CENTA-CITY INDEX CO LTD AND OTHERS v. HONG KONG ECONOMIC JOURNAL CO AND ANOTHER

CACV 232/2014

IN THE HIGH COURT OF THE

HONG KONG SPECIAL ADMINISTRATIVE REGION

COURT OF APPEAL

CIVIL APPEAL NO. 232 OF 2014

(ON APPEAL FROM HCA NO. 237 OF 2011)

________________________

________________________

________________________

J U D G M E N T

________________________

Hon Kwan JA:

1.  On 22 October 2014, Suffiad J handed down judgment in an action for defamation. The plaintiffs sued on two articles published in the Hong Kong Economic Journal on 16 and 17 December 2010 (“the 1^st Article” and “the 2^nd Article”). The defendants raised the defences of justification and fair comment. The judge held that fair comment was established in respect of the 1^st Article and, should it be necessary, he was prepared to hold that the words complained of in the 1^st Article bear a lesser defamatory meaning as pleaded in the defence and the defendants could also establish the defence of justification. However, he rejected the defences of fair comment and justification for the 2^nd Article. He awarded damages of $300,000 for the libel in the 2^nd Article and made an order nisi that the plaintiffs should have two-thirds of their costs.

2.  The only substantial issue in this appeal is whether the judge was right in holding that the defamatory imputation of “出術” (cheating) in the 2^nd Article is a statement of fact or in the nature of comment.  If it is a statement of fact in the proper context of the publication, as held by the judge, the defendants cannot rely on the defence of fair comment, and would have to prove that the defamatory imputation is true in substance and in fact, which they are not able to do. In other words, they would be thrown back to the defence of justification, which the judge has ruled against them.

Background of the publication

3.  I will first relate the background to the publication of the articles.

4.  The 2^nd plaintiff (“Centaline”) is a well-known real estate agency in Hong Kong.  The 1^st and 3^rd plaintiffs are companies in the same group.  The 1^st plaintiff computes, compiles and owns the Centa-City Index (“CCI”) and the Centa-City Leading Index (“CCLI”) (collectively “the Indices”).  The 3^rd plaintiff manages the website in which the Indices are published.  The Indices were jointly developed by Centaline with the City University of Hong Kong in 1998 to reflect residential property price movement in Hong Kong.  City University sold the intellectual property rights of the Indices to Centaline in 2008 and two professors from that university became Centaline’s consultants in relation to the Indices.

5.  CCI is a monthly index of property prices based on data obtained from transactions of constituent housing estates registered in the Land Registry.  CCLI is a weekly index of property prices based on data from provisional agreements in transactions of constituent housing estates handled by Centaline.  As stated in Centaline’s website, the CCLI, published every Friday, reflected the secondary residential property prices for the two week period ending five days before the publication date.  At the material time, the Indices were based on 86 constituent housing estates.

6.  On 19 November 2010, the Hong Kong Government announced new administrative measures which included a special stamp duty to be effective from 20 November 2010 to curb property speculation of the residential property market.  The public sentiment was that property prices would fall as a result.  However, the CCLI published in the next three weeks were all on the rise.

7.  On 14 December 2010, an article appeared in Ming Pao on the apparent anomaly in the rise of the CCLI and reported that the Government had contacted Centaline to find out how the CCLI was computed and why it showed an increase.  On the same day, Shih Wing Ching of Centaline and Professor Chan Yan Chong of City University commented on the subject in their respective columns in AM 730 and Apple Daily.  Shih stated that the trend shown by the CCLI was correct because front-line agents of Centaline had observed a price increase.  Professor Chan, who was one of the founders of the CCLI, indicated that he was unable to provide a reasonable explanation for the rise in the CCLI over the past two weeks.

8.  On 15 December 2010, a further article was published in Apple Daily referring to the apparent anomaly in the CCLI and a number of ambiguous and inconsistent explanations provided by Centaline, Shih and the City University.  It mentioned that a qualification was added in Centaline’s website on 10 December 2010 that the CCLI was based on “anticipated dates of formal agreements for sale and purchase, whereas there is usually a 14 days’ time lag between the formal agreements for sale and purchase and the preliminary agreements for sale and purchase”.

