Singapore Democratic Party v Attorney-General [20 20] SGHC 25 [2020] SGHC 25

By this Originating Summons (“OS”), the appellant seeks to set aside three Correction Directions (“CDs”) issued by the Minister of Manpower. All three CDs were issued on 14 December 2019 pursuant to s 10(1) of the Protection from Online Falsehoods and Manipulation Act 2019 (No. 18 of 2019) (“POFMA”). The appellant’s application is made under s 17(1) of the POFMA, which frames the application as an appeal. Section 17(4) of the POFMA gives the Court power to set aside the three CDs if certain prescribed circumstances are established.

Background

A brief chronology of the relevant facts is as follows.

On 8 June 2019, the appellant published an article (the “SDP Article”) on its online website titled “SDP Population Policy: Hire S’Poreans First, Retrench S’Poreans Last”. This date of first publication is relevant to one of the arguments raised by the appellant because the POFMA only came into force on 2 October 2019. This issue is considered later in this judgment at [51] to [56].

On 30 November 2019, the appellant published a post (the “November Facebook Post”) on its Facebook page titled “Singapore Democratic Party (SDP)” with some text, an image and a hyperlink to the SDP Article. The posts on this Facebook page are accessible to the public.

On 2 December 2019, the appellant published a post (the “December Facebook Post”) on Facebook with some text and an image containing two graphical illustrations. There was also in this post a hyperlink to the SDP Article. This post was visible to certain users of Facebook as a sponsored post paid for by the appellant’s Vice-Chairman, John Tan Liang Joo.

On 14 December 2019, the POFMA Office of the Info-communications Media Development Authority, on the direction of the Minister for Manpower, issued three CDs to the appellant. One CD (“CD-1”) referred to the SDP Article, while the other two referred to the November Facebook Post (“CD-2”) and the December Facebook Post (“CD-3”). All three CDs stated that the material they referred to contained “a false statement of fact”, and directed the appellant to add correction notices at the top of the SDP Article, November Facebook Post and December Facebook Post respectively by no later than 4.00pm on 15 December 2019. The appellant complied with all three CDs.

On 3 January 2020, the appellant applied to the Minister of Manpower under s 19 of the POFMA for the cancellation of all three CDs. On 6 January 2020, the Ministry of Manpower rejected the said application. On 8 January 2020, the appellant filed the present OS to set aside the three CDs.

The statutory framework

Under s 10(1) the POFMA, a CD may be issued if a false statement of fact (referred to in Part 3 of POFMA as the “subject statement”) has been or is being communicated in Singapore, and a Minister is of the opinion that it is in the public interest to issue the CD.

The subject statement is identified based on the material against which the CD is issued. It refers to a false statement of fact which is reasonably discerned from the material. Identifying the appropriate subject statement therefore requires examination of the material and the meaning (or meanings) which that material conveys.

In these proceedings, parties have referred to the subject statement in a whole range of ways. It has been described as being, inter alia, the “meaning” of the material in question, the “interpretation” to be drawn from the material, and even the “appropriate meaning” of the material. The respondent has also used the term “false statement of fact” to specifically refer to the subject statement(s). These definitions, while varied, do generally illustrate what the subject statement is. In short, it is an interpretation, drawn from the material in question, which the Minister asserts is a false statement of fact. Given the differing terminology used on this point, I should explain that the references to “meaning”, “interpretation”, the “false statement of fact” and the “subject statement” in this judgment should be understood in this same way.

In my view, there are thus two questions the Court must examine in relation to the subject statements in each case:

(a) First, can the subject statement be reasonably interpreted from the material in question?

(b) Second, is that subject statement false?

On the issue of whether a subject statement has been “communicated” in Singapore, the ambit of a “communication” for the purposes of a CD is set out in s 3(1) of the POFMA. Under this provision, a statement or material is communicated in Singapore if it is made available to one or more end-users in Singapore on or through the internet.

If a CD is issued, the recipient of the CD may be required to communicate, in a specified form or manner, by a specified time, a notice (referred to in the POFMA as a “correction notice”). The correction notice will contain one or both of the following:

(a) A statement that the subject statement is false, or that the material contains a false statement of fact; and/or

(b) A specified statement of fact, or a reference to a specified location where the specified statement of fact may be found, or both.

Section 17(5) of the POFMA makes clear that the High Court may only set aside a CD on the following grounds:

(a) the person did not communicate in Singapore the subject statement;

(b) the subject statement is not a statement of fact, or is a true statement of fact; and

(c) it is not technically possible to comply with the CD.

The ground relied on by the appellant in the present case as the basis for setting aside all three CDs is the latter limb of s 17(5)(b) of the POFMA, that is, each subject statement is “a true statement of fact”.

The Correction Directions

CD-1 pertains to the SDP Article, and identifies the subject statement from the SDP Article as being that “Local PMET retrenchment has been increasing”. “PMET” is an acronym for professionals, managers, executives and technicians. The basis for identifying this subject statement was further explained in the respondent’s reply affidavit as being primarily a reference to the first sentence in the eighth paragraph of the SDP Article (in bold below), read in context of the entire Article. For ease of reference, the relevant portions of the SDP Article which the parties have referred to be as being relevant and providing the proper context are as set out below:

CD-2 takes issue with the November Facebook Post on the basis that the said post includes a hyperlink to the SDP Article, which is itself the subject of CD-1. The subject statement identified is thus the same as that in CD-1.

CD-3 identifies two subject statements. The first, which relates to a graphical illustration on the bottom left hand side of the December Facebook Post showing a downward pointing arrow on a graph titled “Local PMET Employment”, is that “Local PMET employment has gone down”. The December Facebook Post also includes a hyperlink to the SDP Article (the subject of CD-1). Thus, the second subject statement identified is the same as that in CD-1.

As is clear from the above, the defensibility of CD-1 is closely intertwined with that of the other two CDs. All three CDs relate, at least in part, to the SDP Article.

The correction notices in all three CDs also included links to a government website (the “Factually Website”) setting out, inter alia, data on the number of local PMETs from 2015 to 2019, the number of local retrenchments and local PMET retrenchments from 2015 to 2018, the number of Employment Pass holders from 2015 to June 2019, and the number of retrenched local PMETs per 1,000 local PMET employees from 2015 to 2018.

Both sides are in agreement that the subject statements identified in the CDs are interpretations of the Minister for Manpower, in the sense that the subject statements are the meanings she has identified from the SDP Article, the November Facebook Post and the December Facebook Post. Both sides are also in agreement that it is for the Court to decide whether the Minister’s interpretations can indeed be derived from the words and graphical illustrations used in these published material.

Issues

As outlined at [11] above, the issues that the court has to decide in relation to each CD are as follows:

(a) On a proper interpretation of the SDP Article, the November Facebook Post and the December Facebook Post (“the SDP Material”), do the subject statements identified in the CDs arise?

(b) Should the subject statements arise, are they true or false?

In addition to the issues outlined above, important questions also arise as to which party bears the burden of proof in these proceedings, and what the standard of proof should be.

It also bears note that, even though s 17(5)(b) of the POFMA provides that a ground for setting aside CDs is where the “subject statement is not a statement of fact…” (emphasis added), the appellant did not take the position in its affidavit in support of this application that the subject statements identified by the Minister for Manpower are statements of opinion, and not statements of fact. Further, there was no suggestion in the course of the appellant’s oral arguments that the subject statements were statements of opinion.

It was only in the appellant’s written submissions which were filed on 22 January 2020, post the oral hearing, that the appellant sought to argue in relation to CD-1 that “The statement is absolutely and undeniably true, and the point being made is an opinion based on the true statement” (emphasis added).

