BAWANG INTERNATIONAL (GROUP) HOLDING LTD AND ANOTHER v. NEXT MAGAZINE PUBLISHING LTD

HCA 1109/2010

IN THE HIGH COURT OF THE

HONG KONG SPECIAL ADMINISTRATIVE REGION

COURT OF FIRST INSTANCE

ACTION NO 1109 OF 2010

____________________

____________________

__________________

JUDGMENT

__________________

Index

I Background

II The defamatory meanings of the Article with reference to the Plaintiffs

(i)  Legal principles in determining the meaning of the alleged defamatory words
(ii)  The Plaintiffs’ pleaded meanings
(iii)  Words disparaging to products vis-à-vis the Plaintiffs
(iv)  Reference to the Plaintiffs

III Defence of justification

(i)  Lucas-Box meanings pleaded by the Defendant
(ii)  Toxicological evidence on the alleged harmful effect of 1,4-dioxane in BaWang Shampoo Products
(a)  What is 1,4-dioxane?
(b)  “Human studies” in respect of 1,4-dioxane
(c)  Regulations and safety standards in relation to the presence of 1,4-dioxane in cosmetic products including shampoos
(d)  Different to methodology  to assess the health risk associated with the presence of 1,4-dioxane in consumer products
(e)  Is 1,4-dioxane genotoxic?
(f)  MOA (Mode of Action)
(g)  The policy behind the default option and its limitations
(h)  Are the assessments made by NICNAS, EU Bureau and Health Canada out-dated?
(i)  WHO’s and HKWSD’s safety level of 1,4-dioxane in drinking water and the ‘de minimis’ level advocated by Dr Sawyer
(iii)  Conclusion on toxicology evidence and the defence of justification

IV Defence of honest comment

(i)  Undisputed legal principles
(ii)  The “three comments” are statements of facts
(iii)  “Comments” not based on true facts
(iv)  The first two “comments” do not meet the defamatory stings in the Article
(v)  Malice in publishing the “comments”

V Defence of publication in the public interest

(i)  The development of the principle of “responsible journalism”
(ii)  Human rights jurisprudence
(iii)  The standard of responsible journalism
(iv)  Events leading to the publication of the Article
(a)  How did the investigation begin?
(b)  How was the investigation conducted?
(c)  How was the Article drafted and edited?
(d)  How did the Defendant seek the reply or verification from the BaWang Group?
(e)  The dubious background of “Mr Chan”
(v)  Whether the Defendant’s reporting fell below the standard of responsible journalism?
(a)  Serious nature of the allegations targeting the Plaintiffs
(b)  Serious flaws in the conclusion reached by the Defendant as to the safety level of 1,4-dioxane in shampoo and its failure to verify the scientific basis for the advocated safety limit
(c)  Failure to present fairly or adequately the contrary scientific view and the reply of the Plaintiff
(d)  Inclusion of unnecessary defamatory accusations against the Plaintiffs
(e)  Failure to verify the credibility of the information provided by the complainant
(f)  Other considerations
(g)  Conclusion on responsible journalism

VI Injurious falsehood

(i)  Falsity of the statement
(ii)  Malice in publishing the statement

VII Quantum

(i)  The actual loss of sales and profits
(ii)  Juridical basis in claiming for loss resulting from republications
(iii)  Causation and remoteness of damages
(iv)  Conflict of law issue and the question of double actionability
(v)  Right to claim for loss of the sales and profits in Hong Kong
(vi)  The effect of the said rulings on the assessment of damages
(vii)  General damages
(viii)  Exemplary damages

VIII Conclusion

1. This a claim for libel and malicious falsehood in respect of an article published in the Next Magazine on 14 July 2010 (“the Article”) about the presence of a chemical known as 1,4-dioxane in the Plaintiffs’ shampoos.

2. The quantum of the Plaintiffs’ claim is substantial amounting to about HK$600 million, which may be a record in term of the quantum of a defamation claim.

3. The trial of this case lasted for 39 days.  The court has been asked to determine numerous issues, including the concept of responsible journalism, conflict of law issues relating to the claim for damages caused by overseas third parties’ republications, human right consideration in assessing damages resulting from a piece of investigative journalism and the relevance of the no-reflective-loss principle in assessing the damages for defamation suffered by the holding company and its subsidiaries. Some of these issues are novel which have not been canvassed in the Hong Kong courts before.  Apart from these legal issues, the court has to deal with substantial and complicated expert evidence from toxicologists, chemical engineering experts, accounting experts and Mainland law experts.  Due to the complexity of this case, I find it difficult to keep this Judgment to a reasonable length.

4. I cannot criticise the parties for overburdening the court with all these issues.  After all, the subject Article had seriously affected the business and the reputations of the Plaintiffs and the share price of the 1^st Plaintiff had dropped significantly within a few hours after the publication of the Article.  On the other hand, this is the first time that a journalist is facing such a huge claim for defamation relating to a piece of investigative journalism.  The present case involves human right considerations such as freedom of expression and right of the public to know. All these issues deserve serious consideration.

I: BACKGROUND

5. The 1^st Plaintiff is a company incorporated in the Cayman Islands whose shares are listed in the Hong Kong Stock Exchange.

6. The 2^nd Plaintiff is a company incorporated in the Mainland with its registered office in Guangzhou.  The 2^nd Plaintiff is a wholly-owned subsidiary of the BaWang Group of companies (“the BaWang Group”) of which the 1^st Plaintiff is the ultimate holding company.

7. The Plaintiffs are engaged in the business of designing, manufacturing and marketing of Chinese herbal shampoos and hair care products, skin care products and health care products, including shampoo products under the brand names “BaWang” and “Royal Wind” (“BaWang Shampoo Products”).  The Plaintiffs’ products are sold in the Mainland, Hong Kong and some other countries around the world. 

8. The Defendant is the publisher of a weekly Chinese magazine bearing the name “Next Magazine” (“the Magazine”) which is widely circulated in Hong Kong.

9. The Article involved a report about the presence of an alleged carcinogenic substance known as 1,4-dioxane in BaWang Shampoo Products.

10. The handling reporter of the Article was one Madam Lam Yu Ting (林瑜婷)(“Madam Lam”).  According to Madam Lam, the Article originated from a complaint made by one “Mr Chan” (“Mr Chan”).  Because of his family history, “Mr Chan” was concerned about the presence of 1,4-dioxine in shampoo.  He therefore bought a few shampoos in the market and sent them to surveyors for examination. The reports from the surveyors showed that BaWang shampoo contained 27 ppm (parts per million) of 1,4-dioxane, whereas shampoos of two other brands, Rejoice and Acene, contained 8.8 ppm and 2.9 ppm of 1,4-dioxane respectively.  For the other 3 brands including O’Naomi, the results were “not detected”.  After interviewing the complainant, Madam Lam purchased 3 bottles of BaWang Shampoo Products which were later tested to have contained 10 ppm of 1,4-dioxane.  She also conducted some research about such chemical which was a confirmed animal carcinogen. Some regulatory agencies classified the chemical as a “possible human carcinogen”.

11. In the course of her research, Madam Lam had also spoken, a few times over the telephone, with Professor Lam Hon Wah Michael (林漢華) (“Professor Lam”) who is an associate professor in the Department of Biology and Chemistry of the City University of Hong Kong, and Dr Lau Fei Lung (劉飛龍)(“Dr Lau”) who is the director of the Toxicology Training Centre in United Christian Hospital.

12. Madam Lam produced the first working draft of the Article on about 12 July 2010.  On 13 July 2010, the Magazine sent a letter to the public relation company of the BaWang Group for comment about the presence of 1,4-dioxane in BaWang Shampoo Products.  On the same day, two reporters of the Magazine went to the head office of the BaWang Group in Guangzhou asking for comment or explanation about the issue. 

13. The Article was published in the 1062^nd issue of the Magazine dated 15 July 2010.  It is common ground that that particular edition of the Magazine was available in the market in the morning of 14 July 2010.

14. The heading or the topic of the Article was “霸王致癌” (BaWang causes cancer), which appeared in the cover page of the Magazine and the first page of the Article.  There was also a short passage near the topic in the first page of the Article which reads as follows:

“因成龍一句「動L」而揚名的霸王洗頭水，打着「中藥世家」旗號，標榜防脫髮配方，就算售價貴過不少歐、日名牌洗頭水兩三成，照樣大把人捧場。去年霸王更乘勢來港上市，現時市值逾一百七十億元，創辦人陳啟源夫婦的身家更暴漲至逾百億元。本刊最近發現霸王旗下的多隻洗頭水，不約而同含有名為二噁烷的致癌物質；醫生表示長期接觸可導致皮膚敏感出疹，損害肝腎功能，甚至引發癌症！”

The English translation reads:[1]

“BaWang shampoo, which became famous because of the slogan ‘Dong-al’ by Jackie Chan, flaunted the banners of ‘renowned family of traditional Chinese herbs” and “anti-hair fall formula’.  Even though its price is 20%-30% higher than a number of famous European or Japanese brand shampoos, it is still very popular.  Last year, BaWang further advanced its position and got listed in Hong Kong.  Its current market value is in excess of $17 billion and the wealth of its founders, the Chen Qiyuan couple, was substantially increased to over $10 billion.  This magazine recently found that several kinds of shampoo under BaWang all contained, incidentally, a carcinogen called 1,4-dioxane.  Doctors indicated that persistent contact with the substance could cause skin allergy and eruption, damage to liver and kidney functions or even cancer.”

15. The contents of the Article read as follows:

“近年，霸王洗頭水的風頭一時無兩，儼如中年男士的防脫救星。年約四十歲的陳先生，去年底全家開始使用霸王洗頭水，「我同屋企人都開始有脫髮問題，而我哋一家由細到大都信中藥，霸王標榜中藥防脫，所以轉用霸王。

今年三月，陳先生在報章看到一則關於洗頭水含致癌物質的新聞：「有內地媒體化驗了十支內地品牌的洗頭水，居然有九支含有致癌物質二噁烷！」惟該篇報導沒有披露品牌名稱，令全家都用霸王的陳先生憂心忡忡，「我阿嫲阿爺同姑姐都係肝癌死，我自己有乙型肝炎，人哋話癌症隔代遺傳的機會率好高。自細阿媽已經好小心，食嘢要揀有機，會致癌的食物，好似鹹魚更加唔會食。」

朋友見陳先生愁眉不展，遂建議他把洗頭水拿去化驗，「問咗幾間先搵到香港通用公證行有得驗二噁烷，其他例如標準化驗所都無得驗。」他遂在百佳買了一支全新的霸王洗頭水，交予香港通用公證行化驗，並付了一千五百元化驗費。

二噁烷可穿透皮膚吸收

兩星期後，化驗報告出爐，陳先生一看之下頓時又驚又怒，「霸王咁大間公司，洗頭水居然都有二噁烷！」他遂即時棄用家中的霸王洗頭水，但他始終想繼續用中藥洗頭水，無計可施下，惟有接連將飄柔、澳雪、O’Naomi，屈臣氏及萬寧牌出產的中藥洗髮水拿去化驗，總共花費了七千多元。

結果，飄柔及澳雪均含微量二噁烷，故他現時選用沒有被驗出該物質的O’Naomi洗頭水，「既然人哋都可以無，咁點解霸王咁大牌子會有呢？」

根據陳先生提供的化驗報告，霸王的二噁烷含量為27ppm（即一百萬分之二十七，或每一公升洗頭水含二十七毫升）；飄柔首烏洗頭水則含有8.8ppm；澳雪人參靈芝洗頭水含有2.9ppm；O’Naomoi、萬寧牌及屈臣氏牌中藥洗髮水的化驗結果是「Not Detected」，即未能驗出含有二噁烷。本刊隨後在屈臣氏買了兩支霸王及一款同為霸王生產的追風洗頭水，未經開封便送到香港通用公證所化驗，三支洗頭水的結果皆含有10ppm的二噁烷。

城市大學生物及化學系副教授林漢華表示，二噁烷是有害物質，「如果一次過好大劑量咁喺空氣中吸入，甚至會死。」他續指二噁烷可穿透皮膚，再被人體吸收，「佢可以經皮膚入血，好似透過洗頭水或化妝品，長期接觸二噁烷，有可能會令肝同腎受損，而喺動物身上，亦已證實會致癌。」聯合醫院中毒諮詢科總監劉飛龍補充：「長期接觸，皮膚可能會出疹，亦會影響免疫系統。可以避免的話，就應該盡量避免。」