9.  Obviously, the rise in the CCLI against the Government measures to curb property speculation was a matter of great public interest at the time.  It was against this background that the 1^st and 2^nd Articles were published in the Hong Kong Economic Journal on 16 and 17 December 2010.  They were written by a free-lance writer Law Ka Chung under the pen name “Law Gang”.  They appeared in the column “Finance DNA”, which was in Section B of the newspaper.  Section B was devoted to information on stock prices, funds, financial derivatives and the like with articles containing analysis and comments on finance and business.

The 1^st Article

10.  The 1^st Article appeared under the heading “編樓價？收檔啦！” (Compiling property prices? Pack up!).  The certified translation reads as follows:

“<Ming Pao> on Tuesday reported that “The Centa-City Leading Index, which is always “considered” to be most capable of reflecting the instant property market condition, showed a 1.33% “accumulative increase” in secondhand property prices for 2 consecutive weeks after the Government has imposed tough measures by levying special stamp duty, its performance was quite contrary to the general public’s understanding. Insiders revealed to our newspaper that in fact the Transport and Housing Bureau telephoned Centaline Property Agency last Friday to enquire about the calculation methodology of the Index with a view to being able to know why the Index was different from the market trend”.

The Index was set up jointly by Centaline and City University. On the same day Shih Wing Ching of Centaline stated in <am 730>  <C View> “Concerning how the “Centa-City Index” is calculated, ‘professors from City University have explained clearly in the website, interested readers can take a look on the website.” However, Chan Yan Chong of City University who was involved in founding the Index, stated on the same day in <Apple Daily> <Chan Yan Chong’s Column>: “several reporters asked me why? I am also unable to provide a reasonable explanation…”

It seems that everyone is passing the buck to each other and prevaricating. I always consider these indices in the market not reliable. I therefore visited City University’s website to check out for this “clear explanation”. In fact the Index only covers 50 estates. How many estates are there in Hong Kong? Just grab a book of maps and flip it through, I am afraid there are not only 500 [estates]. Even if the 50 [estates] are all large estates, then how representative could this be? Not mentioned; what kind of 50 [estates]? Also not mentioned. It is surprising that a university do things in such a way and there is not even a single word mentioned about the samples.

What is more ridiculous is that, on the compilation methodology, there is surprisingly only one sentence ‘Regression models are used’. What regression? As for me even I am [only] writing in newspaper, I would still set out the entire regression formula. A professor did regression and yet it could not be revealed to the public. Pack up.

Note: The so-called Centa-City Leading Index calculation methodology of City University:

http://fbweb.cityu.edu.hk/ms/cci/content.htm

lawgang@gmail.com”

The judge’s rulings of fair comment on the 1^st Article

11.  The particulars of comment in the 1^st Article pleaded in the defence are:

(1)  “編樓價？收檔啦！” (Compiling property prices? Pack up!)

(2)  “似乎大家都在互相推搪，含糊其辭。” (It seems that everyone is passing the buck to each other and prevaricating.)

(3)  “更混帳的是編制方法竟一句「Regression models are used」便算。什麼regression? 在下寫報紙尚將regression全條寫出，堂堂教授做regression竟然不見得光？收檔啦。” (What is more ridiculous is that, on the compilation methodology, there is surprisingly only one sentence ‘Regression models are used’.  What regression?  As for me even I am [only] writing in newspaper, I would still set out the entire regression formula.  A professor did regression and yet it could not be revealed to the public. Pack up.)

12.  It is the plaintiffs’ case there were three main stings in the 1^st Article:

(1) the words “收檔” (pack up), connoting the compilation of the Indices was invalid and unreliable, and that the plaintiffs should cease doing so;

(2) the words “互相推搪” (pass the buck), connoting the shifting of blame or burden; and

(3) the words “不見得光” (cannot be revealed), connoting knowing concealment of matters that could not see the light of day.

13.  The judge noted these four elements to the defence of fair comment:

(1) the comment must be on a matter of public interest;

(2) the words used are comment and not fact;

(3) the facts on which the comment is based are substantially true or protected by privilege; and

(4) the comment is one that any man, however prejudiced and obstinate, could honestly hold.

14.  The judge found element (1) was satisfied, as there can be no question but that the subject matter of the 1^st Article related to a matter of public interest.