In this regard, Rule 10 of the Supreme Court of Judicature (Protection From Online Falsehoods and Manipulation) Rules 2019 (“POFMA Rules”) (Cap 322, No. S 665) provides that:

No leave was sought by the appellant to raise this ground of challenging CD-1. The leave that I had granted to the appellant to file written submissions post-hearing was limited to dealing with legal points that had been raised by the respondent’s counsel in his oral submissions, and not to raise new grounds of challenge.

Be that as it may, I find that the contention that the subject statement in relation to CD-1 is a statement of opinion to be a rather hopeless one. The starting point for determining whether a statement is one of fact or opinion is s 2(2)(a) of the POFMA, which provides that a statement of fact is a statement which “a reasonable person seeing, hearing or otherwise perceiving it would consider to be a representation of fact”. Applying this test to the statement that there is a “rising proportion of Singapore PMETs getting retrenched” in the context of the SDP Article, I find that a reasonable person reading the Article, and that sentence in particular, would not consider the statement to be merely an expression of opinion. The statement is not worded as a comment or an expression of opinion by the writer. Rather, in the context of the rest of the SDP Article, the statement appears to be a factual assertion of the state of affairs in Singapore in relation to the retrenchment of “Singapore PMETs”.

That conclusion is buttressed by the fact that the same outcome would be reached if the principles for determining fact and opinion in defamation law were applied. Indeed, the Minister for Law was clear in the Second Reading Speech for the Protection from Online Falsehoods and Manipulation Bill that:

The Minister for Law’s reference to “existing case law” in distinguishing between fact and opinion must obviously be a reference to the principles applicable to the defence of fair comment in defamation law. Applying that body of case law, whether a statement is one of fact or comment will depend on the precise words, the context in which the passage is set out, and the content of the entire publication: Chen Cheng and another v Central Christian Church and other appeals [1998] 3 SLR(R) 236 at [34] and [35]. As stated in Golden Season Pte Ltd and others v Kairos Singapore Holdings Pte Ltd and another [2015] 2 SLR 751 at [100], determining whether a statement is one of opinion or fact will depend on the entirety of the circumstances, and falls to be determined on an objective standard of how the statement would strike the ordinary reasonable reader. Whether a statement is an opinion or a statement of fact is a question of fact (Review Publishing Co Ltd and another v Lee Hsien Loong and another appeal [2010] 1 SLR 52 (“Review Publishing”) at [144]) and it is for the Court to determine whether a statement is one of opinion or fact.

Applying these principles, in my judgment, it is clear from the context of the SDP Article that the subject statement identified in CD-1 has been cast as a statement of fact, and not a comment or opinion. There is nothing which qualifies it or which even indicates that it is based on particular data sources, thus militating against any finding that the subject statement would be regarded as a comment by a reasonable reader: Oversea-Chinese Banking Corp Ltd v Wright Norman and others and another suit [1994] 3 SLR(R) 410 at [32] to [33], Review Publishing at [145] to [148]. It appears, on the contrary, that the SDP Article was quite unequivocal in making an assertion that there is “a rising proportion of Singapore PMETs getting retrenched”. That would be regarded as a factual statement rather than a comment.

The Ministry of Law’s own guidance on the POFMA dated 9 July 2019 may perhaps be instructive in identifying statements of opinion. In one of its examples, it is stated:

Unlike the example above, the SDP Article did not set out any methodology, nor did it outline any references to data. The claim that there was a “rising proportion of Singapore PMETs getting retrenched” can be read only as a straightforward assertion of a factual state of affairs because no supporting data was cited or referenced. I thus find that the clear understanding by a reasonable person, reading the SDP Article in context, would be that the “rising proportion of Singapore PMETs getting retrenched” is a statement of fact. There is no basis to construe it as a statement of opinion.

The burden and standard of proof

Turning next to the question of whether the appellant or respondent bears the burden of proof, the POFMA does not specify which party bears the burden of proof in the present proceedings. As was confirmed by Deputy Attorney-General Hri Kumar Nair (“DAG Nair”) who appeared on behalf of the respondent, there is also nothing in the Parliamentary Debates which sheds light on this issue.

The appellant argues in its written submissions that the respondent ought to bear the burden of proof for three reasons:

(a) First, the constitutional right to free speech means that, if a Minister seeks to impose limitations on a citizen’s right of free speech, it must be for the Minister to establish the existence of grounds for that limitation.

(b) Second, members of the public making various statements online cannot be reasonably expected to discharge the burden of proof because they do not have access to the required documents and data, most of which would be held by the government.

(c) Third, the appellant seeks to draw an analogy to setting-aside applications and the fact that the party seeking leave to serve a writ out of jurisdiction bears the burden of proof even where a challenge is subsequently brought to dispute the jurisdiction of the Court. However, no explanation was provided by the appellant as to why setting-aside would be an appropriate analogy.

On the other hand, DAG Nair argues on behalf of the respondent that the appellant bears the burden of proof for the following reasons:

(a) First, reliance was placed on s 103(1) of the Evidence Act (Cap 97, 1997 Rev Ed) (“EA”) to argue that since the appellant was asserting that the subject statements were true in order to succeed in this appeal, the burden lay on them to prove thus on the balance of probabilities. In this regard, s 104 of the EA states that the burden of proof lies on the person who would fail if no evidence at all were given on either side.

(b) Second, reference was made to the words used in s 17(5) of the POFMA, which sets out the potential grounds of appeal to the High Court as, inter alia, the subject statement being a true statement of fact. The argument made is that, since the words suggest that a party has to prove that the subject statement is true in order to succeed, it must be inferred that the burden of proof lies on the appellant. To buttress this argument, the respondent also refers to s 3(5) of the EA, which provides that a fact is said to be ‘not proved’ when it is neither proved nor disproved. It goes on to argue that s 3(5) of the EA illustrates that if the appellant does not prove that the subject statement is a true statement of fact, the truth of that said statement is ‘not proved’.

(c) Third, the respondent argues that the interactions between the Minister and the statement-maker prior to an appeal to Court being brought would show that the burden of proof must be on the statement-maker. The process was described by DAG Nair in the following way. A person would first have, for his own reasons, made an online communication. The Minister would then have identified what was false about that communication, and informed that person about the falsity of the statement in the form of a CD. Upon reading the CD and the evidence provided by the Minister, the maker of the communication can then decide whether he still wishes to challenge the CD by first applying to the Minister for re-consideration, and thereafter taking the matter to Court under s 17(1) of the POFMA. If the statement-maker elects to bring the matter to Court even after learning of the Minister’s reasons and evidence, logic dictates that he ought to bear the burden of proving that the statement is true in those proceedings.

Having considered the arguments, I am unable to agree with the respondent that the appellant should bear the burden of proof. My reasons are set out below.

First, despite the respondent’s argument to the contrary, I am of the view that s 103(1) of the EA in fact supports the appellant’s position. The starting point in this regard is Art 14 of the Constitution of the Republic of Singapore (1999 Reprint) (the “Constitution”), which provides that, subject to certain restrictions, “every citizen of Singapore has the right to freedom of speech and expression”. It is thus clear from the Constitution that the members and officers of the appellant who are citizens have a right to freely express their views. Section 103(1) of the EA provides that “(w)hoever desires any court to give judgment as to any legal right or liability, dependent on the existence of facts which he asserts, must prove that those facts exist”. Given that it is the Minister who is contending before the Court that the appellant’s constitutional right to free speech should be constrained by the CD because it has made a false statement of fact, I cannot see how s 103(1) of the EA assists the respondent. The constraint on the appellant’s right to free speech in the form of the CD would not exist but for the Minister’s attempt to impose it, and accordingly, it is the Minister who desires this Court to give judgment that the appellant’s rights should be curtailed. This curtailment may in turn give rise to legal liability on the part of the appellant should he not comply with the CD. Section 103(1) of the EA would thus suggest that it is for the Minister, who “desires (this) court to give judgment as to (the) legal right or liability” of the appellant, to prove that facts warranting the curtailment of the appellant’s rights exist, that is, that a false statement of fact has been made.

