails at the close of the proceedings. As this court explained in Ng Eng Ghee v Mamata Kapildev Dave [2009] 4 SLR(R) 155 at [1], “unmerited barriers in the path of recovering reasonably incurred costs might well have the chilling effect of deterring parties, in future, from legitimately pursuing or defending their rights”.

42 On appeal from Coomaraswamy J’s decision in Then Khek Koon, this court held that Coomaraswamy J had correctly identified the principles governing the legal regime on costs: see Maryani Sadeli v Arjun Permanand Samtani and another and other appeals [2015] 1 SLR 496 at [28] and [32]. It emphasised at [29] and [34] that a legal system’s rules on costs, including the question of how legal costs should be recovered in litigation, were necessarily a matter of social policy, and that part of the policy of the law in this respect was to enhance access to justice for all.

43 We turn to consider whether these general principles and aims apply in respect of a claim by the AG for costs where he has intervened in private judicial review proceedings. In our judgment, it is difficult to see how these principles are directly engaged in relation to the AG’s involvement, the nature of which we have described at [34]–[37] above. The compensatory principle is not engaged because the AG has not been compelled, either as a result of a court order or by way of any legal consequences that might be visited upon him, to participate in the proceedings. It is unlike the situation of a plaintiff, who cannot vindicate his rights unless he commences legal process, or a defendant, who might have a judgment in default of appearance entered against him if he chooses not to take any action in response to the claim brought against him, baseless though that claim may turn out to be. In other words, the AG has not been put to the necessity of incurring legal expenses either in initiating or resisting court proceedings to vindicate or defend his rights. Likewise, the policy of enhancing access to justice is not relevant because the purpose of the AG’s intervention is not to vindicate or defend any of his rights that have been interfered with.

44 However, this is not the end of the analysis. Although the compensatory principle and the policy of enhancing access to justice do not apply in the context of the AG’s intervention in private judicial review proceedings, this does not necessarily lead to the conclusion that the AG has no interest at all in participating in such proceedings, or does so as a mere busybody. As we have explained, the AG is the “guardian of the public interest”. That role requires him to intervene in private judicial review proceedings to make submissions on issues of public interest where he considers it necessary and appropriate to do so. Therefore, it would be wholly inaccurate to say that the AG has a completely free hand in deciding whether to participate in such proceedings. On the contrary, if the AG does not intervene after the originating summons and supporting documents have been served on him despite taking the view that there are issues of public interest on which he ought to make submissions to the court, the AG will not be performing his public function as the “guardian of the public interest”. He will in fact be in dereliction of his public duty.

45 We surmise that the appropriate approach must therefore be to strike a balance in the tension separating the two ends of the spectrum – that is, between the fact that the AG is neither the applicant nor the respondent to the proceedings and therefore cannot be regarded either as a “winner” or “loser” in the litigation, and the need to recognise that the AG plays an important statutorily-envisaged role in private judicial review proceedings and supports the public interest by his intervention.

46 Before we elaborate on our preferred approach, we will first discuss another key component of the analysis, which is the statutory regime on costs that the AG relies on in the present proceedings.

47 Section 29(1) of the GPA states as follows:

48 The GPA has its roots in the Malaysian Government Proceedings Ordinance 1956 (M Ordinance 58 of 1956) (“the 1956 Ordinance”). The effect of that Ordinance was extended to Singapore by virtue of the Malaysian Modification of Laws (Government Proceedings and Public Authorities Protection) (Extension and Modification) Order 1965 (“the 1965 Order”), which terminated the effect of the Crown Suits Ordinance (Chapter 12) of Singapore. The 1956 Ordinance took on the force of law in Singapore on 25 February 1965 – the date on which the 1965 Order came into operation – about half a year before Singapore’s independence from Malaysia. Section 29(1) of the GPA is in pari materia with s 31(1) of the 1956 Ordinance, with differences in wording that are only cosmetic in nature. The provision has therefore existed in almost entirely the same form for approximately 60 years.