15.  The judge held element (2) was also satisfied.  First of all, he noted that the 1^st Article was a commentary article in a commentary section of the newspaper, and thus readers are likely to understand the words as being comments made by the author.  As to each of the three stings, he was satisfied that the words which formed the heading of the article “Compiling property prices? Pack up!” are not facts but comment.  They may be said to be criticism, remark or observation, but certainly not facts.  The words “It seems that everyone is passing the buck to each other and prevaricating” were a conclusion drawn by the author from the articles published by Shih and Professor Chan, and from the fact that the City University website failed to provide a formula for the methodology of calculating the CCLI.  As for the last paragraph beginning with the words “What is more ridiculous” and ending with the words “Pack up”, it was the criticism or observation by the author of the failure to set out the compilation methodology used in the regression model, and could only be comment not fact.

16.  In respect of element (3), the judge found that the facts on which the comments were based are substantially true, as those facts were based on the factual background leading to the publication of the 1^st and 2^nd Articles, and the factual background was not in dispute.

17.  For element (4), the judge found that the author honestly held the views that the CCLI was not reliable, that Shih and Professor Chan were shifting responsibility to one another and prevaricating with neither willing to disclose the details of the methodology used in the regression model, and that there was a lack of clarity and transparency in the calculation methodology for the CCLI.  While the words “Pack up” may be thought to be overly harsh, having come to the conclusion that the author did honestly hold the views in the comments expressed, in the words of Lord Nicholls in Cheng & Anr v Tse Wai Chun (2000) 3 HKCFAR 339 at 347J, “a critic need not be mealy-mouthed in denouncing what he disagrees with. He is entitled to dip his pen in gall for the purposes of legitimate criticism”.

18.  He therefore ruled that the defence of fair comment was made out for the 1^st Article.

The 2^nd Article

19.  The 2^nd Article was a follow up to the 1^st Article.

20.  As context is of critical importance to the principal issue on appeal, namely, whether the defamatory imputation is a statement of fact or comment, I will set out the entire article in Chinese and its certified translation.  Those parts in the article pleaded in the defence as particulars of comments are printed in bold and I have added a number to the end of each for ease of reference.

21.  The 2^nd Article reads:

“指數可造市            (1) 羅耕 / 財經DNA

叫人收檔，不是隨便叫的(2)。昨文提到中原與城大合辦了一個樓價指數，當被發現數據有可疑後，兩邊負責人互相卸膊(3)。中原姓施的叫你找城大，城大姓曾的答不知，而城大「中原指數專頁」在編制方法上含糊其辭，還有條連結帶你「重返中原」，真真有趣(4)。

這個指數，在樓市上的地位等同港股裏的恆生指數。假使恆指公司如斯含糊其辭，你猜後果如何？會否被人指控造市？港府監管金融有道，但面對樓市出術卻頓成廢柴(5)。

說其出術不是亂說的。「這檔嘢」還搞了個負擔比率。在下當然抱著懷疑態度(6)，按國際慣例計算負擔比率，將差餉物業估價署的樓價指數除以統計處綜合住戶統計調查的私樓（不包公屋）住戶入息中位數（兩個都是公開資料），再乘個 factor (40000/3) ，即得與「這檔嘢」頗似的負擔比率。如果你信他們，目前樓價僅相當於2000年水平，不買就笨；但若信官方數據，目前樓價其實是等同1997年5月見頂前幾個月的水平，買才呻笨(7)。

如斯分歧，絕不可能是誤差。有心也好無意也罷，這個指數已有造市效果(8)。

lawgang@gmail.com

中原與城大合辦了一個樓價指數，當被發現數據有可疑後，兩邊負責人互相卸膊 (9) (網上圖片)。”

22.  The certified translation reads:

“Index can manipulate the market        (1) Law Gang Finance DNA

To tell people to pack up, [one] doesn’t lightly say so(2). Yesterday’s article mentioned that Centaline and City University had jointly set up a property price index, when it was found that the the data were dubious, the persons-in-charge of both sides passed the buck to each other(3). The one surnamed Shih of Centaline told you to contact City University, the one surnamed Chan of City University said he had no idea, and the compilation methodology stated in City University’s ‘Centa Index Special Page’ is ambiguous in its expressions, and there is even a link which brings you ‘back to Centaline’, really really interesting(4).

This Index, its status in the property market is equivalent to Hang Seng Index in Hong Kong stock [market]. If that the Hang Seng Index company prevaricated as such, what do you think the consequences would be? Would it be accused of manipulating the market? The Hong Kong Government monitors the financial [market] well, but when faced with cheating in the property market, [it] immediately becomes an inept loser(5).