Second, I take the view that the POFMA Rules also support the appellant’s position. Rule 5 of the POFMA Rules provides that the appeal to the Court under s 17 of the POFMA is “by way of rehearing”. I take this to mean that this Court is not bound by the CDs issued, and can exercise its discretion “completely unfettered” by what had previously been decided: Lassiter Ann Masters v To Keng Lam (alias Toh Jeanette) [2004] 2 SLR(R) 392 at [15], which explains what a rehearing means in the context of Registrars’ Appeals. The determination by the Minister to issue a CD was an administrative (and non-judicial) decision, and accordingly, I am of the view that the rehearing referred to in r 5 of the POFMA Rules should involve a more significant degree of scrutiny than that provided for in a Registrar’s Appeal. If I were to agree with the respondent, it would mean that the Minister would succeed in a situation where neither party provides any evidence of truth or falsity simply and solely because of the Minister’s own earlier decision to cause the issuance of a CD. There would be no opportunity for scrutiny by the Court of the Minister’s decision that there was a false statement of fact. In that sense, the Court would in effect be fettered by the Minister’s earlier decision in issuing the CD, and that would not be consistent with the requirement that the Court should approach the matter “by way of rehearing”.

Third, I am not satisfied that Parliament intended that the appellant should bear the burden of proof. There is a clear information asymmetry between the Minister on one hand, and the maker of a statement being challenged under POFMA on the other. Unlike the Minister, who is able to rely on the machinery of state to procure the relevant evidence of falsity, the maker of a statement often has to contend with far more limited resources. For a statement-maker, who may be an individual, to bear the burden of proof would put him in an invidious position. While I accept that the existence of information asymmetry in itself should not in itself determine where the burden of proof lies, it does raise the question of whether Parliament could really have intended to place such an onerous burden on the statement-maker. Section 9A of the Interpretation Act (Cap 1, 1999 Rev Ed) provides that Parliament’s intention is relevant in interpreting a statute, but I do not see any indication that Parliament intended for the appellant to bear the burden of proof notwithstanding the potentially significant difficulty that would cause.

With respect, I am unconvinced by the respondent’s attempt to rely on the wording used in s 17 of the POFMA. I find that the reference to a “true statement of fact” is neither here nor there. The wording does not in and of itself set the burden of proof on the appellant to show that the subject statement is true. In my view, the reference to a “true statement” is simply to show that the Court must be satisfied that the statement of fact is “not false” in order to set aside the CD. If Parliament truly intended for the statement-maker to bear the legal burden of proof, surely the legislative draftsman could have expressed himself more clearly than by the use of the words in s 17(5)(b) of the POFMA. By extension, I find that the respondent’s reference to s 3(5) of the EA also does not assist, since I disagree with the respondent that the reference in s 17(5)(b) of the POFMA to a “true statement of fact” requires the appellant to prove the truth of his statement, rather than for the respondent to prove its falsity.

I am also unconvinced by the respondent’s argument that it is the appellant who, after learning of the Minister’s evidence of the falsity of the statement, must consider whether he still wishes to challenge the Minister’s decision and appeal against the CD. It is clear from my reading of the POFMA that there is no statutory duty on the Minister to provide evidence to show that he is justified in issuing a CD. Regulation 6 of the Protection from Online Falsehoods and Manipulation Regulations 2019 (S 442/2019) (“POFMA Regulations”) only provides that a CD must contain, inter alia, “the basis on which the subject statement…is determined to be a false statement of fact” (emphasis added). Regulation 6 therefore requires the Minister to give the basis, i.e., the reasons, for a subject statement being found to be a false statement of fact. That is not the same thing as providing the maker of the statement with evidence of the statement’s falsity.

A practical example may help illustrate the difference. Hypothetically, if one were to make a statement that 10% of all national servicemen suffer a serious injury during full-time national service and a Minister causes the issuance of a CD in relation to that statement, the basis or reason the Minister could give for the CD may simply be that it is not true that 10% of all national servicemen are seriously injured during their full-time national service because the actual figure is far lower. The evidence, by contrast, might then take the form of statistics or records from the Ministry of Defence to show that only a certain smaller proportion of national servicemen are ever seriously injured.

Thus, a Minister could comply with Regulation 6 of the POFMA Regulations but not actually present any evidence in support of his reasons. If there is a subsequent appeal under s17(1) of the POFMA, I am unable to see why the Minister should then be able to prevail even if he continues not to provide any supporting evidence of the falsity of the statement in question. The respondent’s argument that the appellant should bear the burden of proof means that, in a hypothetical situation where the Minister or respondent completely fails to provide any evidence of the falsity of a particular subject statement, the respondent could still succeed in having the appeal under s 17(1) of the POFMA dismissed. This is despite the fact that it is the appellant’s right to free speech which the Minister seeks to infringe upon.

Accordingly, I find that the respondent bears the burden of proof on the issues raised at [22] above.

Turning to the issue of the standard of proof, I note again that neither the POFMA nor the relevant secondary legislation prescribes the standard of proof. The appellant argues in its written submissions that the “POFMA is a penal statute” which is “professed to be directed towards the protection of public rather than private rights”. Pointing to the fact that the making of false statements could, and the failure to comply with a CD would, attract penal consequences, the appellant argues that the POFMA bears the hallmarks of a criminal process, and that the criminal burden of proof requiring the respondent to prove his case beyond reasonable doubt therefore applies.

With respect, I am unable to accept the appellant’s position. It is too blunt a claim to assert that the POFMA is a “penal statute”. While it has certain provisions which specifically create criminal liability like ss 7 or 15 of the POFMA, different requirements must be established before such liability attaches. The consequences of being issued a CD are not ipso facto penal in nature. It is specifically non-compliance with the CD rather than mere receipt of the CD that triggers criminal liability. This is nothing unusual since non-compliance with an order of court made in a civil case can also give rise to liability, which may result in a fine or imprisonment, but there is no suggestion that one has to meet a higher burden of proof in a civil case to obtain that order of court in the first place.

The appellant also asserts that the issuance of a CD “compels a statement maker to make a forced ‘confession’ of falsity in his statement”, and that this is “a penal sanction just as a fine or imprisonment is”. I disagree. There is simply no basis for the appellant’s conflation of the statutory obligation to put up a correction notice with a fine or imprisonment. The POFMA provisions have clearly delineated certain sections of the legislation as offence-creating, and the issuance of a CD in s 10 is not one of them.

Perhaps, a useful comparison in this regard is the power under s 15 of the Protection from Harassment Act (Cap 256A, 2015 Rev Ed) (“POHA”). Like the POFMA CD, s 15 of the POHA addresses the publication of false statements of fact, and provides that the maker of a false statement of fact may be compelled to publish a notification bringing attention to the falsehood and the true facts instead. There is no suggestion that an order made by the Court under s 15 of the POHA to publish a notification bringing attention to the falsehood in and of itself gives rise to criminal liability. In fact, s 15(3) of the POHA expressly states that the Court must be satisfied “on the balance of probabilities”.

Accordingly, for the reasons above, I find that the applicable standard of proof in these proceedings is on a balance of probabilities.

With these preliminary issues out of the way, I turn now to consider the three CDs in question.

CD-1 in relation to the SDP Article

As its first point, the appellant contends that CD-1 should not have been issued because it applied the POFMA retrospectively to the SDP Article. The SDP Article, having been published in June 2019, pre-dated the coming into force of the POFMA. The appellant thus argues that the POFMA cannot apply to the SDP Article as that would constitute retrospective application of the statute.