49 The primary significance of the Republic of Singapore Independence Act (Act 9 of 1965) (“the RSIA”), which came into operation on 9 August 1965, was of course the transfer of sovereignty and jurisdiction from the Yang di-Pertuan Agong of Malaysia to the Head of State of Singapore. But the RSIA was also essential in maintaining the continuity of law by preserving, through s 13, “all existing laws” subject to such modifications, adaptations, qualifications and exceptions as might be necessary to bring them into conformity with the RSIA and the independent status of Singapore upon separation from Malaysia. This naturally included the 1956 Ordinance. On 21 April 1966, the Government Proceedings (Amendment) Bill (Bill 20 of 1966) (“the Bill”) was presented to Parliament by Mr Yong Nyuk Lin on behalf of the then Minister for Law and National Development, Mr E W Barker. As was explained during the parliamentary debates on the Bill, which took place the day after the Bill was presented, the purpose of the Bill was to amend the 1956 Ordinance, and a good part of the Bill made merely formal amendments which were necessary in consequence of Singapore ceasing to be part of Malaysia (see Singapore Parliamentary Debates, Official Report (22 June 1966) vol 25 at col 135). The Bill also introduced other not insignificant changes to the law governing the institution of proceedings by or against the Government, which we need not delve into for present purposes. However, no amendment was made to s 31(1) of the 1956 Ordinance. In fact, as we noted at [48] above, to date, the provision remains very much in its original form, persisting even through several revisions of the legislation.

50 It may be observed that s 29(1) of the GPA bears a marked resemblance to s 7(1) of the English Administration of Justice (Miscellaneous Provisions) Act 1933 (c 36) (“the English Act”) in terms of both its structure as well as its content. Section 7(1) of the English Act reads as follows:

51 During the debate on the Administration of Justice (Miscellaneous Provisions) Bill in the House of Lords (see United Kingdom, House of Lords, Parliamentary Debates (25 May 1933) vol 87 at col 1051), the Lord Chancellor explained that in proceedings between one subject and another, costs were in the discretion of the court and the unsuccessful party typically had to pay the costs of the successful party, “but proceedings by and against the Crown ha[d] in the past been regarded as an exception”. There was an “old rule” that the Crown did not ask for costs and would not receive costs; and although, in practice, there were many exceptions to the rule, “the general rule still survive[d]”, and “from time to time bitter complaints [were] made by parties who succeed[ed] in proceedings against the Crown and yet [were] obliged to bear their own costs”. The Lord Chancellor did not elaborate further, but it can be seen from the terms of the English Act that it was clearly intended to effect a change in the law for the reasons that the Lord Chancellor identified – s 7(1) clarifies that costs may be awarded to and ordered against the Crown applying the same approach and principles that govern these matters in ordinary litigation between private parties, but proviso (a) removes any doubt that it remains within the discretion of the court, having regard to the relevant circumstances, to order any party to the litigation to pay the costs of the Crown, regardless of the outcome of the proceedings. These elements of s 7(1) of the English Act are identical to those in s 29(1)(a) of the GPA, which likewise prescribes, in express and unequivocal terms, that the court is to have regard to the nature of the proceedings and the circumstances in which the AG appears and to decide whether, in the exercise of its discretion, it should order any other party to the proceedings to pay the AG’s costs, whatever may be the result of the proceedings.

52 We further observe that while s 7(1)(a) of the English Act applies only to proceedings to which the AG, a Government department or any officer of the Crown “is required to be made a party” [emphasis added], s 29(1)(a) of the GPA (and s 31(1)(a) of the 1956 Ordinance that preceded it) is more broadly worded – it applies to proceedings to which the AG or any officer of the Government “is authorised or required to be made a party” [emphasis added in italics and bold italics]. In our judgment, s 29(1)(a) aptly and effectively encompasses the AG’s intervention in the Appellant’s application for judicial review in the present case. It was entirely within the remit of the AG to intervene as he was, in the language of the proviso, “authorised” [emphasis added in bold italics] to do so pursuant to O 53 r 1(3) (as explained at [33]–[37] above). In addition, we note that at no point during the proceedings in the present case, whether at the leave stage or the merits stage in the court below or, indeed, on appeal, did either the Appellant or the Respondent raise any objections to the AG’s intervention and subsequent participation in the proceedings.