To say that they are cheating is not lightly said. ‘This stall’ has also made an affordability ratio. Myself of course took a suspicious attitude(6), I calculated the affordability ratio according to international practice, dividing the Rating and Valuation Department’s property price index by the median household income of private properties (not including public housing) in the Census and Statistics Department’s General Household Survey (both being public information), and then multiplying it by a factor (40000/3), thereby arriving at an affordability ratio similar to that of ‘this stall’. If you believe in them, [that means the] current property prices are merely equivalent to the Year 2000’s, and one would be stupid not to purchase; but if you believe in the official data, current property prices are in fact equal to the level in May 1997, just a few months before reaching the peak. You would groan stupid only if [you] purchase(7).

Such discrepancies cannot possibly be errors. No matter it was so done intentionally or unintentionally, these data have already had the effect of manipulating the market(8).

lawgang@gmail.com

Centaline and City University have jointly set up a property price index, when it was found that the data were dubious, the persons-in-charge of both sides passed the buck to each other(9). (online pictures)”

23.  There were two stings in this article complained of by the plaintiffs.  They concern the words “造市” (manipulation) and “出術” (cheating), which the plaintiffs say connote deliberately creating misleading impression of the property market.

The judge’s rulings of fair comment on the 2^nd Article

24.  The judge was of the view that the 2^nd Article goes further than the 1^st Article.  He held that the words “面對樓市出術” (when faced with cheating in the property market) as well as “說其出術” (to say that they are cheating) constitute statements of fact and not comment, as it is a direct accusation that “they” are cheating.  And to a reader, “they” can only mean the compilers of the CCLI.  It is not in dispute that the plaintiffs are the parties referred to in both articles.

25.  The judge took the view that in choosing the words “出術” (cheating), the author had “overstepped the line” and had “gone beyond what was permissible” by accusing the plaintiffs of cheating and manipulation of the property market.

26.  As there was no evidence to prove the plaintiffs to have knowingly engaged in misleading conduct in the property market, the judge held that neither the defence of fair comment nor the defence of justification could be established.

The approach of the appeal court

27.  On behalf of the plaintiffs, Mr Shieh, SC initially submitted whether the words complained of are allegations of fact or comment is what one would colloquially call a “jury question” (and thus a question of fact), so an appeal court would set aside a finding that the words amounted to fact not comment only if the finding is unreasonable or perverse, in the sense that it was one which a jury could not properly find.  In support, he cited Gatley on Libel and Slander (12^th ed) at §36.19.

28.  The passage in Gatley cited by Mr Shieh relates to an appeal to set aside the verdict of a jury, in which the appeal court is tasked to interpret the jury’s verdict.  In the absence of misdirection by the judge, the appeal court will only interfere or set aside a verdict if it was unreasonable or perverse in that there was no rational explanation for it.  We are here concerned with the speaking decision of a judge, not with a jury’s verdict or finding for which it is impermissible to speculate about the jury’s reasoning but one may only draw inevitable or proper inferences.  Furthermore, the question as to whether the words are fact or comment does not depend on the judge’s assessment of the reliability of witnesses or on the substance of their oral evidence and the material before us is exactly the same as was before the judge.

29.  The correct approach in this situation is as stated in another passage in Gatley at §36.25, cited by Mr Pow, SC for the defendants, which is based on the principle stated by Sir Thomas Bingham MR in Skuse v Granada Television Ltd [1996] EMLR 279 at 286:

“The Court of Appeal should be slow to differ from any conclusion of fact reached by a trial judge. Plainly this principle is less compelling where his conclusion is not based on his assessment of the reliability of witnesses or on the substance of their oral evidence and where the material before the appellate court is exactly the same as was before him. But even so we should not disturb his finding unless we are quite satisfied he was wrong.”

30.  Mr Shieh stressed the importance of the last sentence, and submitted we should not interfere with the judge’s finding if his finding is one that we may not agree with but is still within the realm of reasonable disagreement.

31.  I have no problem with this submission.  We should treat the judge’s decision on this question with all due respect, but if we are satisfied that his finding is wrong, then it is our duty to reverse him (Slim v Daily Telegraph [1968] 2 QB 157 at 186 to 187, per Salmon LJ; British Chiropractors Association v Singh [2011] 1 WLR 133 at §§13 to 15, 31, per Lord Judge CJ; Cruddas v Calvert [2013] EWCA Civ 748 at §§18 to 22, per Longmore LJ).  The determination of this issue does not depend solely on the document, but on the evaluation of the critical words in their context.  We should not depart from the judge’s finding unless it is clear some other finding applies (Cammish v Hughes [2013] EMLR 13 at §§29 to 31).