I am unable to agree. Section 3(1) POFMA makes clear that to “communicate” something within the meaning of the Act entails merely making information “available to one or more end-users in Singapore on or through the internet”. I find that hyperlinking falls within that broad definition because it makes the SDP Article easily available to readers. Hyperlinks to the SDP Article were included in both the November Facebook Post and the December Facebook Post. In fact, both Posts draw the reader’s attention to the hyperlinks and may be seen as inviting readers to access the hyperlinks and read the SDP Article, which is just one click away. I am thus of the view that the hyperlinking of the SDP Article suffices to constitute communication within the meaning of s 3(1) of the POFMA. This manner of communication via hyperlinking in both the November and December Facebook Posts was made after the entry into force of POFMA. As such, no issue of any retrospective application of the POFMA arises.

The position would be similar if the question was whether there has been re-publication of defamatory material in the law of defamation. It is fairly trite that the re-publication of a libel is a new libel, and re-publishers will be liable as if the objectionable statement originated from them: Gatley on Libel and Slander (12th Ed., Sweet and Maxwell) (“Gatley”) at [6.47], Low Tuck Kwong v Sukamto Sia [2014] 1 SLR 639 (“Sukamto”) at [39] to [41]. The question then is whether re-publication has taken place on the present facts by virtue of the appellant’s hyperlinking of the SDP Article in the Facebook Posts of November and December.

In Crookes v Newton [2011] 3 SCR 269 (“Crookes”), the Supreme Court of Canada held that the mere existence of a hyperlink within a document would not ipso facto mean that the document which could be accessed through the hyperlink was incorporated with the main document (which contained the hyperlink). However, McLachlin CJ and Fish J accepted that re-publication may have occurred if the combined text and hyperlink indicates “adoption or endorsement of the content of the hyperlinked text”. Abella J similarly observed that individuals may attract liability for hyperlinking if the manner in which content has been referred to conveys defamatory meaning. Deschamps J, who dissented on other points, was of the view that hyperlinking to defamatory content could satisfy the requirements of publication if “it makes the defamatory information readily available to a third party” (emphasis added). Gatley at [6.13] suggests that English courts may be willing to find that hyperlinking ipso facto gives rise to re-publication, at least in the absence of evidence to the contrary.

On the facts before me, I am of the view that this issue does not require detailed analysis of the applicability of the principles set out in Crookes in Singapore. In my view, it is clear from the November Facebook Post and the December Facebook Post that the appellant endorsed the material in the hyperlinked SDP Article and invited readers to access the said Article.

I am unable to accept the appellant’s argument that the hyperlinking was only to the “policy” aspects of the SDP Article. It is unreasonable to expect that only certain portions of the Article would be read, and it would be artificial to disaggregate the Article into purportedly different segments, only some of which would likely have been read but not others. From the entirety of the context in which the hyperlink was presented, I am satisfied that the hyperlinking of the SDP Article in both the November and December Facebook Posts would have sufficed to constitute re-publication of the whole Article.

I move now to the substantive questions in relation to CD-1 and the SDP Article. The first issue relates to the subject statement which may be identified from the content in the SDP Article.

The respondent claims that the subject statement that can be gleaned from the SDP Article is that “local PMET retrenchment has been increasing”. It sets out the following reasons this subject statement is the correct one:

(a) First, paragraph eight of the SDP Article expressly referred to “a rising proportion of Singapore PMETs getting retrenched.” This claim was immediately followed by the further claim that “[s]uch a trend is partly the result of hundreds of local companies continuing to discriminate against local workers”.

(b) Second, the context of the SDP Article framed a clash between foreign workers entering Singapore to “displace” local PMETs. The very premise of the SDP Article appeared to be to call for “reform”, “pushed for” by the appellant in the form of its “immigration policy”.

(c) Third, the context of the Article, and the specific text in paragraph eight, would cause the reasonable person to interpret the statements in the Article to give rise to the subject statement. In particular, concern might be caused to readers on whether the job prospects for local PMETs are deteriorating given the need for reform.

I will refer to the above interpretation as the “respondent’s primary case”.

In the alternative, the respondent argues that the other reasonable meaning which may be gleaned from the SDP Article is that the share of retrenched local PMETs as a proportion of all local PMET employees has been increasing. The respondent contends that this argument (the “respondent’s alternative case”) accounts for the appellant’s reference to “proportion” in paragraph eight of the SDP Article, and fits with the narrative conveyed by the SDP Article as a whole, which is that more and more local PMETs are losing their jobs.

As a further point, the respondent also argues that, as long as what has been identified is one reasonable interpretation, such that it could be said that a reasonable group of people would have adopted that interpretation, a CD can be issued if that reasonable interpretation is false. In other words, so long as the subject statement put forward is one reasonable interpretation out of many, this would mean that some people would have adopted that interpretation. And, if the claim in that interpretation is in fact false, a CD is warranted, notwithstanding the existence of other reasonable interpretations which may in fact be true.

In making this submission, DAG Nair argues on behalf of the respondent that the Court should not adopt the “Single Meaning Rule” from defamation law in the application of the POFMA. The Single Meaning Rule requires the determination of a single “correct” meaning conveyed to the hypothetical ordinary reasonable reader from the published material, and not a range of possible meanings. DAG Nair highlighted that the Single Meaning Rule has its historical origins in English jury trials for defamation (Charleston v News Group Newspapers [1995] 2 AC 65 at 72), and pointed to the widespread criticism that the Single Meaning Rule has received. In Ajinomoto Sweeteners Europe SAS v Asda Stores Limited [2010] EWCA Civ 609 (“Ajinomoto”), the Single Meaning Rule was criticised as being anomalous, otiose and potentially unjust at [31], and was further described as being “without enduring rationale” at [32]. Hence, in Ajinomoto, the English Court of Appeal rejected the rule as being applicable to claims for malicious falsehood.

Beyond the criticism of the Single Meaning Rule in the context of defamation, DAG Nair argues that the POFMA involves significantly different considerations as compared to defamation such that the Single Meaning Rule cannot be directly transposed into the POFMA context. In defamation, the Court adopts a single meaning in order to allow juries to quantify the loss of reputation and assess damages accordingly. On the other hand, the POFMA is concerned about ensuring the public is not misinformed about matters of public interest. It has a far broader scope than ordinary defamation law, as illustrated by how the statute at s 2(2)(b) of the POFMA provides for the many ways in which a statement’s meaning can be identified as being false: “whether wholly or in part, and whether on its own or in the context in which it appears”. This multiplicity of ways in which a statement’s meaning can be identified as being false sits uneasily with the Court having to identify only a sole meaning under the Single Meaning Rule.

DAG Nair therefore argues that the Court should, in the context of POFMA applications, adopt a “Multiple Meaning Rule” and determine if different meanings can arise from the published statements in the eyes of reasonable persons. According to DAG Nair, if one of the multiple reasonable meanings is false, a CD can be issued notwithstanding the fact that the other possible meanings are true. This would be in line with the policy intent undergirding the POFMA. It would also discourage those seeking to create mischief by making ambiguous statements that might mislead some segments of the public, but not others. In addressing the Court’s concern that this would permit the Minister to fixate on one meaning which is false out of many which are true, and on that basis activate the use of the POFMA, DAG Nair argues that, in reality, there would probably by only two or three possible reasonable meanings that can arise from any statement or depiction.

In summary, the respondent’s arguments on the appropriate subject statement operate on three levels. First, it argues that the meaning conveyed by the SDP Article in the minds of reasonable persons is that local PMET retrenchment has been increasing, in absolute numerical terms. Second, it argues as an alternative position that, if one is take into account the use of the word “proportion” in the eighth paragraph of the SDP Article, the other reasonable meaning of the Article is that the proportion of retrenched local PMETs as compared to all local PMETs has been rising. Third, the respondent contends that, if the words used in the relevant part of the SDP Article have different reasonable interpretations, then so long as one reasonable interpretation of the material is in fact false, a CD can be issued.