原料質素影響含量

林漢華表示二噁烷本身沒護髮或清潔功效，惟為令到洗頭水的質感比較柔滑及油潤，往往會加入聚乙二醇等甘油類化學品，「而制造聚乙二醇的過程之中，可能用到二噁烷。洗頭水若含有二噁烷，可能是因為呢啲聚乙二醇原料上有殘留物質。」

根據美國環境保護局資料，二噁烷屬於「可能引致人類癌症」一類；按加州用作監管有害化學品的條例《第六十五號提案》，二噁烷更直接被歸類為致癌物質，若產品含量超過10ppm，消費者有權控告生產商。

其實多個國家如歐盟、中國及台灣均禁止生產商在個人護理產品及化妝品添加二噁烷，惟若是在製造過程中令產品受到微量污染，則可以接受，但各地均沒界定何為微量。

霸王便以此作為擋箭牌，堅拒認錯，表示產品接照國內現行法律、法規及標準之要求規範生產。記者帶着化驗報告直闖廣州的霸王總部，向負責人質詢為何洗頭水中含有致癌物質，惟對方多個職員均推說負責人不在。

承認早知含二噁烷

最後本刊以電話聯絡到霸王首席執行官萬玉華，她表示早已知道產品含有二噁烷：「是在原料上出現的，無可避免，全行大部分洗頭水都有, 但含量好少，對人體唔會有害。」惟專家不認同，林漢華說：「洗頭水係日日用，而且二噁烷都幾毒，10ppm已經係危險邊緣，20ppm以上直頭係多。」

而且，根據美國食物及衛生局資料，即使二噁烷真的在生產過程中殘留，亦可使用真空剝離(Vacuum Stripping) 辦法抽出。林漢華指出：「原料上唔一定會有殘留，簡單啲講，原料靚自然無雜質同污染。」劉飛龍續指要產品完全不含二噁烷，生產成本可能貴五至六倍，故廠商未必願意。換句話說，覇王之所以含二噁烷，原因可能是原料質素欠佳。

飄柔及澳雪的公關回覆本刊，其產品所含的二噁烷微乎其微，不會危害人體健康，但亦表示會設法減少含量。”

The English translation reads:[2]

In recent years, BaWang shampoo stood out from others as if it was the emancipator for prevention of hair falling in middle-aged men.  “I and my family began to have hair-falling problems.  We all believe in Chinese herbs since we were small.  BaWang flaunted its banners of ‘Chinese herbs’ and ‘anti-hair falling formula’, so[we] changed to BaWang”, said Mr. Chan, who is about 40 years of age and whose family started to use BaWang shampoo since late last year.

In March this year, Mr. Chan read a newspaper article about shampoo containing carcinogens.  “A mainland media arranged to examine 10 Chinese-branded shampoo, surprisingly 9 of them contain the carcinogen 1,4dioxane!”  However, the article did not disclose the brand names, thus causing Mr. Chan, whose family all used BaWang, to be very anxious.  “My grandparents and my aunt all died of liver cancer and I personally have hepatitis B.  People said cancer has high rate of reversion.  My mom has been very careful since [we were] small and would choose organic food [for us].  [We] would not eat food that would cause cancer, like salted fish.”

A friend who found Mr. Chan anxious suggested him to take the shampoo for examination.  “I made inquiries with several [laboratories] and SGS Hong Kong is the only one that provides examination for 1,4-dioxane content.  Such kind of examination is not available in other [laboratories], like [Standard Laboratory].”  Mr. Chan thus bought a new bottle of BaWang shampoo in Park n Shop, handed it to SGS Hong Kong for examination and paid the examination fee of $1,500.

1,4-dioxane can penetrate through the skin and be absorbed

Two weeks later, the examination report was ready.  Mr. Chan felt scared and angry upon reading it, “BaWang is such a big company and yet its shampoo contains 1,4-dioxane!”  He thus immediately stopped using the BaWang shampoo at home, but he still wanted to use Chinese herbal shampoo.  Running out of ideas, he took all brands of Chinese herbal shampoo including Rejoice, Acene, O’Naomi, Watson’s and Mannings to be examined, and spent a total of HK$7,000 odd [for the examinations].

The results, [of the examinations] showed that Rejoice and Acene both contained traces of 1,4-dioxane.  So he now uses O’Naomi shampoo, in which no 1,4-dioxane was detected.  Other [products] do not have [1,4-dioxane], why was it found in a big brand like BaWang?”

According to the examination report provided by Mr. Chan, the 1,4-dioxane content was 27ppm (i.e. 27/1000000 or 27ml per litre of shampoo) in BaWang, 8.8ppm in Rejoice polygonum shampoo, 2.9ppm in Acene ginseng and lingzhi shampoo, and the results for O’Naomi, Mannings and Watson’s brands of shampoo were “not detected”, that is, no 1,4-dioxane content was detected.  This magazine then bought two bottles of BaWang shampoo and a bottle of Royal Wind shampoo (also produced by BaWang) from Watson’s and delivered the bottles unopened to SGS Hong Kong for examination.  The result shows that all three bottles of shampoo contained 10ppm of 1,4-dioxane.

Lam Hon Wah, Associate Professor of the School of Biology and Chemistry of City University [of Hong Kong] indicated that 1,4-dioxane was noxious and “if one inhales a large volume of it in the air in one go, one will even die”.  He further indicated that 1,4-dioxane could penetrate through the skin and then be absorbed by the human body.  “It can go through the skin into the blood.  If one has persistent contact with 1,4-dioxane, say, by  using shampoo or cosmetics, it may cause damage to the liver and kidney, and it has been proved in animals that it can cause cancel.”  Lau Fei Lung, Director of the Toxicology Training Centre of the United Christian Hospital, added, “If one has persistent contact [with it], one’s skin may have eruption, and the immune system will also be affected.  It shall be avoided as far as possible.”

Quality of raw materials affected the content

Lam Hon Wah indicated that 1,4-dioxane itself does not have conditioning or cleansing functions.  However, to make shampoo softer and smoother, very often glycerol kind of chemicals such as polyethylene glycol would be added.  “1,4-dioxane may be used in the process of manufacturing polyethylene glycol.  If shampoo contains 1,4-dioxane, it may be the residue in the polyethylene glycol materials.”

According to information from the Environmental Protection Bureau of the United States, 1,4-dioxane belongs to the kind [of materials] that “may cause cancer to humans”.  According to “Motion No.65” which is the law of California governing toxic chemicals, 1,4-dioxane is directly classified as a carcinogen.  If its content exceeds 10ppm, consumers have the right to sue the manufacturer.

In fact, a number of countries such as the EU, China and Taiwan all prohibited manufacturers to add 1,4-dioxane into any personal skin care products or cosmetics, but it is acceptable if the product is contaminated with traces of [1,4-dioxane] in the course of manufacturing.  However, there is no definition for “traces”.

BaWang thus used this as a shield and firmly refused to admit any fault.  They indicated that their products were produced according to the current laws, regulations and standards of China.  Our reporter went straight to the headquarters of BaWang with the examination reports [intending to] question the person in charge why a carcinogen was found in their shampoo, but several staff members gave an excuse that the responsible person was not around.

Admitted that [they] knew long time ago about the 1,4-doxane content

Lastly, this magazine got in touch with Wan Yuhua, chief executive officer of BaWang, by telephone.  She indicated that they knew long time ago about the 1,4-dioxane content, “It exists in the raw materials used and is unavoidable.  It is there in most of the shampoos produced by our fellow traders but its content is minimal and will not do any harm to the human body.”  Experts however disagreed.  “Shampoo is used every day and 1,4-dioxane is quite toxic.  10ppm is already on the margin of being dangerous and 20ppm is absolutely excessive.” said Lam Hon Wah.

In addition, according to information from the Food and Hygiene Bureau of the United States, even if there is really residue of 1,4-dioxane in the production process, it can be extracted by means of vacuum stripping.  “There is not necessarily residue in raw materials.  Put simply, good raw materials will certainly have no impurity or contamination,” indicated Lam Hon Wah.  Lau Fei Lung added that the production cost may rise by 500 to 600% if [manufacturers] are to make their products completely free of 1,4-dioxane and producers may therefore not be willing to do so.  In other words, the reason for the 1,4-dioxane content found in BaWang products may be the poor quality of raw materials.

The Public Relations Department of Rejoice and Acene replied to this magazine that their products contained very minimal level of 1,4-dioxane which will not cause any harm to human health, but they indicated that they would try to reduce such content.”

16. A caption also appeared next to a whole-page photograph of, amongst others, Mr Chan Qiyuan (the founder of the 1^st and 2^nd Plaintiffs) and his wife, containing the following words:

“霸王洗頭水以中藥及天然作為賣點,但內裡卻含有致癌物質,長期使用會危害健康.”

The English translation reads:[3]

“BaWang shampoo uses Chinese medicine and natural [ingredients] as its selling points but contains carcinogen; prolonged use will be hazardous to health.”

17. The Plaintiffs complain that the words in the Article mentioned in the preceding three paragraphs (“the Words”) are defamatory of them.

18. The Article then went on with the listing history of the 1^st Plaintiff and the family background of Mr Chan Qiyuan, and it was queried in the Article whether Mr Chan Qiyuan did come from a family with Chinese medicine background.  It was also alleged that Mr Chan Qiyuan had made a fortune by reason of the soaring share price of the 1^st Plaintiff.

19. The Article also contained 2 captioned boxes: one containing some technical facts about 1,4-dioxine, and the other one listing out some incidents about the concern for the presence of 1,4-dioxine in consumer products in recent years.

20. The share price of the 1^st Plaintiff dropped by about 14% within a few hours after the publication of the Article. The trading of the 1^st Plaintiff’s shares was suspended in the Hong Kong Stock Exchange in the afternoon of the same day.

21. The Plaintiffs’ main claim is one on defamation. To oppose such claim, the Defendant relies on the following defences:

(i)  the facts and allegations included in the Article are true;

(ii)  certain comments made in the Article are fair comment; and

(iii)  it would be in the public interest to publish the Article, thereby the Defendant is protected by the defence of publication in the public interest or commonly referred to as the “Reynolds privilege”.

22. I will deal with these defences in the latter of this Judgment.  Before I do so, I have to make a finding about the meanings of the Words and whether the Words referred to the Plaintiffs.  Whether the defences of justification, fair comment and publication in the public interest would succeed depends on the meanings of the Words.

II: THE DEFAMATORY MEANINGS OF THE ARTICLE WITH REFEFENCE TO THE PLAINTIFFS

23. In order to succeed in a claim for defamation, a claimant has to show that defamatory matters with reference to the claimant have been published by a defendant.

24. The matters published are defamatory if they expose the claimant to contempt and ridicule by others and cause others to shun him.  The matters would also be defamatory if they lower the claimant’s estimation in the eyes of right-thinking members of society and adversely affect his reputation.[4]

(i)  Legal principles in determining the meaning of the alleged defamatory words

25. There is no dispute that the following principles are applicable for the court in determining the meaning of an alleged defamatory article.

26. In determining the meaning, words in an article are normally construed according to their natural and ordinary meaning, and the judge or jury has to ascertain the single meaning which the publication sued upon would convey to the mind of an ordinary, reasonable, fair-minded reader.[5]

27. The article must be read as a whole.  It is well established that the context and circumstances of the publication must be taken into account.  A plaintiff cannot pick and choose parts of the publication.  For example, a particular sentence may be considered defamatory, but there may be other passages which neutralize or mitigate its sting.  A reasonable reader is assumed to have read the whole article complained of, and a defendant is entitled to have considered as part of the plaintiff’s case the whole of the piece from which the alleged libel is extracted.[6]

28. In reading the article in its entire context, the court will also bear in mind the general tenor of the entire article.  The fact that an article gives a generally uncomplimentary rendering of the claimant will obviously affect a reasonable reader’s understanding of the specific portions or statements in the article.  The court will also consider the effect of the “visual impact” of the article on the readers.