Fact or comment – the law

32.  As Lord Nicholls has stated in Cheng & Anr v Tse Wai Chun at 347A, the title of the defence of fair comment is misleading and it would be more satisfactory to call it “comment” or “honest comment”.  The name of the defence in the Defamation Act 2013 in England has been changed to “honest opinion”.  There is no change to the legislation in Hong Kong and we are still using the traditional terminology.

33.  For the defence to be available, the words must be recognisable as comment, as distinct from an imputation of fact.  Gatley at §12.7 regards the recognisability of a statement as comment as a “key determinant” of the availability of the defence. As stated by Mathews AJ in Harrigan v Jones [2001] NSWSC 623 at §72, “this threshold question, which sounds simple enough, often involves fine nuances of meaning and can be very difficult indeed”.  Lord Phillips of Worth Matravers PSC in Joseph v Spiller [2011] 1 AC 852 at §5 also acknowledged that “jurists have had difficulty in defining the difference between a statement of fact and a comment in the context of the defence of fair comment.”  Mason NPJ made a similar comment in Eastern Express Publisher Ltd & Anr v Mo Man Ching & Anr [1999] 3 HKLRD 530 at 541E to F.

34.  The ultimate determinant of whether the words are comment or fact is how they would strike the ordinary, reasonable reader (Gatley at §12.8).  “All the circumstances are to be taken into account, and it must be remembered that the ordinary reasonable reader does not parse and pore over the published matter as the lawyers are inclined to do at trial.”  (John Fairfax Publications Pty Ltd v O’Shane [2005] NSWCA 164 at §29, per Giles JA).  In determining this question, one takes into account “all the circumstances in which the publication was made, include the manner of speaking, the context, the tone of voice, the relationship between the material relied upon and the alleged comment, and the surrounding circumstances.” (Petritsis v Hellenic Herald Pty Ltd [1978] 2 NSWLR 174 at 182F, per Reynolds JA) The court should approach this in accordance with common sense and first impression (Slim v Daily Telegraph at 179C to D, per Diplock LJ; Waterson v Lloyd [2013] EMLR 17 at §61, per McCombe LJ).

35.  Though comment for this purpose is often equated with opinion, this is an over-simplification.  More accurately, it is “something which is or can reasonably be inferred to be a deduction, inference, conclusion, criticism, remark, observation, etc.” (Gatley at §12.8, citing the dictum of Cussen J at Clarke v Norton [1910] VLR 494 at 499, adopted in Branson v Bower [2001] EWCA Civ 791 at §12, per Latham LJ)

36.  A statement which is potentially one of fact may be one of evaluative opinion depending on the context.  “The context of a publication might alter the nature of its contents”, so “a statement which on its face appears to be an assertion of fact, might be found to be comment if it is the speaker’s conclusion drawn from facts which are contained within the publication.” (Harrigan v Jones at §79, per Mathews AJ) In Cheng & Anr v Tse Wai Chun at 347E to F, Lord Nicholls referred to this example given by Ferguson J in Myerson v Smith’s Weekly Publishing Co Ltd (1923) 24 SR (NSW) 20 at 26:

“To say that a man’s conduct was dishonourable is not comment, it is a statement of fact. To say that he did certain specific things and that his conduct was dishonourable is a statement of fact coupled with a comment.”

37.  In Joseph v Spiller at §5, Lord Phillips added these remarks to the above example:

“To say that a man’s conduct was dishonourable is not a simple statement of fact. It is a comment coupled with an allegation of unspecified conduct upon which the comment is based. A defamatory comment about a person will almost always be based, either expressly or inferentially, on conduct on the part of that person. Judges and commentators have, however, treated a comment that does not identify the conduct on which it is based as if it were a statement of fact. For such a comment the defence of fair comment does not run. The defendant must justify his comment. To do this he must prove the existence of facts which justify the comment.”