The appellant argues that the subject statement identified in CD-1 is not borne out by the SDP Article. The appellant instead contends that the subject statement that should be gleaned from the relevant part of the SDP Article is that there is a rising trend of Singaporean (as opposed to “local”, which would include Singapore Permanent Residents (“SPRs”)) PMETs retrenched relative to all retrenched locals. In other words, it is the proportion of Singaporean PMETs retrenched as a share of all locals retrenched that has been increasing. The appellant’s arguments in support of this interpretation are as follows:

(a) First, the appellant points to a report by the Straits Times and another by Yahoo! News to assert that “[I]t is obvious that our website article (based on the ST and Yahoo! Reports) refers to local PMETs retrenched as a share of all locals retrenched” (emphasis in the original). The Straits Times report is dated 15 March 2019, while that on Yahoo! News is dated 3 October 2019. It bears note that neither of these reports were cited or mentioned in the SDP Article.

(b) Second, the appellant asserts that its reference to “Singapore PMETs” in paragraph eight of the SDP Article should be interpreted as Singaporean PMETs to the exclusion of SPRs. In support of this, it points to the wording in paragraph eight, as well as its claim that, while actually meeting people and knocking on doors to speak to them, people understand “locals” as “Singaporeans” and not SPRs.

The appellant’s proffered meaning to be read from the relevant part of the SDP Article thus differs from the respondent’s case on meaning in two ways. First, the appellant argues that the subject statement should refer to “Singaporeans” only, and not “locals”, which it asserts would include SPRs. Second, the appellant rejects the respondent’s primary case that the Article suggests an increase in the absolute number of PMET retrenchments, and instead claims that the Article refers to Singaporean PMETs retrenched as a share of all locals retrenched.

Having set out both cases on the meaning to be derived from the relevant portion of the SDP Article, I pause to note that the proper determination of whether the subject statement can be derived from the communicated material under the POFMA is strictly a question of interpretation by the Court. The respondent has expressly accepted this in its affidavit, while the appellant impliedly acknowledged this over the course of oral arguments when inviting the Court to determine the correct subject statement to identify from a proper reading of the Article.

There is no prescribed test in the POFMA for the Court to apply in order to determine whether the interpretation or meaning can be derived from the communicated words or depictions in question. However, s 2(2)(b) of the POFMA may assist because it provides that a statement on its own can be false, without needing to read it as part of the whole. For ease of reference, s 2(2)(b) of the POFMA is set out below:

In my view, s 2(2)(b) of the POFMA appears to proceed on the basis that the reader may be taken to have read only parts of the article or even just one or two sentences, since the statement can be false “in part” or “on its own”, rather than in the context in which it appears. This suggests to me that the question of the meaning or subject statement to be identified is thus approached on a broader perspective than that in defamation law, where the publication has always to be considered in its entirety and in line with the “bane and antidote” rule. This well-known rule operates such that a potentially defamatory statement (the ‘bane’) in one part of a publication may be nullified by a statement in another part of the publication (the ‘antidote’) if that other statement has the effect of neutralising the defamatory statement: Sukamto at [44] and [45], and Gatley at [3.31]. The end result if the bane is neutralised by the antidote would then be that the publication is not defamatory as a matter of law. This does not appear to apply in the POFMA context – should a sentence be false, even on its own, it would be fall within the ambit of s 2(2)(b) of the POFMA.

Nonetheless, I think that some of the legal principles under defamation law for determining the meaning of the published statements are still relevant. The meaning falls to be determined by reference to what a reasonable reader would understand from the material in question. This is an objective assessment. In Jeyaretnam Joshua Benjamin v Lee Kuan Yew [1992] 1 SLR(R) 791 at [19], the Court held that, in determining the natural and ordinary meaning of words, “the sense or meaning intended by the (author) is irrelevant”; rather, “it is the natural and ordinary meaning as understood by reasonable members of the audience” which the Court has to assess. I am thus unable to accept the appellant’s contention that the subjective intention of the statement-maker is relevant, such that I had to consider whether the appellant intended to deliberately make a false statement. The appellant’s contention is also at odds with s 11(4) of the POFMA, which clearly states that a CD may be issued even if the statement-maker does not know or had no reason to believe that the statement was false.

I accept DAG Nair’s characterisation of the reasonable member of the audience, which is from defamation law. Gatley at [3.26] states that,

Put another way, the reasonable reader should be neither perverse, nor morbid, nor suspicious of mind, nor “avid for scandal”: Jeynes v News Magazine Ltd [2008] EWCA Civ 130 and Carter-Ruck on Libel and Privacy (6th Ed., LexisNexis) at [4.36] and [4.37]. I accept that these principles on the interpretation of words and depictions are of assistance to the Court in cases under the POFMA to determine the reasonable interpretation(s) of the communicated material.

Applying the abovementioned test, I find the appellant’s position on the subject statement to be derived from the SDP Article to be untenable. My reasons are as follows.

First, the appellant’s attempt to rely on the Straits Times article and the article in Yahoo! News to contour the interpretation of the SDP Article is not permissible. Neither article was referenced in the SDP Article, nor was there any attempt in the SDP Article to draw the reader’s attention to the two news articles. Accordingly, a reasonable person’s interpretation of the SDP Article would not have been shaped by these two news articles published six months apart, one of which (the Straits Times news article) was published more than eight months before the November Facebook Post which had a hyperlink to the SDP Article.

Second, the appellant’s claim that it was referring only to “Singaporeans” and not “locals”, which would include both Singaporeans and foreigners, is not tenable. In the appellant’s supporting affidavit filed by Mr John Tan Liang Joo (“Mr Tan”), under the section titled “Regarding Correction Direction 14 Dec 19 (website)”, he claims that its article was based on a Straits Times report of 15 March 2019. Mr Tan then proceeds to quote from the said Straits Times Report that “[PMETs] made up three in four or 76 per cent of the locals – Singaporeans and permanent residents – who were retrenched last year…” (emphasis original). Critically, Mr Tan added emphasis to the definition of locals as being “Singaporeans and permanent residents” by rendering the quotation in bold, as shown above. Given that the Straits Times report (on which the appellant claims to have based the SDP Article) relies on a definition of “locals” which refers to both Singaporeans and SPRs, and that the appellant has itself seen fit to highlight that definition to the Court, I do not think it is open for the appellant to now try and claim that the SDP Article should be interpreted as referring only to Singaporeans and not SPRs. In my view, the appellant has admitted that “Singapore PMETs” must, to an ordinary reasonable reader, include both categories of persons.

Similarly, I also cannot accept the appellant’s argument that their experience of meeting voters and citizens while knocking on doors contributed to them using the word “locals” as referring only to citizens and not SPRs. Such an argument relies on a subjective approach to interpreting the SDP Article which is at odds with the objective approach required under the POFMA (see [70]). I therefore cannot agree with the appellant’s attempt to constrain the meaning of “locals” to refer only refer to Singaporeans.

Even if I were to accept arguendo that the appellant might have been, at certain parts in the article, referring to Singaporean citizens only and not SPRs, the very first line of the SDP Article refers to the displacement of “local PMETs”. Similarly, paragraph eight of the Article refers to “Singapore PMETs getting retrenched”, not “Singaporean PMETs”. Instead, the distinction being drawn is with foreigners, or people who are not “Singapore PMETs”. I am thus not satisfied that the SDP Article would be read by a reasonable reader as having excluded references to SPRs who are also PMETs from its scope when it was discussing local PMET retrenchment trends.

Third, I found the appellant’s claim that the subject statement should be read as a reference to a proportion to all locals retrenched to be similarly problematic. The SDP Article does not itself say what the “rising proportion of Singapore PMETs getting retrenched” is a proportion of. Also, the crux of the SDP Article relates to PMETs, and the appellant’s attempt to argue that the appropriate comparator for the issue of proportion is “all locals” actually broadens the purported scope of the Article far beyond what its text would suggest.