53 Indeed, we are of the view that proviso (a) to s 29(1) of the GPA in fact embodies the relevant principle that ought to guide the court in determining whether or not, in any given case, costs ought to be awarded to the AG when he participates in private judicial review proceedings. We pause to observe that the principle just mentioned is by no means one that is unfamiliar to lawyers in the context of legal practice. On the contrary, it embodies the concept that applies to the awarding of costs in general – that such an award is within the discretion of the court, and whether an award is in fact made depends, in the final analysis, on the facts and circumstances of the particular case concerned. This concept is embodied in O 59 r 2(2) of the Rules of Court, which establishes that costs are “in the discretion of the Court, and the Court [has] full power to determine by whom and to what extent the costs are to be paid”. The statutory position is thus entirely consonant with the position at common law.

54 We hasten to add, however, that the discretion of the court on costs is not unfettered; it must in every case be exercised judiciously, “guided by the beacon of achieving the fairest allocation of costs in the circumstances of the case at hand”: see the Singapore High Court decision of Re Shankar Alan s/o Anant Kulkarni [2007] 2 SLR(R) 95 (“Re Shankar Alan”) at [17] and [21] per Sundaresh Menon JC (as he then was). We bear this guidance well in mind in outlining the general approach to be applied in determining if the AG is entitled to costs when he intervenes in private judicial review proceedings.

55 In our judgment, the court must bear three factors or guidelines in mind when deciding whether it is appropriate to award the AG costs in circumstances such as the present. Taken together, these guidelines provide a calibrated approach between the imperatives of supporting the AG’s participation in private judicial review proceedings in the public interest and ensuring that the parties are treated fairly in relation to their liability to pay costs.

56 First, as a threshold requirement, the issue(s) pursued by the AG must, objectively speaking, concern the public interest. We have explained at [36] above that the function of the AG in private judicial review proceedings is to bring to the court’s attention issues of public interest that arise in the dispute and to make submissions that he regards as necessary and appropriate on those issues. This is the proper basis and justification for the AG’s involvement in such proceedings. We emphasise that we are not presently concerned with the question of whether the AG can or cannot lawfully intervene if the issues that he raises are not objectively issues of public interest. As mentioned at [33] above, there remain unanswered questions as to whether the court may properly restrict the scope and extent of the AG’s intervention and the circumstances in which it will be appropriate for the court to do so. The dispute presently before us pertains only to the awarding of costs to the AG, and this is a matter that lies entirely within the court’s discretion, both as a matter of general principle and as expressly provided for in s 29(1)(a) of the GPA. Thus, if the court takes the view that the issues that the AG raised do not pertain to the public interest, it is open to the court to refuse to order costs in favour of the AG even if the AG considers that his intervention and submissions were warranted in the public interest.

57 While it is ultimately for the AG to satisfy the court that the issues which he raised are issues of public interest in order to justify his entitlement to costs, we agree with the AG’s submission (see [17] above) that if any party to the proceedings considers that the issues or arguments raised by the AG do not concern the public interest, that party should make his objection known at the earliest possible stage. If a party displays no discomfort regarding the nature of the AG’s submissions at any point until the proceedings come to a close and the issue of costs falls to be decided, we consider this a compelling indication that that party agrees that the AG’s intervention was warranted in the public interest.

58 We note the AG’s suggestion that “public interest” may be defined as “[t]he general welfare of a populace considered as warranting recognition and protection” or “[s]omething in which the public as a whole has a stake”, citing Black’s Law Dictionary (Thomson Reuters, 10th Ed, 2014) at p 1425. Given the breadth of the concept, we do not think it would be useful to venture a complete definition of what “public interest” entails. We would observe only that since the AG intervenes on a non-partisan basis, disinterested in the interests of the specific parties to the litigation, the issues that he raises must logically have a discernible impact beyond those specific parties and the specific dispute before the court. Beyond this, we do not think it would assist to demarcate with greater particularity a definition of “public interest”; whether the AG’s submissions and the issues that he raises concern the “public interest” is a matter for the court’s judgment in each case.