38.  Field J in O’Brien v Salisbury (1889) 54 JP 215 at 216 provided this further elucidation:

“Comment may sometimes consist in the statement of a fact, and may be held to be comment if the fact so stated appears to be a deduction or conclusion come to by the speaker from other facts stated or referred to by him, or in the common knowledge of the person speaking and those to whom the words are addressed, and from which his conclusion can be reasonably inferred … If, although stated as a fact, it is preceded or accompanied by such other facts, and it can be reasonably based upon them, the words may be reasonably regarded as comment, and comment only, and if honest and fair, excusable; and whether it is to be regarded as a fact or comment is a question for the jury, to be determined by them upon all the circumstances of the case.”

39.  Thus, where the statements in question are derived as inferences or conclusions suggested by the material set out in the articles, they are more likely to be construed as comment or statements of opinion, particularly when they are statements expressed in judgmental terms (Petritsis v Hellenic Herald Pty Ltd at 196E).  A similar approach to construction was adopted in British Chiropractors Association v Singh at §§24 to 26, 28 and 33; Harrigan v Jones at §§ 83 to 89; Mitchell v Sprott [2002] 1 NZLR 766 at §§27 to 28.

40.  Statements made in editorials or leaders are more likely to be regarded by readers as comment than matter in news stories, and a fortiori in a “Comment” section of a newspaper, although this is not conclusive (Gatley at §12.11, citing Keays v Guardian Newspapers Ltd [2003] EWHC 1565 (QB)).  See also British Chiropractors Association v Singh at §6; John Fairfax Publications Pty Ltd v O’Shane at §43; Peterson v Advertiser Newspapers Ltd (1995) 64 SASR 152 at 157.

41.  Depending on the context, attribution of motive can often be properly classified as comment.  As one can rarely establish what is going on in the mind of another and is therefore dependent on inference, where it is clear to a reader that the writer is drawing an inference as to another’s motivation, it would be reasonable to classify the imputation of motive as comment rather than fact (Pena v Tameside Hospital HHS Foundation Trust [2011] EWHC 3027 at §§27 to 28, per Eady J).

If the judge was in error in his approach

42.  Mr Pow submitted that the judge was in error in construing the words “面對樓市出術” (when faced with cheating in the property market) and “說其出術” (to say that they are cheating) as constituting statements of fact.  He had failed to construe these words in the context of the whole sentences (“港府監管金融有道，但面對樓市出術卻頓成廢柴” (The Hong Kong Government monitors the financial [market] well, but when faced with cheating in the property market, [it] immediately becomes an inept loser) and “說其出術不是亂說的” (To say that they are cheating is not lightly said)) and, even more importantly, in the context of the whole 2^nd Article.

43.  Although the judge did mention in §145 of the judgment that the words “出術” (cheating) appeared in the context of the sentences which he then quoted, I think there is clearly substance in Mr Pow’s criticism that the judge did not appear to have construed the defamatory imputation in the whole context of the 2^nd Article, which was a follow up to the 1^st Article published only the day before.  Further, in coming to his conclusion in §§147 to 149 that the words constituted statements of fact and amounted to a direct accusation the plaintiffs were cheating, his primary focus would appear to be on the natural and ordinary meaning of the words “出術” (cheating) (in §148), which he regarded as the writer having “overstepped the line” (in §149), rather than the context in which the defamatory words appeared so as to arrive at a proper conclusion whether the imputation was a statement of fact or in the nature of comment.  In stressing that the imputation was a direct accusation of the plaintiffs that they were cheating and manipulating the property market (in §§147 and 152), the judge appeared to have missed the focus of the inquiry.  An accusation could be a statement of fact.  Equally it could be a comment, if it is a conclusion that could clearly and reasonably be inferred from or based upon facts stated or referred to in the publication.

44.  It is also pertinent to note that when the judge set out the elements of the defence of fair comment in §115, he only mentioned four elements, which he proceeded to consider if each was satisfied in respect of the 1^st Article.  Because of his ruling that the defamatory words were fact not comment, he did not go through a similar exercise for the 2^nd Article. In Cheng & Anr v Tse Wai Chun at 347C to 348A, Lord Nicholls mentioned five elements which made up the objective limits of this defence.  The element the judge omitted to mention is that “the comment must explicitly or implicitly indicate, at least in general terms, what are the facts on which the comment is being made.”[1]  Again, the judge’s omission seems to me indicative of his failure to have proper regard to the context in which the defamatory imputation was made.

45.  I am satisfied the judge was in error in his approach.  As the judge had not approached the question of construction correctly, it falls upon this court to consider the question afresh.