Further, the appellant’s concept of proportion does not cohere with the message conveyed by the entirety of the SDP Article, which is that Singapore PMETs are getting retrenched and being displaced by foreign workers. If the appellant was intending to refer to the proportion of retrenched local PMETs to “all locals retrenched” as opposed to “all local PMET employees”, that figure would have been statistically unhelpful in showing how retrenched local PMETs are suffering or in a deteriorating position. DAG Nair posited a hypothetical which illustrated this. If there are three locals retrenched and one of them is a PMET, the proportion on the appellant’s case would be 1/3. However, if there are 10 locals retrenched, and three of them are PMETs, the relevant proportion would be 3/10, which is smaller than the former’s 1/3 despite the latter situation having a worse number of PMETs retrenched (3 as compared to 1) and a worse number of locals retrenched (10 as compared to 3). Thus, if I were to accept the interpretation of the SDP Article advanced by the appellant, I would have to accept a concept of proportion which would not make much sense to the ordinary reasonable reader, and hence would not have been in the mind of the said reader when he read the Article.

For the above reasons, I am of the view that the interpretation of the SDP Article which the appellant has invited me to adopt is not borne out on an objective reading of the said Article.

That said, I am also of the view that the respondent’s attempt to frame the subject statement as meaning that there is a rise in local PMET retrenchment in absolute numerical terms is also not supported by an objective reading of the SDP Article.

First, the key sentence relied on by the respondent in supporting its characterisation of the subject statement unequivocally contains the term “rising proportion” (emphasis added). In my judgment, the Court will be slow to adopt a meaning which ignores a term which the Article clearly includes, especially when the term in question is not an unduly technical word or jargon.

Second, while DAG Nair is right to underscore that the “reasonable reader” is likely to engage in some “loose thinking” (Ng Koo Kay Benedict v Zim Integrated Shipping Services Ltd [2010] 2 SLR 860 at [13], citing Review Publishing at [30]) and will not be conducting minute linguistic analysis of every phrase used (Jeyasegaram David (alias David Gerald Jeyasegaram) v Ban Song Long David [2005] 2 SLR 712 at [27]), I am unable to accept that this gives the respondent licence to ignore inconvenient words in the Article. In particular, while a layman would not understand words in the same way a professional might (Sukamto at [29]), the concept of a “proportion” is not a complex one and does not involve a convoluted deductive process based on interpreting a voluminous document (cf Lim Eng Hock Peter v Lin Jian Wei and another [2009] 2 SLR(R) 1004 at [92] and [101]).

Third, given the weight that DAG Nair invites the Court to attach to the specific word “amidst” in paragraph eight of the SDP Article (see [94] below), I found it rather inconsistent that the respondent is inviting me to effectively “read-out” the word “proportion” in its framing of the subject statement.

Accordingly, I am not prepared to accept the respondent’s primary case as the meaning which arises arise from an objective and reasonable interpretation of the SDP Article.

Turning to the respondent’s alternative case, however, I am of the view that it represents a reasonable interpretation of the relevant portion of the SDP Article.

First, it accounts for the appellant’s use of the word “proportion” in paragraph eight of the SDP Article. Critically, it coheres with the broader sentiment evoked in the article, which is that local PMETs should be concerned about more of them being retrenched. An ordinary reasonable reader can discern the message that more local PMETs as a proportion of the total number of local PMETs employed have been retrenched. Thus, the respondent’s alternative case does not involve the “reading out” of any part of the SDP Article, and identifies a logically defensible comparator.

Second, I also accept the argument by the respondent that, by looking at the share of retrenched local PMETs vis-à-vis the number of all local PMET employees, one can gain a meaningful statistical indicator as to the vulnerability of local PMETs to being displaced and losing their jobs. An ordinary reasonable reader can make sense of the proportion and hence understand the statement being conveyed that more local PMET jobs are in danger of being lost. That being the case, it is more likely that the reader would come away from reading the SDP Article with such an interpretation.

For the above reasons, I accept the respondent’s alternative case as being a reasonable interpretation of the relevant portion of the SDP Article.

Thus, of the three interpretations of the SDP Article put forward, I am of the view that the respondent’s alternative case is the right one. The other two interpretations are, in my view, flawed for the reasons outlined above. Given the absence of other reasonable interpretations of the SDP Article raised before me, there is strictly no need for me to consider whether the Multiple Meaning Rule as argued by the respondent applies in the POFMA context. However, having heard the arguments on this issue, I must say that my preliminary view is that the Single Meaning Rule which applies in defamation law may not be directly applicable in the POFMA context because the underlying rationale of the legislation is quite different. Nonetheless, given the importance of this point, the absence of full arguments on this legal issue because the appellant was unrepresented by counsel, and the fact that it is not essential to my decision on the facts, I do not come to any definitive conclusion about the applicability of the Multiple Meaning Rule.

Having established the subject statement that can be derived from the SDP Article should be, I turn to consider the truth or falsehood of the said statement.

The respondent refers to statistics from the Ministry of Manpower (“MOM”) in support of its argument that the number of retrenched local PMETs per 1,000 local PMET employees has decreased from 2015 to 2018. The statistics were sourced from Labour Market Surveys conducted by MOM over the relevant timeframe.

The respondent also provided statistics in tracking the number of retrenched local PMETs in absolute terms from 2015 to 2018, though this was not strictly relevant given my findings as to the appropriate subject statement above.

The critical factor on the question of truth or falsity is that the appellant did not challenge the veracity of the statistics put forward by the respondent as evidence of falsity. While it did call on the respondent to make further data which pertained solely to Singapore citizens available, the appellant did not at any point assert, or even suggest, that the statistics were inaccurate. Rather, the appellant’s approach to the issue of truth and falsity was to challenge the respondent’s reliance on the statistics in two regards:

(a) First, the appellant alleges that the respondent is “cherry-picking” its statistics by showing figures for retrenched local PMETs per 1,000 local PMET employees from only 2015 to 2018. The appellant provided its own graph showing “PMET retrenchment per 1000 PMET” which spanned 2010 to 2018 and which illustrated a statistical increase from 2010 to 2018. The appellant further argues, by reference to ministerial statements from the period around 2010, that the issue of foreigners displacing Singaporean PMETs had started to cause comment from government officials around 2010, and that the relevant trend thus fell to be assessed from 2010 instead of 2015.

(b) Second, the appellant argues that, even if the matter fell to be assessed based on the most recent statistics, these show an uptick of retrenchments of locals from 2,510 in the fourth quarter (“Q4”) of 2018 to 3,230 in the first quarter (“Q1”) of 2019.

I am unconvinced by the arguments made by the appellant. First, the appellant’s argument that the relevant data should stretch back to 2010 is problematic in a number of regards. As emphasised by the respondent, the appellant described its policy proposals as coming “amidst” a rising proportion of PMET retrenchment. This suggests an urgency and contemporaneity with the time of the communication of the subject statement (i.e., November 2019), which appears to be discordant with its approach of now referring to such a broad swathe of time, that is, from 2010 when considering the evidence of the truth of the statement.

Further, there is nothing in the SDP Article which offers any hint that the “rising proportion of PMET retrenchment” should be understood with such a long timeframe rather than by reference to the most recent period of time. The appellant relies on reference to external ministerial statements from 2010 to justify that time period, but it would be difficult to imagine that the average reasonable reader of the Article in November 2019 would be cognizant of (or remember) those statements and have them in mind when they read the Article.

Third, even if one takes the appellant’s case at its highest and relies on the data from 2010 to 2018, it bears note that the appellant’s reference to 2010 as a starting point is somewhat arbitrary. Understandably, any timeframe may, to some extent, be criticised as being arbitrarily selected. However, the appellant’s own reasoning could warrant a starting point of 2009, or even 2008. Using 2009 as a starting point would show an overall decrease in unemployed local PMETs from 2009 to 2018, and this illustrates how using 2010 as a starting point is itself quite arbitrary.