29. As regards the meaning of the alleged defamatory words, it is trite that it is a question of fact to be read “in the sense in which ordinary persons, or in which we ourselves out of court… would understand them” and whether the words are capable of defamatory meaning is for the judge.[7]

30. In Jeynes v News Magazines Ltd[8], Sir Anthony Clarke MR summarised the nature of the exercise:

“(1) The governing principle is reasonableness. (2) The hypothetical reasonable reader is not naïve but he is not unduly suspicious. He can read between the lines. He can read in an implication more readily than a lawyer and may indulge in a certain amount of loose thinking but he must be treated as being a man who is not avid for scandal and someone who does not, and should not, select one bad meaning where other non-defamatory meanings are available. (3) Over-elaborate analysis is best avoided. (4) The intention of the publisher is irrelevant. (5) The article must be read as a whole, and any ‘bane and antidote’ taken together. (6) The hypothetical reader is taken to be representative of those who would read the publication in question. (7) In delimiting the range of permissible defamatory meanings, the court should rule out any meaning which, ‘can only emerge as the produce of some strained, or forced, or utterly unreasonable interpretation…’ (8) It follows that it is not enough to say that by some person or another the words might be understood in a defamatory sense.”

31. It is well established that the words are to be construed in their natural and ordinary meaning, i.e. the meaning in which reasonable people of ordinary intelligence, with the ordinary person’s general knowledge and experience of worldly affairs, would be likely to understand them.[9]

32. The natural and ordinary meaning does not rest upon any technical process of construction or analysis.  In some circumstances, the defamation does not lie in the actual words used, but in an inferential or implied meaning that the words are capable of bearing without the need to have recourse to extrinsic facts.[10]

33. It has been considered that the range of legitimate meanings is a matter of impression, “provided that the impression is not of what the words mean but of what a jury could sensibly think they meant… Such an exercise is an exercise in generosity, not in parsimony”[11].

34. In defamation cases, the “single meaning rule” applies. The words complained of were deemed to bear only one natural and ordinary meaning which was the meaning that the words would convey to the reasonable reader.  In Oriental Daily Publisher Ltd & Anor v Ming Po Holdings Ltd[12], Lord Neuberger NPJ observed that, in ascertaining a single meaning of the words, the first question to be considered must be: what did the statement mean?  His Lordship also said: “Conventionally, the question admits of only one answer, to be arrived at by reference to the words used when assessed in their context, documentary, factual and common sense. This is true even where more than one meaning is possible.”[13]

(ii)  The Plaintiffs’ pleaded meanings

35. In the Re-Re-Amended Statement of Claim, it is pleaded that the Words in the Article, in their natural and ordinary meaning, meant and were understood to mean:[14]

(i)  the 1^st and/or 2^nd Plaintiffs knowingly manufactured and marketed products which would, through normal and regular use, cause serious illnesses and even cancer to users of these products;

(ii)  the 1^st and/or 2^nd Plaintiffs knowingly used cheap and inferior raw materials in the manufacture of its products in order to save costs even though the end product would, even through normal and regular use, cause serious illnesses and even cancer to its users;

(iii)  the 1^st and/or 2^nd Plaintiffs ignored or paid little regard to the health and safety of consumers who used or purchased its products;

(iv)  the 1^st and/or 2^nd Plaintiffs exploited legal and regulatory loopholes and relied on the absence of legal sanction or regulation as an excuse to justify its conduct in manufacturing and marketing products which would, through normal and regular use, cause serious illnesses and even cancer to users of those products;

(v)  the 1^st and/or 2^nd Plaintiffs had unjustifiably refused to acknowledge that it had done anything wrong; and

(vi)  by reason of the aforesaid conduct, the 1^st and 2^nd Plaintiffs were irresponsible, incompetent and greedy commercial organisations which lacked commercial morality.

36. Mr Pow SC, counsel for the Plaintiffs, submits that although there are 6 pleaded meanings, all of them are interrelated.  According to him, the 1^st meaning is the core meaning and all the other meanings actually “flow” from the 1^st meaning.

37. The 1^st pleaded meaning is that the Plaintiffs knowingly manufactured and marketed products which would, through normal and regular use, cause serious illnesses and even cancer to users of these products.  The Plaintiffs contend that such meaning contains the following “stings” on both the Plaintiffs’ products and their conduct of business:

(i)  BaWang Shampoo Products would cause serious illness or cancer through normal and regular use;

(ii)  this fact was known to the Plaintiffs; and

(iii)  nonetheless, they went ahead to manufacture and market  such products.

38. Mr Yu SC, counsel for the Defendant, submits that the Article does not impute the first sting as mentioned above.  According to him, the Article only carries the meaning that BaWang Shampoo Products contain carcinogenic substance known as 1,4-dioxane and long term use of such products may have the possibility of getting cancer. Further, it was written in the Article that in a number of places such as European Union (“EU”), Mainland and Taiwan, it was acceptable for consumer products to contain traces of 1,4-dioxance and there was no definition for “traces”, and so the Article does not suggest that normal or regular use of BaWang Shampoo Products would cause cancer or other illnesses.

39. I disagree.  The meaning of the Article goes far beyond that.  The topic of the Article as shown in the cover page of the Magazine and the first page of the Article clearly states that BaWang Shampoo Products cause cancer.  The artwork for the words “致癌” (cause cancer) in the first page of the Article further heightened the readers’ alarm.  The Article vividly described the awe and fear harboured by “Mr Chan” as to the prospect of developing cancer through using BaWang Shampoo Products.  In the Article itself, it has been repeated many times that prolonged and regular use of BaWang Shampoo Products would be hazardous to human health, causing cancer and damage to liver and kidney and other illnesses, including the statement: “霸王洗頭水以中藥及天然作為賣點,但內裡卻含有致癌物質,長期使用會危害健康” (BaWang shampoo uses Chinese medicine and natural [ingredients] as its selling points but contains carcinogen; prolonged use will be hazardous to health).  Further, the Article tells the readers that the presence of 10 ppm of 1,4-dioxane in shampoo is already on the margin of being dangerous and 20 ppm is absolutely excessive.  As the Article claims that BaWang Shampoo Products contain as much as 27 ppm of 1,4-dioxance, the clear message conveyed by the Article is that normal and regular use of BaWang Shampoo Products would cause serious illness and cancer.  I do not believe that Mr Yu is seriously suggesting otherwise.

40. The main dispute between the parties is whether the Article is also imputing that the Plaintiffs knew about the alleged harm of their products but nevertheless went ahead to manufacture and market the same.  Although the Article does not expressly say so, I find that an ordinary and reasonable reader would have understood the Words as carrying such meaning.

41. The Article first began with stating that Madam Wan Yuhua (“Madam Wan”), who was the Chief Executive Officer of the BaWang Group and the wife of Mr Chan Qiyuan, admitted that all along, she knew that BaWang Shampoo Products contained 1,4-dioxane.  The Article then queried why shampoos of such big brand name did contain such substance. The Article also stated that the Plaintiffs made use of the loophole of not having a uniformed definition of “traces” of 1,4-dioxane in shampoos and used this as a shield refusing to admit fault and justifying the presence of that high level of 1,4-dioxane in BaWang Shampoo Products.  Finally, the Article stated that inferior raw materials might be a source for the contamination of 1,4-dioxane and manufacturer might not be willing to increase the costs in purchasing more expensive raw materials to eliminate the presence of 1,4-dioxane in the production of shampoos.

42. In my judgment, the meaning of all these allegations is clear.  The management of BaWang, which was a big brand name, should have known about the harmful effects of 1,4-dioxane as alleged in the Article.  However, since there was no regulatory regime governing the presence of 1,4-dioxane in shampoos, the Plaintiffs, in order to save costs, used raw materials of inferior or at least low quality for the production of BaWang Shampoo Products.  In other words, the Plaintiffs had made a conscious decision of using raw materials of inferior quality for the production of BaWang Shampoo Products, which resulted in the contamination of 1,4-dioxance to a level that was harmful to human health.

43. It is trite law that a hypothetical reasonable reader can draw inference not only from the words used but also from the tenor and context of the whole article.  Given the general context of the Article imputing culpability on the part of the Plaintiffs over the incident, an ordinary and reasonable reader would certainly have understood the Article as accusing the Plaintiffs of knowing the health risks posed by BaWang Shampoo Products; hiding behind unclear regulatory regimes; putting a blind eye to such risk for mercenary reasons; and obstinately refusing to admit fault even when exposed.

44. Further, the words “堅拒認錯” (firmly refused to admit any fault) could only be understood on the premise of a “wrong” on the part of BaWang which they adamantly refused to admit or acknowledge.  The suggestion of “wrong” is echoed by the description of “便以此作擋箭牌” (used this as a shield) which gives a common colloquial nuance of “putting up lame excuse”.  I therefore find that the Words do carry the 1^st meaning as pleaded by the Plaintiffs.

45. I then turn to the 2^nd pleaded meaning which relates to the Plaintiffs’ use of inferior raw materials for the production of BaWang Shampoo Products.  According to the Plaintiffs, the 2^nd meaning went further to impute the following misconduct on the part of the Plaintiffs in manufacturing the Products: (i) the harmful products were the result of using cheap or inferior raw materials; and (ii) the Plaintiffs did so with a mercenary motive to save costs.

46. I also agree with Mr Pow that the above additional stings are borne out by the ordinary and natural meaning of the Words.  It was suggested in the Article that a probable reason for BaWang Shampoo Products containing 1,4-dioxane was the unsatisfactory quality of the raw materials.  This was then connected to the earlier quotation of Dr Lau that raw materials with absolutely no contamination would cost 5-6 times higher and manufacturers may not be willing to shoulder such high costs.  A combination of these two statements gave rise to a clear impression that the problem arose out of “costs implication” and hence imputed a mercenary motive or agenda on the part of the Plaintiffs.  Reasonable readers were likely to draw such inferential meaning given the general context of the Article imputing culpability on the part of Plaintiffs over the incident.

47. In so far as the Defendant is seeking to rely on the phrase “原因可能是” (the reason may be) to argue that readers would not have understood the suggestion relating to the use of inferior raw materials as a definite assertion, I do not accept that such argument can assist the Defendant’s case.

48. In Gatley, the learned authors said the following:[15]

“It is immaterial whether the defamatory imputation is conveyed by words of direct assertion or by suggestion, for insinuation may be as defamatory as an explicit statement and even more mischievous. Similarly words may be defamatory even though they are used in an interrogative form, and in principle the same should apply if they are used in a hypothetical way, so long as in the context in which they are used they may reasonably be interpreted to convey the truth. The tendency and effect of the language, not its form, is the criterion and a defendant cannot defame and escape the consequences by dexterity of style.”

49. In my judgment, though it was not worded as a direct assertion, the Article coveys a clear message to ordinary readers that the Plaintiffs, in order to save costs, used inferior raw materials in the production of BaWang Shampoo Products which resulted in the contamination of 1,4-dioxane to a level that was harmful to human health.  Hence, I find that the Words would have understood by an ordinary and reasonable reader to carry the 2^nd meaning as pleaded by the Plaintiffs.

50. The Plaintiffs’ 3^rd pleaded meaning is that the Plaintiffs ignored or paid little regard to the health and safety of consumers.  In discussing the 1^st and 2^nd pleaded meanings, I have already found that a reasonable reader would have understood the Article as accusing the Plaintiffs of knowing the health risks posed by BaWang Shampoo Products; hiding behind unclear regulatory regimes; putting a blind eye to such risk for mercenary reasons; and obstinately refusing to admit fault even when exposed.  The 3^rd meaning naturally flows from the defamatory stings contained in the 1^st and 2^nd pleaded meanings.  A manufacturer of consumer products accused of the conducts mentioned in the 1^st and 2^nd meanings would certainly be considered by reasonable readers as “paying little or no regard to the health and safety of consumers”.

51. In his submission, Mr Yu argues that the 3^rd meaning is inaccurate because by quoting Madam Wan’s response, the Words only suggest that Bawang has taken a view that consumers should not be concerned about the presence of 1,4-dioxane in its products.  However, such argument ignores the immediate context following the said quotation.  Madam Wan’s viewpoint was immediately negated by a disagreement said to have come from an “expert” (“惟專家不認同”).  Reasonable readers are bound to understand the Article as suggesting that Madam Wan’s point of view was “unacceptable”. The purported viewpoint of Professor Lam was immediately juxtaposed which suggested 20 ppm was “clearly beyond the danger point”.  Far from leaving readers with the view of Madam Wan, the Article clearly presented Madam Wan’s view as being unacceptable when further read in conjunction with the prior nuances of “refusing to admit wrong” and “putting up lame excuse”.