Applying the law to the facts of the case

46.  Mr Shieh accepted that to say someone was cheating (出術) may or may not be a statement of opinion, depending on context, but it is certainly capable of being a statement of fact. He submitted we should approach the question of construction in this way.

47.  He asked the court to note that the words “出術” (cheating) were not preceded by judgmental phrases such as “疑似、令人覺得、我認為” (seemingly, would make people think, I consider). Whilst these words are not essential, he contended that their absence in the article is relevant.

48.  Next, Mr Shieh submitted that as a matter of context, the first reference to cheating was in “樓市出術” (cheating in the property market) and this appeared as a lead up to a criticism of the Government (“港府監管金融有道，但面對樓市出術卻頓成廢柴” (The Hong Kong Government monitors the financial [market] well, but when faced with cheating in the property market, [it] immediately becomes an inept loser)).  The latter part of the sentence criticising the Government as an inept loser is a statement of comment about the Government.  Counsel submitted it is clear that the preceding words “出術” (cheating) are relied on by the author as a factual averment in support of the comment about the Government.  The connotation of this sentence is that because some people like the plaintiffs “出術” (were cheating) as a matter of fact and the Government could not or did not do anything about this fact, the author proffered the comment that the Government was “廢柴” (an inept loser).

49.  Mr Shieh added that the defendants’ argument that statements as to the plaintiffs’ motivation are capable of being inferences is beside the point.  The sentence, properly construed, involves a direct allegation of deliberate wrongdoing against the plaintiffs.  He said it is not an attempt to analyse underlying facts and to proffer an opinion that the plaintiffs were deliberate in their publication of a misleading index.

50.  As for the second reference to “出術” (cheating) in “說其出術不是亂說的” (To say that they are cheating is not lightly said), Mr Shieh submitted this simply compounds the earlier allegation of cheating.  He argued that the second reference to cheating (and the words that followed) is not an attempt to support the earlier allegation of cheating by underlying facts.  The first reference to cheating refers to the CCLI.  The second reference to cheating concerns a different subject matter and a different index, namely, the affordability ratio (“負擔比率”).  By the second reference to cheating, properly construed, what the author was saying is that not only are the plaintiffs cheating in the CCLI (which fact leads to the earlier comment that the Government was an inept loser), but in the ensuing paragraph the author can give readers another incident of the plaintiffs cheating and that is about the plaintiffs’ affordability ratio.  By the same reasoning, if the first reference to cheating is a statement of fact, the second reference to cheating must likewise be a statement of fact.

51.  We should approach this question of construction in the way how it would strike an ordinary, reasonable reader.  For my part, I doubt very much if an ordinary, reasonable reader would dissect and analyse the words in the way Mr Shieh invited us to do.  I do not think it likely that this ordinary reader would only focus on particular words or phrases in the article.  An ordinary reader would not have latched on to words criticising the Government as comment and therefore regard the preceding words as fact.  Nor is it likely that it would strike the ordinary reader the second reference to cheating was in respect of a different subject matter.  Rather, the ordinary reader would form a general overall impression of what the article seeks to convey and would have viewed the defamatory imputation of cheating in that light.

52.  I agree with Mr Pow it is important to have regard to the fact that both articles appeared in a commentary section of the newspaper, where the ordinary reader would expect to find comments, discussion and analysis on financial matters rather than facts conveyed in news stories.  The judge noted in §120 of the judgment that the 1^st Article was in the commentary section and stated that readers are likely to understand the words complained of as comments made by the author.  He mentioned in §138 that everything he said about the 1^st Article would be applicable to the 2^nd Article.  That being the case, one should attach appropriate, although not conclusive, weight to this matter.

53.  It is also important to bear in mind that the 2^nd Article was a follow up to the earlier article. At the beginning of the 2^nd Article, the author expressly reminded the readers of the gist of what he had criticised in the article only the day before.  It was just further criticism on the same theme, that the CCLI was unreliable.  The judge had found the defamatory imputations in the earlier article are all in the nature of comments, not statements of fact.  If the publication had not been split up into two articles, and they were published in one go, there is much to be said for the view that taking into account the overall context and the relationship between the defamatory imputations and the materials relied on, the defamatory imputation in the 2^nd Article should be considered along the same lines as the defamatory words in the 1^st Article.  As it is, I think it is appropriate to consider that readers who had read the earlier article would have viewed the 2^nd Article as containing further criticisms by the author in the same vein.