59 The second factor that we identify serves as a control mechanism for the first. The issue of public interest raised by the AG must be relevant to, and go towards, the determination of the actual dispute before the court. It cannot be an issue that is unrelated or only tangentially related to the dispute, or worse – entirely hypothetical. It is true that the AG’s focus is on the proper consideration and treatment of issues of public interest, and not on the resolution of the specific dispute at hand. But it would be wrong to ignore the fact that the AG’s submissions are made in the context of an ongoing dispute. The hearing of that dispute is not a forum in which the sole, or even the main, purpose is the ventilation of issues of public interest. The court’s aim, as described at [54] above (citing Re Shankar Alan), must be to “achiev[e] the fairest allocation of costs in the circumstances of the case at hand”. It would not be fair to the parties in terms of their time and costs expended if the AG were in a position to make a claim for costs even though his submissions were untethered to the resolution of the dispute. In addition, we are of the view that the AG should not be entitled to costs if, as a result of the issues or arguments that he raised, he unnecessarily or unreasonably protracted or added to the costs or complexity of the proceedings, by way of analogy to O 59 r 6A of the Rules of Court. We note that the AG himself accepts in his written submissions that this is the correct position (see [18] above).

60 The above analysis will require, of course, careful consideration of the precise issue(s) which the AG raised in the proceedings and the role which such issue(s) played in the context of the proceedings as a whole. We will explain in due course why this also furnishes the court with some guidance as to the quantum of costs that ought to be awarded to the AG.

61 Before moving to the issue of quantification of costs, we will outline the third factor that the court should bear in mind. In order to be entitled to costs, the AG must have succeeded in his submissions on the issue(s) raised. As a matter of fairness to the parties, it is difficult to justify making a costs order in favour of the AG – with such costs to be paid by one or the other (or both) of the parties – if the court has seen no merit in the AG’s argument and has rejected it. Importantly, we are of the view that the AG’s success or failure for the purposes of awarding costs should be considered in the context of his submissions on the issue(s) of public interest raised, and not on the basis of his being an overall “winner” or “loser” in the litigation. This is because the AG cannot be regarded as an overall “winner” or “loser”; as we have explained at [36]–[37] above, the AG does not take an interest in the outcome of the specific dispute before the court. The nature and effect of the AG’s submissions are based solely on his assessment of where the public interest lies. An argument- or issue-based assessment of the AG’s success or failure therefore reflects the fact that the arguments advanced and the issues raised by the AG are based purely on what, in his view, the public interest demands, and the fact that he conducts himself purely on a non-partisan basis.

62 By way of an example, the AG may make submissions that favour the applicant on a particular issue, while making submissions that favour the respondent on another issue. If the applicant opposes the AG’s submissions on the second issue and the respondent opposes the AG’s submissions on the first, and both parties fail on the respective issues opposed, then it would be appropriate (subject to the other requirements identified above) for each of the parties to pay costs to the AG on the respective issues opposed. Looked at in this light (and depending on the precise facts and circumstances), it might well be the case that the AG’s submissions on a particular issue do not impact either party’s position in the litigation and thus neither party opposes his submissions; or alternatively, the AG’s submissions might be adverse to both parties’ interests and thus both parties oppose those submissions (unsuccessfully). In the former situation, it might be appropriate that neither party pays the AG’s costs, and in the latter situation, that both parties contribute towards the costs to be awarded to the AG.