52. For the above reasons, I find that when the Words are read in context, they clearly convey the message that the Plaintiffs paid little regard to the health and safety of consumers.

53. The 4^th and 5^th pleaded meanings relate to the Plaintiffs’ exploitation of legal and regulatory loopholes, and reliance on the absence of regulatory regimes to justify their conducts.  In my judgment, these meanings naturally flow from the earlier meanings.  For the same reasons given in respect of the earlier pleaded meanings, I also find that the Words would have been understood by an ordinary and reasonable reader to carry the 4^th and 5^th pleaded meanings.

54. The Plaintiffs’ 6^th pleaded meaning is that, based on the aforesaid conducts, the Plaintiffs were irresponsible, incompetent and greedy commercial organisation which lacked commercial morality.  Again, the imputation of “irresponsible” is borne out by the sting that the Plaintiffs knowingly manufactured products which is seriously unsafe for consumers’ health, and sought to justify its conduct by exploiting legal and regulatory lacuna or loopholes.  The above conduct, together with the Plaintiffs’ alleged decision to use raw materials of inferior quality, can at the very least be described as “incompetent”.  The Plaintiffs are further accused of using cheap and inferior raw materials out of a mercenary motive, thus an imputation of “greedy” can readily be inferred.  Finally, the imputation of “lacked commercial morality” can be readily inferred if one considers the stings of greediness and lack of regard for customers’ safety.  Hence, I also find in favour of the Plaintiffs in respect of the 6^th pleaded meaning.

55. Based on my finding on the meanings of the Words contained in the Article, it is obvious that these meanings are defamatory to the Plaintiffs.  It cannot be seriously argued that such meanings do not lower the Plaintiffs’ status in society and in the eyes of the public.

(iii)  Words disparaging to products vis-à-vis the Plaintiffs

56. The Plaintiffs rely on two causes of action, namely libel and injurious falsehood.  One of the main distinctions between these two causes of the action is that words which merely disparage a person’s goods or property, but in no way reflect on his personal or trading character, are not defamatory; whereas a defendant who maliciously published a false statement in disparagement of another person’s goods may be liable for malicious falsehood.[16]

57. To oppose the claim for libel, the Defendant therefore seeks to argue that the Plaintiffs cannot sue for such tort as the Words are disparaging merely of the products, namely BaWang Shampoo Products, and not against the Plaintiffs.  Mr Yu relies on the classic exposition of the distinction between libel and malicious falsehood as set out by Lord Esher MR in South Hetton Coal v North-Eastern News Association[17] to argue that the imputation in the Article was only with regard to the Plaintiffs’ goods only, namely BaWang Shampoo Products were hazardous to human health, and so the proper cause of action against the Defendant should be one based on malicious falsehood.

58. However, it is trite law that a statement which on its face is a disparagement of goods may still be read by reasonable readers as also disparaging the claimant’s conduct and reputation.  This is especially true when the claimant in question is the “manufacturer” of the goods disparaged.  In Gatley, the learned authors said:[18]

“Thus an action will lie for an imputation that a trading company is in an insolvent condition, or that the company or its directorate is composed of alien enemies. So an imputation on the goods sold or manufactured by a trading company may involve a reflection on the company in the way of its business, and if a statement be made as to the mode in which a trading company conducts its business, such as to lead people of ordinary sense to the opinion that it conducts its business in a dishonest, improper or inefficient manner, the law is the same as in the case of an individual, and the company can maintain an action without proof of special damage.”

59. A libel on a thing may therefore constitute a libel on a person, because the language may import malpractice or other impropriety on the part of the person who manufactures or sells the goods. That is the reason why the learned authors in Gatley said: “Care should however be taken not to read too much into Lord Esher’s words”[19].  Examples of some of the imputations on goods reflecting the conduct of a company’s business can be found in Gatley at §2.48.

60. There is also a local decision on this particular issue.  In Ming Kee Manufactory Ltd v Man Shing Electrical Manufactory Ltd[20], the plaintiff’s competitor (the defendant) issued a letter to certain companies who were customers of both the plaintiff and defendant, advising that the plaintiff’s electrical products were fitted with unapproved fuselinks and were being withdrawn from the market.  It was argued by the defendant that any disparagement related only to the plaintiff’s goods and hence the plaintiff’s reputation could not be defamed by the words used.  However, DHCJ Jones (as he then was) held that the offending words of the defendant’s letter were not merely disparaging of the plaintiff’s goods but were disparaging of its reputation because the necessary inference from the letter was that the use of unapproved fuses was deliberate or negligent, and the use of unapproved fuses affected a product’s safety and was not something a reputable manufacturer would do.  Under such circumstances, disparagement contained in the defendant’s letter went well beyond merely that of the plaintiff’s goods.

61. Applying the same approach, looking at the nature of the accusation that “Bawang Shampoo Products can cause serious illness and even cancer” in the entire context of the Article, reasonable readers could only have read them as imputing culpability on the part of the Plaintiffs: they were either deliberate, reckless or negligent in respect of the safety of their products.  The accusation that “Bawang Shampoo Products can cause serious illness and even cancer, and is unsafe for consumer use” is clearly something which a reasonable reader of the Article would regard as something which a reputable manufacturer would not do.  In other words, it is clear that a reasonable member of the public would no longer regard a manufacturer of cancerous and unsafe shampoo products as reputable.  The presentation of this accusation went beyond mere disparagement of Bawang Shampoo Products alone.  The Words surely reflected badly on the trade and business reputation of the Plaintiffs.

62. Furthermore, the aforesaid accusation is just one of the several imputations contained in the Article.  The Article also accused the Plaintiffs of:

(i)  knowingly manufactured and sold such unsafe products without regard to customers’ health;

(ii)  using cheap and inferior raw materials in order to make mercenary gain;

(iii)  exploiting the legal lacuna or loopholes as excuse for their conduct; and

(iv)  refusing to admit their wrong when confronted.

63. All these additional allegations went far beyond disparaging Bawang Shampoo Products and surely reflected on the way in which the Plaintiffs carried on their business of manufacturing and selling the Products.  Even the Defendant’s own witness, Madam Lam, agrees in her testimony that if someone suggests that a product is harmful to health, and that this product was manufactured by a company, it is a serious accusation against the company itself.  Hence, the Words are clearly actionable in libel.

(iv)  Reference to the Plaintiffs

64. Another requirement for a cause of action in libel is that the defamatory words must be with reference to the claimant or claimants.

65. There is no dispute that the Article refers to the 1^st Plaintiff being the holding company of the BaWang Group.

66. There is no express reference to the 2^nd Plaintiff in the Article.  However, the Plaintiffs rely on the following evidence to show that persons acquainted with the Plaintiffs would reasonably believe that the 2^nd Plaintiff was referred to in the Article:

(i)  information provided in the prospectus and the annual reports of the 1^st Plaintiff showing that the 2^nd Plaintiff was the manufacturer of BaWang Shampoo Products and the main operating subsidiary in the BaWang Group; and

(ii)  the packaging of BaWang Shampoo Products indicating that the 2^nd Plaintiff was the manufacturer of the products.

67. So far as the reference to the 2^nd Plaintiff is concerned, the following legal principles are relevant:

(i)  the primary test is “whether the plaintiff may reasonably be understood to be referred to by the words”;[21]

(ii)  where the plaintiff is not specifically named in the defamatory statements, the test is whether persons acquainted with the plaintiff would reasonably in the circumstances believe that the plaintiff was referred to in the article;[22]

(iii)  it is not necessary to adduce direct evidence from persons acquainted with the plaintiff to say that they understood the words as referring to the plaintiff, and the court is entitled to infer from all the evidence that there were probably persons (to whom the words were published) who would reasonably believe that the defamatory statements refer to the plaintiff;[23]

(iv)  a corporation may be defamed and its business damaged even though the readers have no idea of its formal legal name.  It is sufficient to prove publication to persons who “would know who the owners were”, and this requires identification of ownership and the plaintiff, but does not require knowledge of the plaintiff’s name;[24]

(v)  if a defamatory statement made of a “class or group” can reasonably be understood to refer to every member of it, each one has a cause of action[25]: an example is when a defamatory statement is made in respect of a “group of companies”;[26] and

(vi)  however where the words complained of do not specifically indentify the claimant company and that company is one of several in a larger group with similar names, the claim may fail unless a reasonable person could think that the claimant company was sufficiently referred to or that the words sufficiently identified all, or some of, the companies in the group.[27]

68. The Article refers to “BaWang”, though there is no express reference of “BaWang” as a group of companies.  The 1^st Plaintiff is the holding company of the BaWang Group listed in the Hong Kong Stock Exchange.  Being a listed company of considerable size, a reasonable reader acquainted with the Plaintiff should have understood that the BaWang businesses in Hong Kong and the Mainland were actually operated through its subsidiaries, though they might not know the exact names of the various subsidiaries involved.  Even someone acquainted with Plaintiffs might not have bothered to read all the details in the prospectus and the listing documents of the 1^st Plaintiff in order to ascertain the names of each subsidiary.

69. From the nature of the alleged defamatory statements and the context of the entire Article, which was about an attack on the quality of BaWang Shampoo Products, reasonable readers should have understood that the allegations contained in the Article were directed against the holding company and the other subsidiaries in the BaWang Group responsible for the manufacturing and marketing of BaWang Shampoo Products in Hong Kong including the 2^nd Plaintiff.  As the 2^nd Plaintiff was the main operating subsidiary responsible for the manufacturing of BaWang Shampoo Products in Hong Kong, I find that the Article did refer to the 2^nd Plaintiff even though the general public might not know its specific name.

70. Mr Yu submits that the 2^nd Plaintiff cannot maintain a cause of action for libel because there is no direct evidence that readers of the Article identified the publication with the 2^nd Plaintiff.  Relying on Palace Films Pty Ltd v Fairfax Media Publications Pty Ltd[28], a decision of the Supreme Court of New South Wales, Mr Yu argues that it is necessary for the 2^nd Plaintiff to establish that it was identified by readers with knowledge of extrinsic facts at the time of the publication or read the public information concerning the 2^nd Plaintiff.

71. I disagree.  Palace Films is a case applying the principles of Channel Seven Sydney Pty Ltd. v. Parras[29] upon its specific facts.  In Palace Films, the court held that reference in the statement complained of to “Palace Film” was not enough, without resort to extrinsic facts, to establish sufficient reference to the plaintiff company.  As any person with knowledge of the extrinsic facts would also have known that Palace Films Pty Ltd permitted the name “Palace Films” to be used in connection with film distribution to another corporate entity, Palace Enterprises Pty Ltd, which conducted the film distribution business discussed in the matters complained of, the particulars provided were therefore not capable of identifying the plaintiff.  In other words, there might be several entities which might be related to the name “Palace Film”, and there was uncertainty as to which entity the libel referred to.

72. The present case is quite different.  There was no confusion that the attack in the Article was directed against the holding company and the subsidiaries responsible for manufacturing and marketing of BaWang Shampoo Products in Hong Kong, and so the Plaintiffs should not be required to adduce other extrinsic evidence to prove that the readers had detailed knowledge of the corporate structure of the BaWang Group and the specific names of the subsidiaries responsible for manufacturing and marketing of BaWang Shampoo Products in Hong Kong.

73. Furthermore, Palace Films reinstates the principle that in order to establish reference to the claimant, in case of uncertainty (which is not the case here), the claimant only need to establish that at least one person to whom the article was published understood from extrinsic facts that the article referred to that plaintiff.[30] In the present case, there is no dispute that the 2^nd Plaintiff was clearly stated in the listing documents of the 1^st Plaintiff and the packaging of BaWang Shampoo Products as the manufacturer of the Products.  As listing of the 1^st Plaintiff had attracted a lot of publicity in Hong Kong and BaWang Shampoo Products were widely sold and consumed in Hong Kong, I agree with Mr Pow that such undisputed evidence is sufficient for the court to draw an inference that there must have been at least one person who would have knowledge of the 2^nd Plaintiff being the manufacturer of BaWang Shampoo Products and who would have read the Article.  In other words, it is not necessary to call direct evidence from a particular reader to establish reference, which can be established by implication from the established facts in the case.