Fourth, turning to the appellant’s reliance on Q4 2018 and Q1 2019 data, those figures are not applicable to the subject statement identified above. As accepted by the appellant, the quarterly data refers to the overall number of retrenchments in Singapore. This is not limited to “locals” or even “Singaporeans”, and is also not limited to PMETs. Even on the appellant’s own case of what the subject statement should be, these quarterly figures would not be applicable since they refer to a fundamentally different data set.

Fifth, the definition of a “false statement” within the meaning of POFMA also includes misleading statements: see s 2(2)(b) of POFMA. The relevant sentence in the SDP Article did not include or specify any timeframe. Instead, it used the word “amidst”, as already referred to above. I am of the view that the subject statement is misleading, at least in part, in the context in which it appears given that it does not mention that the data from the most recent period of time (2015 to 2018) actually contradicts the statement that local PMETs’ retrenchment rate has increased.

I pause to note at this point that the appellant has accused the respondent of seeking to foist a particular date range (2015 to 2018) on the data. I am of the view that this accusation is a misunderstanding of the respondent’s approach to this case. The respondent is merely seeking to explain how the recent and contemporary statistical data supports the Minister’s position, and has acknowledged that it is simply presenting a few years’ data. The respondent has a reasonable basis, by virtue of the appellant’s use of the term “amidst” to describe the retrenchment trend, to focus on a more recent timeframe. On the other hand, the appellant has no basis to place reliance on data from a timeframe stretching back to 2010.

In sum, I am satisfied the respondent has adduced sufficient evidence to show that the subject statement as interpreted above is false on the balance of probabilities. I note once again that the evidence relied on by both parties was largely the same, in the form of MOM’s statistics. The contention which arose was over the relevant timeframe, and I am of the view that the respondent’s characterisation of the relevant timeframe was more consistent with what could be reasonably interpreted from the relevant part of the SDP Article.

In my judgment, the respondent has met its burden of proof in relation to the SDP Article. There is therefore no basis to set aside CD-1.

I have a final observation in relation to CD-1, which would also apply to CD-2 and CD-3 since the latter two CDs are also concerned about the hyperlinking to the SDP Article. CD-1 and its correction notice arguably does not appear to directly address the subject statement for the SDP Article, as determined in my judgment, which is the respondent’s alternative case. This is perhaps unsurprising, given that CD-1 presumably proceeded on the basis of the respondent’s primary case.

I note, however, that the Factually Website linked in the correction notice does include the relevant data on the number of retrenched local PMETs per 1,000 local PMET employees, and that this fact directly addresses the subject statement identified by the Court. Hence, the appellant having perused the data in the linked Factually Website would understand the identified subject statement in CD-1 (“Local PMET retrenchment has been increasing”) and the correction notice (“There is no rising trend of local PMET retrenchment”) as referring also to the proportion of local PMET retrenchment as a share of all local PMET employees. A member of the public reading the correction notice together with the data found on the Factually Website would also have that same understanding. As such, I am of the view that the CD-1 and its correction notice are, on balance, worded wide enough to address the subject statement that has been found by the Court to arise from the SDP Article.

Nevertheless, the Minister may still wish to consider whether to exercise her power under s 19 of the POFMA, at her own initiative, to vary the CDs and correction notices to make the identification of the subject statement clearer. In any event, there is nothing in the POFMA which provides that a CD may be set aside because the subject statement has not been set out in the CD with sufficient precision. In any event, the appellant also did not seek to argue otherwise.

CD-2 in relation to the November Facebook Post

CD-2 relates to the November Facebook Post. The issue being taken with the Post is its inclusion of a hyperlink which links to the SDP Article. The contention raised by the respondent is that, insofar as the November Facebook Post re-communicates the SDP Article, and the SDP Article contains a false statement of fact, the Minister was entitled to issue CD-2. The respondent’s argument is that the falsehood in the SDP Article has been ‘imported’ or ‘incorporated’ into the November Facebook Post.

The appellant broadly accepts that, given the respondent’s approach to the false statement of fact in the November Facebook Post, the defensibility of CD-2 would turn on whether CD-1 was defensible.

The appellant did argue that the attempt to frame the alleged false statement of fact as being the hyperlinked SDP Article was an afterthought, and that the false statement of fact initially alleged by the respondent related to the text in the November Facebook Post that “local PMET unemployment has increased”. As its basis, the appellant refers to the fact that, in the affidavit filed by the MOM officer in these proceedings, a copy of the November Facebook Post was exhibited that showed a red box drawn around the text “local PMET unemployment has increased”.

I cannot accept the appellant’s argument. First, the respondent has already clarified in the MOM officer’s affidavit that the Minister’s basis for CD-2 stemmed from the hyperlinking to the SDP article. Second, the red box drawn around the text “local PMET unemployment has increased” in the copy of the November Facebook Post exhibited in the respondent’s affidavit had been for internal use, according to DAG Nair. This copy was not included in any prior correspondence to the appellant. Therefore, the rendering of the November Facebook Post in the respondent’s affidavit could not have influenced the appellant’s understanding of CD-2 at the time it filed this appeal. Third, the text of the CD pertaining to the November Facebook Post was in pari materia with that of CD-1 which relates only to the SDP Article, thus illustrating that the two CDs were to deal with the same concern. There is thus simply insufficient basis for me to accept that the respondent had attempted to identify a falsehood ex post facto.

Overall, I reiterate the view that the hyperlinked SDP Article contains the subject statement dealt with in relation to CD-1 above. I find that the hyperlinking of the SDP Article in the November Facebook Post thereby re-communicated the Article’s content. I am therefore satisfied that the November Facebook Post contains a false statement of fact, and that there is accordingly no basis to set aside CD-2.

CD-3 in relation to the December Facebook Post

As mentioned, the respondent has identified two subject statements in the December Facebook Post. One relates to a hyperlink to the SDP Article, while the other relates to a graphical illustration which shows a downward arrow under the heading “local PMET employment”.

It bears note at this point that, under the POFMA, given that more than one subject statement is identified, a CD can be issued in respect of the December Facebook Post even if only one of the subject statements identified from the communicated material is shown to be false.

Turning to the subject statement that “local PMET retrenchment has been increasing”, the false statement of fact alleged is identical as that in the SDP Article. The hyperlinking of the SDP Article in a Facebook post has also been considered in the analysis on CD-2 above. Given that this subject statement is the same as for CD-1, I reiterate my view that the falsehood in the SDP Article was communicated via the December Facebook Post by reason of the hyperlink. Accordingly, there is a proper basis for CD-3 on that ground.

In relation to the subject statement concerning the graphical illustration, the respondent’s case is that it means that “local PMET employment has gone down”. This interpretation stems from the downward-pointing graph, as well as the label applied to the graph.

In response, the appellant argues that the graph, and the December Facebook Post as a whole, would lead a reasonable reader to believe that the proportion of Singaporean (and not local) PMET employment has decreased. Although it is not entirely clear, the appellant appears to be referring to the proportion of Singaporean PMETs as compared to all other PMETs in Singapore. The appellant puts forth the following reasons in support of that interpretation:

(a) First, the December Facebook Post contains numerous references to “Singaporeans”, from the very first sentence down to the call for a “more measured approach towards allowing foreigners to work in Singapore”. Accordingly, the appellant argues that a reasonable reader would construe the word “local” in the graph’s heading to mean Singaporeans and not include SPRs.

(b) Second, the appellant asserts that the graph must refer to the falling proportion of Singaporean PMETs being employed because any other interpretation would be “misleading” and “not take into account the huge population increase fed by a large increase in the number of foreigners into the ‘local’ population”.