63 We do not propose to close the door entirely to the possibility that in an appropriate case, even though the AG advanced submissions on an issue of public interest and such submissions were opposed by a party and ultimately rejected by the court, he might nevertheless be entitled to costs. Although we do not wish to delimit the boundaries of the circumstances in which this might be the appropriate outcome – given that this is ultimately a matter for the court’s discretion in each case – we think that there might be good grounds to make such a costs order if: (a) the issue of public interest raised by the AG was one of obvious and considerable public importance; (b) the grounds relied on by the AG in support of his arguments were sound and possessed some degree of merit; and (c) the AG’s submissions were, in the final analysis, of substantial assistance to the court in reaching its decision even though it ultimately did not agree with the AG’s position. In addition, the court should also take into account the novelty or complexity of the issue in deciding if it is appropriate to award the AG some costs despite his unsuccessful arguments. We reiterate that this is not intended to be an exhaustive or categorical statement of the relevant factors, but rather, a guide to the exercise of the court’s discretion.

64 In any event, even if the AG is unsuccessful in relation to the issue raised, we do not consider that he should generally be ordered to pay costs given that his role in the dispute is, as already emphasised, non-partisan in nature, and given that he has participated in the proceedings only to the extent that it was necessary for him to do so in the public interest. We note that the parties are in agreement on this point. The decisions in Top Ten and, more recently, Ang Pek San Lawrence v Singapore Medical Council [2015] 2 SLR 1179 (“Ang Pek San Lawrence”) provide helpful guidance, even though the facts of these cases involve somewhat different contexts. We will briefly describe these two decisions and the principles set out therein.

65 Top Ten concerned an order of the Singapore High Court that the Respondent (who was the appellant in that case) pay costs to the complainant after the Respondent unsuccessfully opposed the complainant’s application to the High Court for a review of a decision by the Council of the Law Society dismissing the complaint. The High Court found that there was sufficient material to justify an investigation into the complaint, and ordered the Respondent to pay 50% of the complainant’s costs in the proceedings. The appeal against the High Court’s decision centred on the application of the principle identified by the English Court of Appeal in Baxendale-Walker, the so-called “Baxendale-Walker principle”. In Baxendale-Walker, the English High Court had ordered the Law Society of England and Wales (“the LSEW”) to pay costs to a solicitor following the latter’s successful defence against an allegation made against him. On appeal to the English Court of Appeal, Sir Igor Judge P (as he then was) held that the LSEW was not liable to pay costs, reasoning as follows (at [34] and [39]):

66 On the facts of Top Ten, this court found at [24] that the Respondent had been acting as a regulatory body and had therefore been protecting the integrity of the legal profession, which was in the interests of the public. The Respondent should therefore “not be equated with an ordinary litigant who litigates to enforce or protect his or her private interests”. It concluded at [33] that the Baxendale-Walker principle should have been applied as a starting point to determine if costs should have been ordered against the Respondent, and not the principle that costs follow the event.

67 Ang Pek San Lawrence was the Singapore High Court’s judgment on costs following its decision to allow a medical practitioner’s appeal against the decision of a Disciplinary Committee of the SMC convicting him of professional misconduct. After finding that the Disciplinary Committee had the power to order the costs of the disciplinary proceedings against the SMC, and that the High Court likewise had the power to order the costs of both the disciplinary proceedings and the appeal against the SMC, the court turned to consider whether such costs should be ordered against the SMC in the light of the Baxendale-Walker principle and the decision in Top Ten. Sundaresh Menon CJ, delivering the judgment of the court, cited with approval the decision of Lord Bingham of Cornhill CJ (as he then was) in City of Bradford Metropolitan District Council v Booth [2000] COD 338. Menon CJ drew the following points of principle from Lord Bingham’s decision (at [42]):

After describing the decision in Baxendale-Walker, Menon CJ remarked as follows (at [44]):

Menon CJ then summarised (at [55]–[57]) the principles to be applied when determining whether to make an adverse costs order against the SMC in relation to disciplinary proceedings:

68 In the result, the High Court ordered the SMC to bear the costs of both the disciplinary proceedings and the appeal. Menon CJ observed at [61]–[63] that not only had no reason been provided as to why the SMC had persisted in pursuing disciplinary proceedings despite the earlier decision of a Complaints Committee dismissing the complaint, the SMC had also failed to sufficiently particularise the charges and had caused the practitioner to endure two tranches of proceedings and thereby suffer substantial financial loss.