74. For the above reasons, I am satisfied that an ordinary reader would have understood the Words as referring to the 2^nd Plaintiff which was the manufacturing arm of the BaWang Group.  Although such finding allows the 2^nd Plaintiff to maintain a cause of action for libel against the Defendant, it may have serious implication on the issue of quantum as to whether the 1^st Plaintiff should be allowed to recover the loss suffered by its subsidiaries responsible for the distribution of BaWang Shampoo Products in Hong Kong.  I will deal with such matter in length in the latter part of this Judgment on quantum.

75. It is not disputed that the Article was published by the Defendant.  As the Words in the Article are defamatory with reference to the Plaintiffs, they can maintain a cause of action in libel against the Defendant unless the latter can establish any of the pleaded defences.

III:  DEFENCE OF JUSTIFICATION

76. The first pleaded defence is justification.  The burden would be on the Defendant to establish such defence.

(i)  Lucas-Box meanings pleaded by the Defendant

77. The Defendant has pleaded alternative Lucas-Box meanings[31] about the contents of the Article and is seeking to justify those meanings:[32]

(i)  the presence of 1,4-dioxane in shampoo can harm the health of those using such shampoo, including harming the users’ livers or kidneys, or even causing cancer;

(ii)  several kinds of shampoo sold under the brand name of “霸王” (“BaWang”) contain the substance 1,4-dioxane;

(iii)  the levels of 1,4-dioxane in some of the shampoos sold under the brand name of “霸王” (“BaWang”) were greater than those found in various other shampoos sold to the public;

(iv)  a possible cause of the presence of 1,4-dioxane in shampoos is the poor quality of raw materials used in the production of the shampoos;

(v)  some manufacturers and/or vendors of shampoos, such as Rejoice and Acene, have expressed willingness to reduce the content of 1,4-dioxane in their products but the Plaintiffs take the view that the presence of 1,4-dioxane in shampoo is unavoidable, minimal and not harmful to users; and

(vi)  although the Plaintiffs knew that shampoos sold under the brand name of “霸王” (“BaWang”) contained 1,4-dioxane, the Plaintiffs were not concerned about the health and safety of consumers who used or purchased their products because the levels of 1,4-dioxane in their shampoos were, according to the Plaintiffs, within legal safety limits; and

(vii)  the Plaintiffs did not acknowledge that the presence of 1,4-dioxane in shampoo could harm the health of those using such shampoo.

78. Based on my finding on the meanings of the Article, I am satisfied that the Words, in their ordinary and natural meaning, bear stings defamatory to the Plaintiffs.  In considering the defence of justification, the court then has to determine: (i) whether Words bear the Lucas-Box meanings as contended for by the Defendant; (ii) if so, whether as a matter of fact that any of all of the Lucas-Box meanings is true; and (iii) whether the proved meanings are sufficient to meet the libellous stings conveyed by the Words in the Article.  It is only when each of the stings is properly met with true facts that the defence of justification succeeds.[33]

79. The 1^st pleaded Lucas-Box meaning seems to suggest that any concentration or amount of 1,4-dioxance in shampoo can harm the health of users, even causing cancer.  However, such a sweeping statement cannot be true because the Defendant’s own toxicology expert accepts that concentration below a certain limit has “de minimis” health risk.

80. On the first day of the trial, the Defendant therefore added the following particulars in the pleading restricting the pleaded meaning:[34]

“Regular and continuous use of shampoo which contains 1,4-dioxane at 6.4 ppm or 10 ppm or 27 ppm of 1,4-dioxane materially increases the risk of developing cancer.”

81. Hence, the meaning that the Defendant now seeks to justify is that shampoo containing 6.4 to 27 ppm of 1,4-dioxane can “materially increase” the risk of cancer.  As I see it, this issue is the main battlefield between the parties so far as the defence of justification is concerned.  The question has to be resolved with the help of expert evidence on toxicology, which I will deal with in details in the latter part of this Judgment.

82. The 2^nd pleaded Lucas-Box meaning is uncontroversial as it is common ground that BaWang Shampoo Products did contain “traces” of 1,4-dioxane.  However, there is a serious dispute between the parties as to whether such level of 1,4-dioxane does pose a health risk to consumers.  It is clear from the Words in the Article that the level or concentration found in BaWang Shampoo Products was hazardous to human health, and so the justification of the 2^nd Lucas-Box meaning per se cannot meet the stings in the Article.  Ultimately, the court has to resolve the question as to whether regular and continuous use of BaWang Shampoo Products would materially increase the risk of developing cancer.

83. The Defendant’s 3^rd pleading meaning is that the levels of 1,4-dioxane in some of the BaWang Shampoo Products were greater than those found in various other shampoos sold to the public.

84. Again the issue here is not whether the levels of 1,4-dioxane found in BaWang Shampoo Products were higher than those found in other shampoos.  The most important sting conveyed by the Article is that the levels of 1,4-dioxane found in BaWang Shampoo Products were dangerous to human health, and so the comparison of the levels of 1,4-dioxane in different brands is neither here nor there.  Ultimately, the court has to resolve the basic question so as whether regular and continuous use of the BaWang Shampoo Products would materially increase the risk of developing cancer.

85. However, one of the factual issues which may be relevant to the 2^nd and 3^rd Lucas-Box meanings is the reliability of the test results provided by the complainant “Mr Chan” to Madam Lam.  According to the SGS test reports, 6 bottles of shampoo of different brands were found to have contained the following levels of 1,4-dioxane: (i) BaWang: 27 ppm; (ii) Rejoice: 8.8 ppm; (iii) Acene: 2.9 ppm; (iv) O’Naomi, Mannings, Watson: not detected.  As 3 bottles of BaWang Shampoo Products independently tested by the Defendant were later found to have contained only 10 ppm of 1,4-dioxane, there are factual issues as to whether the BaWang shampoo submitted by “Mr Chan” for testing (“the Dubious Sample”) did in fact contain 27 ppm of 1,4-dioxane, and whether the levels of 1,4-dioxane found in BaWang shampoo Products were indeed higher than those found in other brands.

86. In my judgment, it is not necessary for the court to resolve these factual issues in considering the question of justification.  So far as this particular defence is concerned, in order to meet the stings conveyed by the Article, the key question is whether regular and continuous use of BaWang Shampoo Products would materially increase the risk of developing cancer or other illnesses.  Whether BaWang Shampoo Products did contain a higher level of 1,4-dioxane than the other brands is quite irrelevant.  In answering such key question, both toxicology experts agree that 10 ppm, contrary to what was said in the Article, is not the proper safety limit.  According to the Plaintiffs’ expert, shampoo containing less than 100 ppm, or desirably 30 ppm, of 1,4-dioxane should be considered as safe, whereas the Defendant’s expert is of the view that shampoo containing as little as 6.4 ppm is still unsafe.  In such circumstances, whether the Dubious Sample did contain as much as 27 ppm of 1,4-dioxane becomes an academic issue.

87. On the other hand, whether the Article should contain the reference to 27 ppm may be a matter relevant in considering the issue of responsible journalism.  In the latter part of this Judgment dealing with the defence of publication in the public interest, I will deal with the possibility of tampering with the dubious samples submitted by “Mr Chan” for testing and whether Madam Lam should include the test result of 27 ppm in the Article.

88. In any event, the Defendant would bear the burden of establishing the facts in support of the defence of justification.  If, contrary to my analysis above, it would be necessary for the court to make factual findings as to: (i) whether the Dubious Sample submitted by “Mr Chan” for testing did contain 27 ppm of 1,4-dioxane; and (ii) whether the level of 1,4-dioxane found in BaWang shampoo was higher than those of the other brands, it would be the burden on the Defendant to establish these facts.  Based on the various dubious features about the samples submitted by “Mr Chan” for testing as pointed out by me in the latter part of this Judgment on the issue of responsible journalism, I find that the Defendant has failed to prove these facts in support of the defence of justification.[35]

89. The Defendant’s 4^th pleaded meaning is that a possible cause of the presence of 1,4-dioxane in BaWang Shampoo Products is the poor quality of raw materials used in the production of the shampoos.

90. The evidence from the chemical engineering experts is most relevant to this particular issue.  The Plaintiffs’ expert is Dr Richard Parent (who is also the Plaintiffs’ toxicology expert)(“Dr Parent”) and the Defendant’s expert is Mr Robin Jones (“Mr Jones”). There is an issue as to whether Dr Parent is a qualified expert on chemical engineering.  As there is substantial interplay between toxicology and chemical engineering and Dr Parent had related working experience in this field, I accept him to be a qualified expert in chemical engineering.  In any event, I am putting very little weight on his evidence because of the lack of relevancy to the remaining live issues about the manufacturing process of BaWang Shampoo Products.

91. The Defendant seeks to rely on the particulars pleaded in §§10(15) and 10(15A) of the Defence[36] to justify its 4^th Lucas-Box meaning.

92. In §10(15), the Defendant relies on “the lack of 1,4-dioxane in 3 of the shampoo allegedly tested by SGS” as the only factual basis of an inference that the cause of presence of 1,4-dioxane in Bawang Shampoo Products was possibly caused by poor quality of raw materials. I agree with Mr Pow that this is plain non sequitur.

93. First, it is not disputed that the fact that SGS stated the 1,4-dioxane level to be “not detected” in a particular shampoo sample does not mean that it contains no 1,4-dioxane.  It just means that it contains less than 2.5 ppm of 1,4-dioxane, which is the minimum detectable level of the SGS testing equipment.

94. Second, as agreed by Mr. Jones himself, it is not possible to ensure the complete removal of 1,4-dioxane from the ammonium laureth sulfate (“ALS”), which is indisputably a very common ingredient for making shampoo.  So even the best ALS available in the world would contain traces of 1,4-dioxane.  Its presence in shampoo products is inevitable.  Therefore, the “quality” of the raw materials is not the cause of the presence of 1,4-dioxane in Bawang Shampoo Products.  Rather, it is the “type” of raw material involved.

95. Third, the evidence of Madam Jiang Ri Qiong (蔣日瓊)(“Madam Jiang”), the formula engineer of the BaWang Group, dispels the Defendant’s allegation that the 2^nd Plaintiff manufactured BaWang Shampoo Products using raw materials of poor quality.

96. As stated in the Affirmation of Madam Jiang, the presence of 1,4-dioxane in Bawang Shampoo Products is due to a necessary ingredient, ALS, which forms 25% by volume of the final shampoo products. This is confirmed by the experts.  The 2^nd Plaintiff had a supplier vetting process, and Hunan Resun “湖南麗臣”(“Hunan Resun”), the chosen supplier which had about 20 years’ experience in supplying ALS, had both the requisite equipment and experience in producing high quality ALS and controlling the amount of 1,4-dioxane.  The evidence also shows that Hunan Resun had a good track record with many international brands including Johnson & Johnson, Colgate, Procter & Gamble and Uni-lever.

97. The ability of Hunan Resun to be a reliable supplier of high quality raw material is not challenged in cross-examination.  In particular, Mr. Jones opines that Hunan Resun gives the appearance of a modern industrial plant and is capable of producing raw materials to a good standard.

98. As at July 2010, Hunan Resun had contractually guaranteed to the 2^nd Plaintiff that the ALS supplied would not contain more than 40 ppm of 1,4-dioxane.  As ALS forms 25% by volume of the final shampoo product, the level of 1,4-dioxane in the final product should be in the region of 10 ppm.[37]  According to Madam Jiang, the contractual guarantee of 40 ppm was technically the best that ALS suppliers in the Mainland vetted by the 2^nd Plaintiff were able and willing to contractually guarantee, a matter which has not been challenged by the Defendant.  There is also no evidence that in 2010, there was another ALS supplier which was willing  to guarantee a 1,4-dioxane concentration of lower than 40 ppm.  Under such circumstances, the raw material ALS sourced from Hunan Resun that the 2^nd Plaintiff used in July 2010 in manufacturing BaWang Shampoo Products was already the best quality that the 2^nd Plaintiff could obtain in term of the minimisation of 1,4-dioxane concentration.  Further, Madam Jiang testifies that over the years, the 2^nd Plaintiff kept negotiating and working with the ALS suppliers, and with advancement in technology and skills, Hunan Resun was able to guarantee as low as 20 ppm of 1,4-dioxane since early 2013.

99. There is also complaint that there was inaction on the part of BaWang to deal with the problem relating to presence of 1,4-dioxane in shampoo.  According to the Defendant, it was doubtful whether the management of BaWang had given specific instruction to Madam Jiang to reduce the 1,4-dioxane in BaWang Shampoo Products, in particular after the publication of the newspaper report about the presence of 1,4-dioxane generally in the shampoos marketed in the Mainland in Guangzhou Daily on 15 March 2010.  Further, Madam Jiang does not know about all the 1,4-dioxane incidents mentioned in the news articles obtained by Madam Lam during her research.