54.  It would be apparent to the ordinary reader from the overall contents and structure of the 2^nd Article that it is in the nature of a commentary, with the author putting across his views and seeking to persuade the readers that his views are well founded.  Hence, the author brought up the affordability ratio of the plaintiffs, and set about constructing his own ratio by internationally accepted methodology, using data available from the public domain.  For the readers, he plotted a graph presenting the plaintiffs’ ratio and the ratio that he constructed.  He made the observation there were big discrepancies between the two ratios as of December 2010 and postulated that potential home buyers might take diametrically opposite decisions, depending on which ratio they chose to adopt.  He made the remark that the discrepancies of the ratios could not be due to mathematical errors.  As noted by the judge in §142, for present purpose it is not necessary to find which of the two affordability ratios is more accurate or reliable.

55.  An ordinary reader reading the article as a whole would not have lost sight of the fact that the author had put forward calculations and analysis to back up his views.  It is not material that the defamatory imputation in this article is stronger and goes further than the imputations in the earlier article, as the judge had thought.  What matters for this exercise is whether the imputation, albeit couched in pungent and pejorative terms, should be regarded as a statement of opinion in all the circumstances.  I agree with Mr Pow the readers would most likely understand in the references to “出術” (cheating) in the 2^nd Article, that the author was putting forward his deduction, inference and remark, which he sought to support by the statistical calculations, analysis and observations he mentioned in the article.

56.  For the above reasons, I would hold that the defamatory imputation in the 2^nd Article is in the nature of comment for the purpose of the defence of fair comment.  I am satisfied the judge’s ruling that the imputation is a statement of fact is wrong and this cannot be regarded as being within the realm of reasonable disagreement.

57.  I wish to make one last observation.  Mr Pow urged upon us not to take a narrow approach in construing whether the defamatory imputation is fact or comment, as that would impose an unjustified restriction on the freedom of expression and “the right of fair comment is a most important element in the freedom of speech” (Cheng & Anr v Tse Wai Chun at 345D, per Li CJ).  Mr Shieh submitted to the contrary that the court should not lean in favour of holding that the words are comment but should simply apply the well established tests to the facts of the case.  In holding in favour of the defendants on this question of construction, I have simply applied the guidance and principles established in the cases cited.  I have not found it necessary to widen the scope of fair comment to accommodate the construction that I arrived at.

Other elements of the defence of fair comment

58.  Mr Shieh did not make any substantive submissions that the other elements of fair comment are not satisfied in respect of the 2^nd Article. It would suffice to deal with them succinctly.

59.  It is not in dispute that the comment was made on a matter of public interest.  The facts on which the comment was made are substantially true.  They are the facts which formed the factual background leading to the publication of the articles, the affordability ratio constructed by the author, and the big discrepancies between the plaintiffs’ ratio and the author’s ratio.  The comment is not a bare comment.  There is express indication in the publication of the factual basis on which the comment was made.  The judge did not reject the author’s evidence of his subjective belief (quoted in the judgment at §146) that the plaintiffs had knowingly engaged in misleading conduct.  A comment may be honest however exaggerated or even prejudiced in the language of the criticism; it may be honest even if it is irrational, stupid or obstinate, or expressed in pungent and offensive tones (Gatley, §12.27).  Had the judge asked the proper question, he would have held that the comment is one which could have been held by an honest person, and was honestly held by the author.

Conclusion

60.  I would uphold the defence of fair comment for the 2^nd Article and allow this appeal.  I would set aside the judgment and award of damages in favour of the plaintiffs and the costs order of the judge and make an order that the plaintiffs should pay the defendants’ costs in this appeal and below, with a certificate for two counsel in both instances.

Hon Chu JA:

61.  I agree with the reasons given by Kwan JA and the orders proposed.

Hon Poon JA:

62.  I agree with the judgment of Kwan JA and the orders that she has proposed.

[1] Lord Nicholls went on to state, obiter, at 347H and 352E, the factual basis of the comment must be sufficient to enable the reader to judge for himself how far the comment was well founded. The Supreme Court in Joseph v Spiller disapproved of this part of the element stated by Lord Nicholls, holding that it would be sufficient so long as the reader could understand from the factual basis what the comment is about and the commentator can, if challenged, explain by giving particulars of the subject matter of his comment why he expressed the views that he did. For the purpose of this appeal, it is not necessary to go into the question what should be the degree of particularity required for the factual basis of the comment.