69 Before proceeding further, it is of the first importance to highlight that in both Top Ten and Ang Pek San Lawrence, the Respondent and the SMC in those respective cases were actual parties to the proceedings, unlike the AG in the present proceedings. This is a distinction that is significant as a matter of principle because, as we have explained, the AG cannot be regarded as a “winner” or “loser” when he intervenes in private judicial review proceedings. Top Ten and Ang Pek San Lawrence nevertheless remain useful authorities for present purposes because they explain the effect of the fact that the losing party was discharging a public function in the public interest on the court’s decision on costs. The preferred approach in Ang Pek San Lawrence is a multifactorial analysis that places a considerable amount of significance on the fact that the losing party was exercising a public or regulatory function. But it also requires an assessment of whether that function was carried out in an honest, sound and reasonable manner, the financial prejudice to the successful party and any other relevant circumstances. The overriding aim of the court is to render a costs order that is just and reasonable.

70 In our judgment, these factors remain relevant in determining whether costs should be ordered against the AG when he intervenes in private judicial review proceedings and his submissions do not succeed. However, when one considers not only the fact that the AG is discharging his public function as “guardian of the public interest” but also the fact that he is not properly to be regarded as a disputant in the litigation at all, it becomes clear that it is appropriate to make such adverse costs orders only in very exceptional circumstances. Indeed, it is difficult to think of a situation in which it might be apt to make such orders, save where the AG has intervened with improper motives, such as to deliberately advance the interests of one of the parties to the litigation. That would take the AG out of his capacity as a non-partisan advocate of the public interest and into the shoes of a de facto party to the litigation.

71 We now turn to the issue of quantification, assuming for present purposes that the court decides to exercise its discretion to award costs to the AG.

72 Given that the AG’s position is sui generis inasmuch as his participation in private judicial review proceedings is in the public interest and is, looked at in that light, non-partisan in nature, we consider that the costs awarded to the AG ought generally to be on a lower scale than those awarded to the winning party in the proceedings (which costs are, as already explained, intended to compensate that party). This reflects the fact that the nature and scope of the AG’s involvement in the litigation is substantially more limited than that of the actual parties to the dispute, and also serves to control the quantum of costs to be paid by the party who loses on the issue(s) raised by the AG.

73 As mentioned at [60] above, it is important for the court to have regard to the precise issue(s) which the AG raised in the proceedings and the role which such issue(s) played in the overall context of the proceedings. For instance, if the court determines that the issue raised by the AG was of some, but limited, relevance to the determination of the dispute, or would impact only a narrow class of persons or arise only in uncommon scenarios, the court ought to factor these matters into its decision on the quantum of costs to be awarded to the AG. This, like other aspects of the court’s assessment, is a matter for the court’s careful judgment on the facts and circumstances of each case.

74 We now summarise the approach that we have set out. First, as a threshold requirement, the court must be satisfied that the issues raised by the AG are, objectively speaking, of public interest. If this requirement is not satisfied, then there would not even be a connection between the AG’s public function and his intervention in the private judicial review proceedings concerned. Second, the issues of public interest raised must be relevant to, and go towards, the determination of the dispute. This is necessary as a matter of fairness to the parties, who are engaged in litigation in order to resolve their differences and who do not simply provide a platform for the general ventilation of matters of societal concern. Third, the AG must have succeeded in his submissions on the issue(s) raised. An issue- or argument-based approach to ascertaining the AG’s success or failure and, hence, his entitlement to costs, rather than an assessment of whether the AG is the overall “winner” or “loser” in the litigation, should be applied as a matter of principle because it reflects the nature of the AG’s role in the proceedings – which is to raise relevant issues of public interest for the court’s consideration, and not to argue in favour of a particular outcome in the dispute at hand. However, there are circumstances in which it may be proper for the court to award the AG costs even though his submissions are unsuccessful.