100. But according to Madam Jiang, from 2008 onwards, the 2^nd Plaintiff had been trying to improve the shampoo formula and to see if the level of 1,4-dioxane could be reduced, including the formula suggested in an email from Acene.  However, doing so would compromise the quality of the shampoo products, for example making the final product not stable and not foamy, resulting in the possible non-compliance with the shampoo implementation standard QB/T 1974 required by the Mainland authorities.  Mr Jones agrees that reformulation would involve a change in the property or quality of the shampoo.

101. During cross-examination, it has been suggested that the 2^nd Plaintiff could lower the contractual guarantee of 40 ppm to, say 30 or 20 ppm, by just increasing the price.  This is because Hunan Resun could either just disregard a batch which contained more than the contractually guaranteed concentration, or mix a “good” batch with a “bad” batch (as suggested by Mr. Jones) to get an average which is within the contractual limit.  But as Madam Jiang explains, it is not a question of price, but a question of technology.  She has already explained that as at 2010, the state of technology of Hunan Resun was that it could not make a guarantee of concentration of lower than 40 ppm.  The fact that Hunan Resun could, in some batches, attain less than 40 ppm is neither here nor there, because there was no way to ensure that Hunan Resun would be able to meet a contractual guarantee of, say 20 ppm, even if the 2^nd Plaintiff was willing to pay more.  For example, what if, despite its best efforts, Hunan Resun was only able to produce batches of ALS with over 30 ppm in a particular month?  It means that there would be no delivery to the 2^nd Plaintiff in that month.  Not only would Hunan Resun be in breach of the contractual guarantee, but the 2^nd Plaintiff would not have enough raw materials to continue production.  No reasonable commercial entities in the shoes of Hunan Resun and the 2^nd Plaintiff would have entered into such a deal in the first place.

102. Further, Mr Yu relies on the supply contracts for ALS made between the 2^nd Plaintiff and Hunan Resun from 2009 to 2011 to show that there was a corresponding increase in the contract price when Hunan Resun was prepared to lower the contractual guarantee for the presence of 1,4-dioxane in the ALS supplied.  According to Mr Yu, this tends to support that the Plaintiffs were able to obtain better ALS by paying more.

103. However, such argument has ignored the evidence of Madam Jiang that, Hunan Resun, being one of the most reliable suppliers of the ALS in the Mainland at the material time, was only prepared to give a contract guarantee of 40 ppm.  As mentioned above, it had to do with technology rather than price.  Further, there are a lot reasons to explain for the increasing trend in contract price, for example the increasing labour and transportation costs faced by the industries in the Mainland at the material time as shown by the evidence from the accounting experts mentioned in the latter part of this Judgment.[38]

104. There are some attacks against the credibility of Madam Jiang as a witness.  According to the Defendant, there is some confusion in her evidence as to whether BaWang had consulted toxicologists relating to the presence of 1,4-dioxane in shampoo, whether anyone from BaWang’s management had given her instruction in 2010 or any time to reduce or eliminate 1,4-dioxane in shampoo, and the exact time when the BaWang Group negotiated with Hunan Resun for a limit of 20 ppm for the ALS to be supplied in 2011.

105. In my judgment, there is no reason for me to doubt the credibility of Madam Jiang’s evidence. Despite the vigorous cross-examination, she answers the questions in a straightforward manner without any exaggeration.  Madam Jiang is only a formula engineer, and yet she has been asked various questions covering vast area including the negotiation for the contracts for the supply of raw materials and her knowledge about what happened around the world relating to the presence of 1,4-dioxane in shampoo. She has been trying her best to answer the questions.  It is not surprising that she is confused about a few facts, and I do not accept that such immaterial confusion would undermine the overall credibility of her evidence.

106. In such circumstances, the Defendant has simply failed to discharge the burden of proving that the Plaintiffs used raw materials of poor or inferior quality in manufacturing the BaWang Shampoo Products.

107. The Defendant also relies on the particulars pleaded in §10(15A) of the Defence to justify the 4^th pleaded Lucas-Box meaning, saying that the Plaintiffs had not taken all necessary steps to devise manufacturing process or to adopt a formula or process that would either ensure the absence of 1,4-dioxane in their products or one which removed or reduced the amount of 1,4-dioxane to a minimum.

108. Again, the justification of these particulars does not meet the defamatory stings of the Words.  The message conveyed by the Article is not just that 1,4-dioxance is an unpleasant substance and manufacturers should minimize the level of 1,4-dioxane as much as possible.  The Article went further and positively alleged that BaWang Shampoo Products were hazardous to human health.  Obviously, if BaWang Shampoo Products were unsafe, then it was justified for the Defendant to complain that the Plaintiff should have devised a manufacturing process or adopted a formula to ensure the absence of 1,4-dioxane.  However, if the Products were not unsafe, there was no obligation on the Plaintiffs and no commercial reason for them to eliminate 1,4-dioxane from its products at all costs.  Furthermore, as confirmed by Dr Parent, ALS has certain function in the manufacturing process of shampoos.  In fact, Madam Jiang’s unchallenged evidence is that there were certain specific regulations relating to shampoos in the Mainland, for example those relating to the stability and foaming property of the shampoos.  The use of ALS would help the Plaintiffs to meet these specific regulations in the Mainland.

109. The Defendant also alleges that because the Good Manufacturing Practice (“GMP”), which has been defined by regulatory authorities in the United States and elsewhere as good manufacturing practice for the cosmetic industry, was not in place in the Plaintiffs’ production line in 2010, the Plaintiffs were not able to detect any presence of 1,4-dioxane in their products and reject them.  But in my judgment, the Plaintiffs had no obligation to detect and reject products containing any 1,4-dioxane.  As agreed by Mr. Jones, even with the GMP in place, it is still not possible to ensure the complete removal of 1,4-dioxane from ALS.  As further explained by Madam Jiang, the amount of 1,4-dioxane in BaWang Shampoo Products depended on the amount present in the raw material ALS.  The Plaintiffs’ mandate was to keep the level of 1,4-dioxane in the end products below 30 ppm, reference taken from the Australian NICNAS[39] which was considered to be safe.  The 2^nd Plaintiff would require Hunan Resun to check every batch of ALS delivered.  The 2^nd Plaintiff would also periodically send samples of finished products to independent agent to test for the level of 1,4-dioxane to ensure that the safe level would be achieved.  Although the Plaintiffs did not conduct any in-house examination to monitor the level of 1,4-dioxance in the final products, which apparently is Mr Jones’ only criticism on BaWang’s production process, the Plaintiffs did monitor the quality of ALS supplied by Hunan Resun.

110. In any event, the regulatory authority and the cosmetic industry in the Mainland did not require shampoo products to comply with GMP in 2010, a matter which is accepted by Mr Jones himself.  The Plaintiffs eventually achieved that standard in 2013.

111. Obviously, BaWang was able to ensure the safety of its final products in 2010 even without adopting the GMP.  This is supported by the 6.4 ppm and 10 ppm test results given by SGS.  As Mr. Jones opines, such small variance in the level of 1,4-dioxane in the raw material is normally expected in the manufacturing process.  Mr. Jones’ concern appears solely relates to the 27 ppm test result.  As further elaborated below, the Dubious Sample is highly questionable and any Defendant’s argument or criticism which is based on the 27 ppm result should therefore be approached with caution.

112. The Defendant also complains that the 2^nd Plaintiff “overly depended upon” Hunan Resun, because it did not allow for “possibilities” such as “mistakes in chemical analysis such as sampling errors undetected or other accident; presence of other ingredients which could contain 1,4-dioxane, such as ethylene glycol distearate and deterioration of ingredients before use”[40].  However, as Mr. Jones admits in cross-examination, these are only theoretical possibilities in general cases.  There is simply no evidence to show that these theoretical possibilities actually occurred in the present case.[41]  On the contrary, it is clear from Madam Jiang’s evidence and the various testing reports adduced that the 2^nd Plaintiff had maintained a record of controlling the presence of 1,4-dioxane in its shampoo products to no more than 10 ppm.

113. Finally, the Defendant criticises the 2^nd Plaintiff’s supplier evaluation protocol, in that it is “more appropriate” for BaWang to focus on the weight or quality of the ingredient, rather than the value of orders placed.[42]  However, for the ALS sourced from Hunan Resun, it falls anyway within the most stringent category (Group A) of suppliers.[43] Hence, this complaint does not apply.  In any event, the Defendant has not raised this complaint during the cross-examination of the Plaintiffs’ witnesses.

114. The Defendant’s 5^th pleaded Lucas-Box meaning is that some manufacturers or vendors of shampoos, such as Rejoice and Acene, had expressed willingness to reduce the contents of 1,4-dioxane in their products but the Plaintiffs took the view that the presence of 1,4-dioxane in BaWang Shampoo Products was unavoidable, minimal and not harmful to users.  It should be noted that the Article also stated that Rejoice and Acene considered the quantities of 1,4-dioxane in their products, namely 2.9 to 8.8 ppm, were very minimal (微乎其微) and not harmful to the health of users.

115. This meaning merely purports to set out the alleged difference in attitude between BaWang and other brands as to how to deal with the issue.  In any event, any attempt to justify such meaning cannot assist the Defendant’s case, as this meaning is essentially not defamatory towards the Plaintiffs who took the view that the presence of minimal quantity of 1,4-dioxane in shampoo is unavoidable and not harmful to the health of users.  Justifying the truth of this statement does not meet the stings of the libel mentioned in the earlier part of this Judgment.[44] Again this begs the same question as to whether regular and continuous use of BaWang Shampoo Products would materially increase the risk of developing cancer.  If the Products were safe, the view taken by the Plaintiffs was justified.

116. The 6^th pleaded Lucas-Box meaning is that although the Plaintiffs knew that BaWang Shampoo Products contained 1,4-dioxane, the Plaintiffs were not concerned about the health and safety of consumers who used or purchased their products because the levels of 1,4-dioxane in their shampoos were, according to the Plaintiffs, within legal safety limits.

117. Again, the justification of such meaning depends on the safety of the BaWang Shampoo Products. If the Products were not safe, there might be some justification for the Defendant to say that the Plaintiffs were not concerned about the health and safety of consumers.

118. I would make one further observation here.  No matter whether BaWang Shampoo Products were safe judging from the new scientific materials available after the publication of the Article, the evidence shows that the Plaintiffs were concerned about safety of their consumers taking into account the scientific materials available before the publication of the Article.  According to Madam Jiang, even though as at July 2010, the Mainland State Food & Drug Administration (“SFDA”) had not formally specified a legal safety threshold for the level of 1,4-dioxane in shampoo products, she had already studied the Australian NICNAS Report and adopted the 30 ppm limit stated therein.[45] BaWang was not just relying on a standard proposed by regulatory authority.  It was relying on a scientific study (which apparently was the only one available at that time apart from the EU and Health Canada Reports[46]) about the safety level of 1,4-dioxane in shampoo.  Further, Madam Jiang accepted the 40 ppm contractual limit guaranteed by Hunan Resun on the basis that the end product would not contain more than 10 ppm of 1,4-dioxane, well under the 30 ppm threshold recommended by NICNAS.

119. At the trial, even Madam Lam agrees that any institutions which followed the NICNAS and EU standards cannot be said to be irresponsible.

120. There is some suggestion at the trial that the Plaintiffs should have engaged a toxicologist to advise them at the relevant time.  However, it is common ground that all the relevant and up-to-date scientific reports available at the time of the publication of the Article (i.e. July 2010) specifically on the safety of 1,4-dioxane in consumer products, including reports from the regulatory authorities in Australia, EU and Canada, showed that shampoo containing as much as 27 ppm of 1,4-dioxane was considered as safe.  If the Defendant had engaged a toxicologist, that would probably have been the conclusion of the toxicologist by that time.

121. The 7^th pleaded Lucas-Box meaning is that the Plaintiffs did not acknowledge that the presence of 1,4-dioxane in shampoo would harm the health of those using such shampoo.  Again, this begs the key question as to whether regular and continuous use of BaWang Shampoo Products would materially increase the risk of developing cancer.  If the Products were safe, there was no reason for the Plaintiffs to make such acknowledgment.  On this issue, I would also repeat the same observations I made in respect of the 6^th pleaded Lucas-Box meaning about the safety limit recommended by the most up-to-date scientific reports available at the time of the publication of the Article.