I am unable to accept the appellant’s interpretation that the meaning conveyed by the graphical illustration concerns a “proportion”. For the December Facebook Post, the issue being taken by the Minister in relation to this particular subject statement is the graphical illustration, and that alone. Hence, the Court’s attention must be focused on what meaning that graphical illustration conveys to a reasonable reader. I make reference again to s 2(2)(b) of the POFMA, which allows the Minister to act even if one part of a communicated statement, read on its own, is false. The appellant’s proposed interpretation requires me to read into the downward pointing arrow depicted in the graphical illustration a reference to a proportion of Singaporean PMETs when there is simply nothing in illustration itself which would give rise to that inference. Unlike the SDP Article, where the word “proportion” is expressly used, I am invited to effectively insert the word “proportion” into the graphical illustration when there is no apparent basis to do so. For that reason, I do not accept that a reasonable reader would read the graphical illustration as indicating a “proportion”.

Further, the appellant’s attempt once again to disaggregate “Singaporean” and “local” PMETs is, to my mind, problematic. While I accept that the term “Singaporean” is used in the December Facebook Post, a reasonable reader is likely to understand the phrase “local PMET employment” to refer to the employment rates of local PMETs (including SPRs) because the publicly available statistics on local PMET employment in official data published by the MOM refer to both Singaporeans and SPRs when representing “local PMET” employment figures. There is in fact no official data or statistics which the appellant can point to in the evidence in these proceedings which disaggregates the term “local” in the way that it invites the Court to accept, and it thus seems odd that the appellant nonetheless seeks to persuade the Court that “local” should be understood in the manner it advances.

The juxtaposition of “Local PMET employment” against “Foreign PMET employment” in the December Facebook Post, which is shown in the adjacent graphical illustration as having an upward pointing arrow, would also be difficult to explain on the appellant’s reasoning. If SPRs were in fact a wholly distinct category from “local” PMETs, then the decline in “local” (Singaporean-only) PMET employment might be explicable by a rise in SPR PMET employment, not necessarily a rise in “Foreign PMET employment”. There is, after all, no suggestion from the December Facebook Post that “Foreign PMET employment” includes SPRs as well. Hence, a reasonable reader would take away the message from the two graphs placed side by side that local PMET employment, which includes SPRs, is decreasing while at the same time, foreign PMET employment is increasing. The first part of this comparison is precisely the subject statement argued by the respondent as arising from the Post.

In addition, the appellant’s own affidavit acknowledges that there has been a population increase “fed by a large increase in the number of foreigners into the ‘local’ population”. The term “local” was placed in quotation marks by the appellant, and this suggests a recognition of the fact that the “local population” can be increased by SPRs. The appellant’s own use of quotation marks for the term “local” illustrates that it has to clarify the meaning of “local” it relies on, given the possibility that a reader may misunderstand the term. If there were as much of a bright-line distinction between “locals” (meaning only Singaporeans) and “non-locals” (including SPRs) and that distinction was in fact commonly-held, there would be no need for the appellant to take pains to explain the term. In other words, I do not find that the ordinary reasonable reader would adopt the appellant’s distinction between Singaporeans and SPRs as readily as the appellant suggests.

On the whole, I cannot agree with the appellant’s interpretation, which would require me to (a) read a concept of proportion into what the graph depicts, and (b) distinguish between “Singaporeans” and “SPRs” in the term “local” when there is little basis from the Post to do so.

Having addressed the issue of the appropriate interpretation of the graphical illustration in the December Facebook Post, I turn to the issue of falsity. The respondent relies on data from 2015 to 2018 to show that the number of local PMETs employed, in terms of absolute numbers, has been steadily increasing. The appellant does not take issue with the veracity of this data. I should add that the respondent has, in seeking to refute the appellant’s interpretation of the graphical illustration in the December Facebook Post, also referred to data from those same years that show that local PMETs employed as a proportion of all local employment has been increasing. However, this latter data is not strictly relevant given my findings as to the meaning to be derived from the graphical illustration.

The appellant contends that the graph cannot be said to be “false”, and makes two arguments in support of this. First, the appellant argues that since there is nothing on the axes of the graph showing the timeframe, a broad timeframe should be adopted in ascertaining whether there is a falling proportion of Singaporean PMETs. The appellant specifically contends that the data provided by the respondent in its CD, which was from 2015 to 2019, is from “too narrow” a timeframe.

Further, the appellant presents three sets of figures from which it invites the Court to conclude that “the proportion of Singaporean PMET employment has gone down”. The figures that the appellant highlights are (i) the drop in labour force participation rate of residents aged 15 and over, (ii) the number (in absolute terms) of the foreign workforce holding either an Employment Pass or an S Pass, and (iii) the number of Permanent Residencies granted between 2015 and 2018. All three sets of figures adopt the 2015 to 2018 timeframe used by the respondent. The appellant’s reasoning is that since (i) there has been a drop in labour force participation rate of residents aged 15 and older, (ii) there has been an increase in the number of the foreign workforce holding either an Employment Pass or an S pass, and (iii) there has been an increase in the number of Permanent Residencies granted, the proportion of Singaporean PMET employment must have gone down. The Court is invited to draw that conclusion based on the cumulative weight of all three sets of figures.

I am unable to accept the appellant’s arguments, even if one is to leave aside the fact that the appellant is attempting to justify a meaning (in relation to “proportion”) which I find is not borne out by the graphical illustration. I do not see any reason why the appellant, who has chosen not to label the time period for its graph, should have carte blanche to assert any timeframe of its choosing as being the applicable one. I note that the appellant has, in the statistics that it relies on, itself used the time period from 2015 to 2018. I am of the view that the time period from 2015 to 2018 relied on is a reasonable one, and agree with the respondent that the ordinary reasonable person reading the December Facebook Post would interpret the graph to be reflecting a present troubling trend which the SDP’s policy proposals were aimed at addressing.

Further, I do not see how the three data sets cited by the appellant lead inexorably to the conclusion that it seeks to draw. Setting aside questions of statistical significance, I note that the variables used in the three data sets are all quite different from the conclusion the appellant is inviting me to draw. For example, the appellant’s reference to labour force participation is a distinct issue from employment (or retrenchment) rates. Similarly, the appellant’s reference to “number of PR granted between 2015 and 2018” is a blunt instrument which does not disaggregate that number into PMETs and non-PMETs. I am thus not satisfied that the three data sets cited by the appellant detract from the directly applicable evidence in the affidavit filed by the MOM officer that shows the falsity of the subject statement that local PMET employment is decreasing. Again, I reiterate that the appellant does not challenge the MOM’s statistics in this regard as being inaccurate. Accordingly, I find that the second subject statement for the December Facebook Post has been shown to be false.

On balance, I am satisfied that there is proper basis for CD-3, both in relation to the hyperlink to the SDP Article, and the graphical illustration on local PMET employment. Accordingly, I see no basis to set CD-3 aside.

Conclusion

I am of the view that subject statements which were identified from the SDP Material are in fact false in the face of the statistical evidence against them. I reiterate that the appellant has not challenged the accuracy of the statistical evidence, and has instead sought to critique it on other grounds. I do not find those grounds convincing, and thus find that the respondent has discharged its burden of proof. There is thus no basis for me to set aside the CDs.

I would emphasise that the role of the Court in this context is to interpret the legislation, not to comment or adjudicate on the desirability of particular policies. In that sense, the Court is constrained by what the legislation compels. Where there is doubt as to the precise ambit and contours of the legislation, the ordinary rules of statutory interpretation apply.

I pause here to highlight that both parties attempted to cast aspersions on each other’s intentions and motivations, with labels such as “disingenuous” and “dishonest” being bandied about. I underscore that the POFMA necessitates an objective approach based on the wording of the material in question. The issues are whether the subject statement(s) are borne out by the words and/or depictions in the communicated material, and then whether those subject statement(s) are true or false. The intentions of the parties in relation to the POFMA are thus, sensu stricto, irrelevant when there is no question before the Court of any criminal liability.

For the reasons given above, I dismiss the Originating Summons.