75 In closing, we come back to the language of s 29(1)(a) of the GPA, which stipulates that the court should “have regard to the nature of the proceedings and the circumstances in which the Attorney-General or [the Government] officer appears and may in the exercise of its discretion order any other party to the proceedings to pay the costs of the Attorney-General or such officer whatever may be the result of the proceedings”. We have sought to provide some guidance on how the court ought to make its costs determination in relation to the AG. We reiterate that the core of the general principle set out in s 29(1)(a) requires the court to look at the precise facts and circumstances of the case itself. This simultaneously furnishes the court with the requisite flexibility and discretion that are, in fact, the hallmarks of a decision to make an award of costs generally. Such flexibility is particularly apposite and, indeed, necessary in view of the fact that the AG’s role in private judicial review proceedings is a unique one. That role is to assist the court by making submissions on matters relating to the public interest, of which the AG is the guardian.

76 We will now set out our decision on the appropriate costs orders to be made in the present case, beginning with the order as between the Appellant and the AG, followed by that as between the Appellant and the Respondent. As a preliminary matter, we note that the Judge ordered on 8 August 2016 that the costs of the hearing before him be dealt with after or at the appeal, subject to any other directions that might be given. Therefore, no decision on costs has yet been made by the High Court.

77 In determining the AG’s entitlement to costs in the present case, we will, as a guide to the exercise of our discretion on costs, apply the approach that we have described above to the facts and circumstances before us.

78 We note, first, that the point raised by the AG that clearly related to the public interest – and that was relevant to the determination of the dispute between the parties – concerned the locus standi of the Appellant to make a complaint to the Respondent. For the reasons that we identified at [34] of the Judgment (CA), we recognise that this is a question of general importance because it potentially circumscribes the class of persons who can make complaints against lawyers to the Respondent and thereby initiate the disciplinary process set out in s 85 of the LPA. As described at [9] above, the AG argued that the Appellant must show that he had standing to make such a complaint, and that since the Appellant was a “stranger” to the disciplinary proceedings against Dr Lim, he therefore did not have standing: see the Judgment (HC) at [54]. However, the AG’s argument in this particular regard was rejected by the Judge (see the Judgment (HC) at [76]) and was not pursued on appeal.

79 If this were the only issue of public interest for which costs could be awarded to the AG, we would be minded to make no order as to costs, having regard to the specific facts and circumstances surrounding the arguments (as well as the Judge’s decision) with regard to this particular issue. In particular, we note that the party which ultimately prevailed in the judicial review application (viz, the Respondent) did not support either the AG’s position or his arguments on the issue of locus standi. Neither, of course, did the Appellant, who took the position that a person who wished to make a complaint against a solicitor did not need to establish standing before he or she could do so: see the Judgment (HC) at [52]. In this connection, we reiterate that the Appellant was the party who did not prevail at the conclusion of the matter either here or in the court below, although, vis-à-vis the AG, he succeeded before the Judge on the issue of locus standi.

80 However, the AG also made submissions on other issues that, in fact, largely centred on the substantive merits of the judicial review application itself. Amongst these issues, the matter that most clearly pertained to the public interest concerned the nature and scope of the professional and ethical duties owed by lawyers in making claims for party-and-party costs. This issue was considered both in the High Court (see the Judgment (HC) at [85]–[104]) and on appeal (see the Judgment (CA) at [39]–[42]). We find that this is an issue of undeniable public importance given that the filing of claims for party-and-party costs is a cornerstone of our litigation system. Lawyers must ensure that they are acting in accordance with their professional and ethical responsibilities whenever they file such claims on behalf of their clients. We do not think that either the Appellant or the Respondent can be taken to have adopted or expressed any contrary view on the general importance of this issue of principle, given that neither of them raised any objections at any point in time to the AG’s participation in the proceedings both here and below. The AG’s oral and written submissions on this issue were accepted, were plainly relevant to the resolution of the dispute at hand and, indeed, were of considerable assistance to the court both at first instance and on appeal. In the circumstances, we consider it appropriate to accord proper recognition to the AG’s contribution on this issue in making our decision on costs. However, this is tempered by the fact that the parties broadly agreed on the correct legal position on this issue (see [48] of the Judgment (CA)). The heart of the dispute was whether the RC had misstated the correct position in law in the Decision Letter (ie, the first ground of the Appellant’s application for judicial review), which ultimately boiled down to a question of interpretation regarding what was said in the Decision Letter. The costs to be awarded to the AG for his success on this issue should therefore be reduced to reflect this fact.