(ii)  Toxicological evidence on the alleged harmful effect of 1,4-dioxane in BaWang Shampoo Products

122. Having addressed the 7 Lucas-Box meanings pleaded by the Defendant, I return to the key question in the defence of justification, namely whether regular and continuous use of BaWang Shampoo Products would materially increase the risk of developing cancer.  As I see it, whether the Defendant can successfully establish the defence of justification hinges upon the determination of this issue.

123. The court can only resolve this issue with the assistance of the expert evidence on toxicology. The Plaintiffs’ toxicology expert is Dr Parent.  The Defendant had originally engaged the late Dr Bojan Flaks (“Dr Flaks”) as its expert on toxicology. Unfortunately, Dr Flaks passed away before the trial and so the Defendant’s present expert is Dr William Sawyer (“Dr Sawyer”).

124. There is no dispute that both Dr Parent and Dr Sawyer are qualified experts in toxicology.

125. The experts have prepared the following reports for this case:

(i)  expert report of Dr Parent dated 10 February 2012 (“Dr Parent’s Report”);

(ii)  expert report of Dr Sawyer dated 9 October 2014 (“Dr Sawyer’s Report”);

(iii)  1^st and 2^nd supplemental expert reports of Dr Parent dated 7 November 2014 and 13 February 2015 respectively (“Dr Parent’s 1^st and 2^nd Supplemental Reports”); and

(iv)  joint expert report of Dr Parent and Dr Sawyer dated 26 January 2015 (“Toxicology Joint Expert Report”).

126. Dr Sawyer has also reviewed the following earlier expert reports prepared by Dr Flaks and he agrees with Dr Flaks’ opinion as stated in these reports:

(i)  expert report of Dr Flaks dated 26 January 2012;

(ii)  joint expert report of Dr Parent and Dr Flaks dated 26 July 2012 (“Parent/Flaks Joint Expert Report”); and

(iii)  expert report on “not agreed issues” of Dr Flaks dated 15 August 2012.

127. The toxicology experts have been asked to express their opinion on 8 issues:

(a)  Is 1,4-dioxane a probable carcinogen for human beings?

(b)  Might 1,4-dioxane be genotoxic to human beings?[47]

(c)  What harm can 1,4-dioxane cause to humans? Is there any medical evidence or reported case on such harm caused to humans?

(d)  What level of intake of, exposure to or contact with (through skin, scalp or inhalation or otherwise) 1,4-dioxane could cause harm to humans?  Please state also any assumptions underlying such answer including but not limiting to time, period and concentration.

(e)  Is raw material of poor quality a possible cause for the presence of 1,4-dioxane in shampoos? Are there other possible causes of the presence of 1,4-dioxane in shampoos?

(f)  Is it possible to ensure that shampoos do not contain 1,4-dioxane or is the presence of 1,4-dioxane unavoidable?

(g)  What are the regulations, requirements and safety standards in relation to the presence of 1,4-dioxane in cosmetic products including shampoo in USA, EU and (where applicable) other jurisdictions identified by the expert?

(h)  Will the presence of 6.4 ppm of 1,4-dioxane in commercial shampoo cause serious illness or cancer to human beings on the assumption that the shampoo (a) comes into contact with the human skin for a short period of time and will be rinsed off and (b) on a regular basis?

128. Since all these questions are interrelated, I propose to deal with them together.

(a)  What is 1,4-dioxane?

129. According to the data in the Toxicological Review of 1,4-dioxane and Technical Fact Sheet published by the United States Environmental Protection Agency (“USEPA”) in September 2013 and January 2014 respectively, 1,4-dioxane is a synthetic industrial chemical that is completely miscible in water.  It is primarily used as a solvent for celluloses, organic products, lacquers, paints, varnishes, paint and varnish removers, resins, oils, waxes, dyes, cements, fumigants, emulsions, and polishing compositions.  1,4-dioxane has been used as a solvent in the formulation of inks, coatings, and adhesives and in the extraction of animal vegetable oil.

130. 1,4-dioxane can also be found as a contaminant in ethoxylated surfactants (ALS is a typical example of an ethoxylated surfactant) that are used in cosmetics, detergents, shampoos and food additives.[48] 1,4-dioxane is formed as an accidental by-product of the ethoxylation process.  Ethylene oxide is added in the manufacturing process of various skin care products, including shampoos, to reduce the harmful effect of other chemical used in the manufacturing process.  For example, Sodium Lauryl Sulphate, a chemical used in shampoo as foaming agent that is harsh on the skin, is often converted to the less harsh chemical Sodium Laureth Sulphate (the “eth” denotes ethoxylation), which can contaminate the product with 1,4-dioxane.[49]

131. There is no dispute that 1,4-dioxane is found to be an animal carcinogen.[50]

132. However, both experts agree there is insufficient reliable published scientific or medical data or direct epidemiological evidence to show that 1,4-dioxane is carcinogenic to human.[51] It is impossible to conduct human studies in respect of the carcinogenicity of 1,4-dioxane.  Even with animals studies, global ethical and economic considerations dictate that animal carcinogenicity testing be carried out using limited number of laboratory animals.[52]  Hence, the lack of adequate human epidemiological studies is not the result of negative studies but rather a lack of conclusive human studies.

133. Many international regulatory authorities have classified 1,4-dioxane as “likely to be carcinogenic to human”, “reasonably anticipated to be a human carcinogen” or  “possibly carcinogenic to humans”, including USEPA, US Department of Health and Human Services, and International Agency for Research on Cancer.

134. When a chemical is identified as an animal carcinogen, both experts agree that it is the policy of most regulatory bodies that the chemical should be designated as a possible or probable human carcinogen.  If a chemical has been shown to be carcinogenic to humans in scientifically sound clinical and epidemiological studies, then it is designated as a human carcinogen.  1,4-dioxane has not been designated as a human carcinogen or genotoxin.

135. The toxicology experts further agree that the designation of 1,4-dioxane as a probable or possible carcinogen does not necessarily mean that any exposure to 1,4-dioxance will produce cancer in humans.  It just means that in abundance of caution, the public and scientists should be aware of this property for the chemical and further evaluate the risks involved.  In the Toxicology Joint Expert Report, Dr Sawyer says that “[in] virtually all public health assessments, confirmed animal carcinogens classified as probable human carcinogens or reasonably anticipated to be carcinogenic on the basis of animal evidence are always treated as probable carcinogens until conclusive human epidemiological evidence is available.  There are many such chemicals, and in the absence of hard data, all health professionals are required to use caution when producing health risk assessments.”[53]

136. From what the experts agreed, it is clear that “designation or nomenclature” reflects only a theoretical as opposed to proven risk based on animal studies which showed carcinogenic effects on animals.  Both experts agree to “uncertainty” as it applies to humans.  The “designation or nomenclature” is not based on experiments on humans.  Rather it is mainly for regulatory purpose.

(b)  “Human studies” in respect of 1,4-dioxane

137. Despite the agreement of the experts about the insufficiency of reliable published scientific or medical data or direct epidemiological evidence to show that 1,4-dioxane is carcinogenic to humans, Dr Sawyer seeks to rely on a study by Hansen (1993)(“Hansen’s Study”)[54] to support the effect of 1,4-dioxane on humans.

138. The Hansen’s Study was available because Denmark kept a database recording the number of new cases of liver cancer each year, and a comparison was made between people who were exposed to 1,4-dioxane and developed liver cancer and the incidence of liver cancer in the general population in Denmark, which showed an increase of incidence of cancer of 64%.  Dr Sawyer claims that the Hansen’s Study is “the most robust epidemiological study published to date”.

139. However, there are serious limitations about the Hansen’s Study.  Firstly, the workers in that study were exposed to a combination of both 1,4-dioxane and 1,1,1-trichloroethane. Secondly, the dose and exposure levels of the worker are not known.  This is of particular relevance to the present case as the main dispute between the parties is whether the dose in BaWang Shampoo Products was dangerous.  Thirdly, when a latency period (minimum 10 years) was incorporated in the analysis, the standardised proportionate incidence ratio was reduced to 1.15.  Fourthly, it was a retrospective survey of 19,000 cancer cases in the Danish cancer registry which observed liver cancer in workers using 1,4-dioxane from 1970 to 1984.  It was not 19,000 people exposed to 1,4-dioxane.[55]  Fifthly, some other epidemiological studies, namely studies by Buffler (1978)[56] and Thiess (1976)[57], tend to support that there is no indication of any increase in cancer in workers producing 1,4-dioxane.

140. Dr. Sawyer says that since 1,1,1-trichloroethane is not classifified as a human carcinogen, such that no confounding factors identified can explain the increase in cancer risk.  The Hansen’s Study therefore “does support human cancer from exposure to 1,4-dioxane”.[58]  But as Dr. Parent has pointed out,[59] although 1,1,1-trichloroethane is not a carcinogen, it is a liver toxin acting on the same organ as 1,4-dioxane.  So it produces liver toxicity and may account for the increase in cancer risk in that study.

141. In my judgment, what is most important is that the Hansen’s Study had been considered as inconclusive by other scientists.  As pointed out by the scientists conducting the risk assessment in the NICNAS Report[60], uncontrolled factors such as the potential for exposure to other carcinogenic chemicals, particularly 1,1,1-tricholoroethane, and the lack of quantitative exposure data for 1,4-dioxane confound any conclusions regarding a causal association with liver cancer in Hansen’s Study.[61]

142. Even USEPA, the approach of which is heavily relied upon by Dr Sawyer, accepts that the results of these “human studies” are inconclusive.  In the Toxicology Review (2013), USEPA stated:

“Studies in humans are inconclusive regarding evidence for a causal link between occupational exposure to 1,4-dioxane and increased risk for cancer; however, only two studies were available and these were limited by small cohort size and a small number of reported cancer cases (Buffer et al., 1978; Thiess et al., 1976)”[62]

143. The passage did not refer to the Hansen’s Study.  However, if USEPA considered that such study was of any relevance, they would certainly have had included such study in the Toxicology Review.

144. Hence, as agreed by the experts, there is insufficient reliable published scientific or medical data or direct epidemiological evidence to show that 1,4-dioxane is carcinogenic to humans.  The possible risk of harm to humans is just a matter of inference from animal studies, which I will deal with in some details in the latter part of this Judgment.

(c)  Regulations and safety standards in relation to the presence of of 1,4-dioxane in cosmetic products including shampoos

145. The toxicology experts have also been asked to supply information about the regulations and safety standards in relation to the presence of 1,4-dioxane in cosmetic products in countries such as the United States and EU.

146. The United States Food and Drug Administration (“USFDA”) has been monitoring levels of dioxane in cosmetics for some time but has not found it necessary to set limits, stating that the dioxane levels that they have found as part of their monitoring programme for cosmetics and personal care products “do not present a hazard to consumers”.  USFDA allows levels of dioxane in glycerides and polyglycerides of hydrogenated vegetable oils as food additives but not to exceed 10 mg/kg (10 ppm).[63] The USEPA allows for up to 4 milligrams per litre (ppm) in drinking water for a 1-day only limit.  The US National Academy of Sciences establishes a specification of 10 ppm for 1,4-dioxane in a food addictive called polysorbate.[64]

147. In EU, trace amounts of 1,4-dioxane are allowed as impurities from raw materials under the European Cosmetic Directive, 76/768/EEC.[65]

148. The Association of Southeast Asian Nations (ASEAN) Cosmetic Directive allows for trace amounts of 1,4-dioxane to be present in cosmetics provided that, inter alia, “such presence is technically unavoidable in good manufacturing practice”.[66]

149. Apart from exposures to shampoos and cosmetics, many regulatory agencies consider that human can be exposed to 1,4-dioxane in working environment at levels that are considered safe.[67]  I do not propose to set out all these limits in this Judgment, except to say that the Plaintiffs rely on these limits to support that these regulatory agencies have assumed threshold exposure for 1,4-dioxane below which there are no health effects.

150. 1,4-dioxane is included in Canada’s “Cosmetic Ingredient Hot List” and is prohibited at any level in cosmetics.[68]

151. As 1,4-dioxane is classified as a possible human carcinogen, there has been some concern about the presence of 1,4-dioxance in consumer products such as shampoos and cosmetics, and as a result a number of health and regulatory agencies had carried out studies on such subject.