81 In addition, we think it important to highlight the fact that not only was the Respondent (ie, the Law Society of Singapore) the governing body that is centrally involved in such matters, it was, of course, itself a party to the proceedings. It was fully able to and did make relevant and, indeed, similar submissions to those made by the AG on this issue. In the circumstances – and notwithstanding the fact that we derived a good measure of assistance from the AG’s submissions in the present case – we take the view that in future cases involving similar complaints, where the Respondent is a party to the proceedings and is ostensibly in a position to make the necessary submissions on the issues at hand, the AG may wish to take a more conservative view of whether his participation is required. We reiterate that the AG should not expect to receive any or any substantial amount of costs if the requirements that we have summarised at [74]–[75] above have not been satisfied. If the AG’s assistance is rendered tangential or negligible in the light of the submissions made by the parties to the dispute, that is a relevant factor for the court’s consideration.

82 In these somewhat unusual circumstances, we consider it appropriate to award the AG costs of $15,000 (inclusive of disbursements) in respect of the proceedings both here and below. These costs are to be paid by the Appellant, who unsuccessfully opposed the AG’s submissions both here and below (leaving aside the issue of locus standi, in respect of which, as we have explained, it is appropriate to make no order as to costs).

83 Having succeeded both before the Judge and on appeal, the Respondent seeks the costs of the proceedings here and below from the Appellant, and submits that the costs of the hearing before the Judge should also be determined by this court rather than remitted to the High Court for the Judge’s decision. The Respondent further argues that the issue of whether the RC’s decision was susceptible to judicial review – an issue which the Respondent raised before the Judge and on which it was ultimately unsuccessful (see [9] above) – was not canvassed or argued unreasonably, especially in the light of the fact that it was a novel issue. It also points out that it did not continue to pursue the issue on appeal.

84 The Appellant argues that the Respondent should bear its own costs in respect of both the proceedings below as well as the appeal. He highlights that he was successful on the issue of the susceptibility of the RC’s decision to judicial review, and also on the question of whether or not the RC should have relied on WP’s purported clarification that Mr Yeo was not involved in the taxation (see [11] above).

85 In our judgment, it is appropriate that the costs of the proceedings below be decided by this court, alongside the costs of the appeal. This course of action is clearly open to us pursuant to s 38 of the Supreme Court of Judicature Act (Cap 322, 2007 Rev Ed), which establishes that this court “may make such order as to the whole or any part of the costs of appeal or in the court below as is just” [emphasis added]. Although we do not consider that the Respondent unnecessarily or unreasonably added to the costs of the proceedings by contesting the issue of the susceptibility to judicial review of the RC’s decision, we note that this issue took up a considerable amount of time and argument in the court below and ultimately formed a substantial part of the Judge’s written grounds: see the Judgment (HC) at [22]–[49]. We therefore find it appropriate to reduce the costs awarded to the Respondent to reflect this fact. In relation to the Appellant’s submission that a further reduction in costs is justified on the basis of his success on the issue regarding WP’s clarification about Mr Yeo’s involvement in the taxation, we have explained in our judgment on the appeal that the Appellant’s argument on this issue was ultimately immaterial because he was not able to establish in the first place that there was any reason to impugn the RC’s decision that Mr Yeo was not guilty of professional misconduct: see the Judgment (CA) at [31] and [70].

86 In the circumstances, we think it appropriate to order the Appellant to pay the Respondent costs of $60,000 (inclusive of disbursements) for the proceedings below and costs of $30,000 (likewise inclusive of disbursements) for the appeal, and we so order.

87 For the foregoing reasons, we make the orders that we have detailed at [82] and [86] above. These orders are intended to be dispositive of the issue of costs both here and below.