152. The first report was conducted by National Industrial Chemicals Notification Assessment Scheme (“NICNAS”) in Australia which was published in June 1998.  Its conclusion was that “[the] presence of 1,4-dioxine (up to 30 ppm) as an impurity in consumer products is therefore not considered to pose a significant health risk to the general public”[69]. The report also mentioned that “[whilst] it is desirable that the level of 1,4-dioxance in consumer products be limited to 30 ppm, a level of 100 ppm in consumer products is considered toxicologically acceptable”[70].  In its final recommendation, NICNAS said:[71]

“In the protection of public health, it is recommended that levels of 1,4-dioxane in consumer products be limited to 100 ppm.”

153. The next report was published by the European Chemicals Bureau (“the EU Bureau”) with the title “European Union Risk Assessment Report – 1,4-dioxane” in 2002 (“the EU Report”).  The object of the EU Report was again to conduct a risk assessment associated with the exposure of human to 1,4-dioxane.  In respect of the exposure to 1,4-dioxane through the use of shampoo, the EU Report assumed a 1,4-dioxane concentration of 50 ppm (resulting in a total dose of 0.92 μg/kg bw/day) and 300 ppm (resulting in a total internal dose of 5.53 μg/kg bw/day) as a very worst scenario.  Even after taking into account intra- and inter-species differences, the EU Bureau was of the view that there would be no concern for consumers even if the concentration of 1,4-dioxane in shampoo is up to 300 ppm.

154. Health Canada published another report entitled “Screening Assessment for the Challenge – 1,4-Dioxane” in March 2010 (“the Health Canada Report”).  The purpose of the report was to conduct screening assessment of 1,4-dioxane to determine whether such substance presents or may present a risk to the environment or to human health.  The report would estimate the exposure of human to 1,4-dioxane through different media, including the use of shampoos and other personal care products.  Based on a number of reference materials, shampoos selling in Canada contained a range of 0.05 up to 45.5 ppm of 1,4-dioxane.  The maximum concentration of 45.5 ppm was therefore used in the study to derive at the exposure estimates.  The conclusion of Health Canada is that “1,4-dioxane is not entering the environment in quantity or concentration or under conditions that constitute or may constitute a danger in Canada to human life or health”[72].

155. One of the underlying approaches for the said 3 reports is that they adopt a threshold approach in the assessment of health risk in the case of relatively low dosage absorbed by human through the use of shampoos and other consumer products.

156. Another agency which closely monitors the health effect of 1,4-dioxane is USEPA.  It published technical fact sheets and toxicological reviews in respect of 1,4-dioxane from time to time.  For the protection of public, USEPA adopts a linear extrapolation approach as a default option in assessing the health risk possibly caused by a chemical substance in the absence of positive scientific proof.  I will further explain this methodology in the latter part of this Judgment.

157. The Mainland authority adopts the safety guideline proposed by NICNAS.  According to the announcement of the Mainland SFDA dated 16 July 2010[73], the ideal limit of 1,4-dioxane in daily consumer products is 30 ppm.  Further, as long as the content of 1,4-dioxane does not exceed 100 ppm, it “would be regarded as an acceptable content level in toxicology”.  On 30 December 2011, Mainland SFDA announced that it intended to set a residue limit of 30 ppm on 1,4-dioxane in cosmetic products.

158. There is no regulatory safety limit relating to the presence of 1,4-dioxane in consumer products in Hong Kong.

159. 1,4-dioxane exists in natural environment and so water may also contain such substance.  The World Health Organisation (“WHO”) has issued guideline for the safety limit of 1,4-dioxane in water which is currently 50μg/litre.

160. The Water Services Department in Hong Kong (“HKWSD”) adopts the WHO 2011 guideline value which is 50 μg/litre.  According to the article “Drinking water quality for the period of October 2013 and 2014” published by the Water Science Division of the HKWSD between 2013 and 2014, the minimum, maximum and average amount of 1,4-dioxane for the water supplied in Hong Kong was no more than 12.5 μg/litre.  As demonstrated in the latter part of this Judgment[74], these figures relating to drinking water would be relevant in determining whether BaWang Shampoo Products are considered safe.

161. As one can see from the above, there has been very little statutory control over the amount of 1,4-dioxane present in consumer products, health care products or shampoos. According to the newspaper reports obtained by Madam Lam in the course of her research, it seems that Qatar and a few Middle East countries are the only countries which have statutory control over the amount of 1,4-dioxane in consumer products including shampoos.  The limit is 10 ppm.  But having considered all the scientific materials presented to the court in this case, I cannot find any scientific support for this limit of 10 ppm.[75]

(d)  Different methodology to assess the health risk associated with the presence of 1,4-dioxane in consumer products

162. The central issue in respect of the defence of justification is whether chronic use of BaWang Shampoo Products would materially increase the risk of getting cancer.  Since it is not possible to conduct experiments on humans, scientists have to rely on data from animal studies to assess the health risk on humans.

163. In this regard, the parties have overwhelmed the court with voluminous scientific materials.  The court is not here to write a medical literature regarding the carcinogenicity of 1,4-dioxane.  The task of the court is to make a determination about the safety of BaWang Shampoo Products based on the existing scientific evidence presented by the parties.  It may be the case that further research in the future may contradict the finding made by the court today, but the court can only do the best it can to determine this question based on the existing scientific evidence presented before the court.

164. Scientists use animals, most commonly rats and mice, to test the carcinogenicity of a substance.  They cannot test animals with low dosage.  If someone wants to test the effect of low dosage, for statistical purpose, one may need to use thousands if not millions of animals for the experiment which is something not feasible or ethical.  Scientists therefore test the animals with huge dosage.  If the result is positive, they would plot the result on a graph with two axes: one with the number of animals affected and the other one with the different dosages administered.  Scientists would use such data to study the effect of a particular chemical on the subject animals.

165. I am given to understand that scientists will only use the statistically significant data in the graph for their studies and discard those are not.

166. At this stage, I would introduce two terms used by scientists in analysing the data which are relevant in assessing the health risk on human.

167. The first one is “lowest-observed-adverse-effect level” (“LOAEL”), that is the lowest dosage that adverse effect has been observed in respect of the experiment on the animal.  Obviously, such level cannot tell the scientists as to whether there would be adverse effect at lower levels.  Due to the sample size and other parameters of the experiment, it just means that that is the level with the lowest observed adverse effect.

168. The second one is “no-observable-adverse-effect level” (“NOAEL”).  If the scientists are able to find out a NOAEL from a particular experiment, it may support a threshold safety limit under which the chemical presents no health risk to humans.  However, due to the limited sample size and other parameters of the experiment, it may not be possible for scientists to find a NOAEL in every study or experiment.

169. In assessing the possible effect of a chemical on humans, scientists have to make some adjustments in analysing the data from animal experiments.  Firstly, scientists have to take into account the physical difference between humans and animals.  For example, the weights of humans and animals are different.  Secondly, scientists have to consider the ways in which the humans and animals absorb the chemical.  In the case of using shampoo by humans, the intake is by way of inhalation and dermal absorption which is more indirect as compared with ingestion in the case of animal experiments.

170. Dr Parent is the opinion that there is a threshold safety limit of 1,4-dioxane in consumer products under which the substance poses no health risk to human.  In the animal studies, huge dosages were given to the animals resulting in saturation of the metabolic capabilities of the animals causing damage to liver, nasal passages and other organs.  This liver damage has been demonstrated to be a precursor to the development of liver cancer.  The combination of 1,4-dioxane being non-genotoxic and only showing cancer at very high dosing levels in animals would indicate that there is a threshold dose below which no cancer occurs.  This threshold has been demonstrated in animal studies and a mode of action (“MOA”) for liver cancer development has been proposed based on the saturation of metabolic capacity and hepatotoxicity as precursors to cancer development.  In the opinion of Dr Parent, the combination of a threshold dose and the metabolic saturation resulting in hepatotoxicity strongly support the view that exposure to low levels of 1,4-dioxane in shampoos or cosmetic products results in no significant risk of cancer or any other diseases.

171. Dr Parent is not the only toxicologist that supports the threshold approach.  The scientists compiling the NICNAS Report, the EU Report and the Health Canada Report all accepted that there is a threshold dose below which there is no risk to human health.  In the earlier part of this Judgment, I have already set out their conclusions in their studies relating to the safety of 1,4-dioxane in consumer products including shampoos.

172. I first start with the approach taken by NICNAS.  One of the major considerations for the scientists in considering whether to adopt a threshold approach is whether the substance is considered to be genotoxic[76]. NICNAS adopted the “threshold approach” to drive at a “margin of safety” (“MOS”) because they considered 1,4-dioxane to be non-genotoxic:

“It is generally considered appropriate that risk characterization of non-genotoxic carcinogens … … should be treated differently to genotoxic carcinogens, in that the former group of chemicals (which include 1,4-dioxane) may be treated as requiring a threshold dose to elicit effects. Therefore an MOS approach is generally recommended for non-genotoxic carcinogens. Risk assessments for genotoxic carcinogens have traditionally been carried out using low-dose extrapolation models (e.g. linear Multistage (LMS) model), often referred to as quantitative risk assessment or mathematical modelling techniques.”[77]

173. The methodology is set out in the NICNAS Report. MOS is derived by dividing what they consider as the NOAEL (observed from animal studies) by the estimated human dose (“EHD”).

174. The EHD involved in possible exposure to human through cosmetic products (including shampoo) was then calculated.  Exposure to 1,4-dioxane (by dermal contact and inhalation) through the use of shampoo product containing 30 ppm of 1,4-dioxane was also calculated.  The total absorption would be 0.36 µg/kg/day for a 60 kg person.  Then similar calculations were made for exposures to 1,4-dioxane through body lotion/cream (0.24 µg/kg/day) and household detergents (0.72µg/kg/day).[78]

175. NICNAS then considered a “worst case scenario” which assumed a person making use of 10 consumer products a day (each containing 1,4-dioxane).  Using the highest figure of exposure by household detergents 0.72µg/kg/day as denominator, the total daily exposure would roughly be 7 µg/kg/day for a 60 kg person.[79]

176. Then, a NOAEL was chosen from animal study by Yamazaki (1994)(“Yamazaki’s Study”)[80], namely 10-40 mg/kg/day.[81]  The lower figure of 10 mg/kg/day was used.

177. By dividing the NOAEL by the EHD under the worst case scenario above, the MOS derived was about 1,500.  The significant of MOS was explained as:

“The MOS provides a measure of the likelihood that a particular adverse health effect will occur under the conditions of exposure. As the MOS increases, the risk of potential adverse effects decreases. In deciding whether the MOS is of sufficient magnitude, expert judgment is required. Such judgments are usually made on a case-by-case basis, and should take into account uncertainties arising in the risk assessment process, such as the completeness and quality of the database, the nature and severity of effect(s) and intra/inter species variability.”[82]

178. As mentioned above, NICNAS picked the NOAEL from the Yamazaki’s Study in 1994 (which another study by Kano (2009)(“the Kano’s Study”)[83] was based on).  This is explained in §9.2.5 of NICNAS Report which set out the study on liver tumours:

“Hepatocellular adenomas and carcinomas were significantly increased in rats at 0.1% and 0.5% 1,4-dioxane, respectively (Yamazaki et al., 1994). Cholangimoas were also reported in a single rat study at 1% (Kociba et al., 1974). A NOAEL (oral) for all liver tumours in rats was determined between 0.01% and 0.02% 1,4-dioxane (equivalent to 10-40mg/kg/day) in studies by Kociba et al. (1974) and Yamazaki et. Al (1994), based on the increased incidence (dose-related) of adenomas in male animals at and above 0.02% (statistically significant at 0.5%).

Hepatocellular adenomas and carcinomas were also significantly increased in mice at the lowest dose level, 0.05% (equivalent to 40-70 mg/kg/day) 1,4-dioxane. However, a clear dose-response relationship for adenomas was not evident in mice (Yamazaki et al., 1994). A NOAEL (oral) was not identified in this mouse study.”

179. After identifying the NOAEL, NICNAS derived the MOS.  The conclusion was stated under the section “12.2 Assessment of public health risks”:

“In rat studies (Section 9.2.5) the overall chronic NOAEL was 10-40 mg/kg/day (0.01-0.02% 1,4-dioxane in drinking water).

Compared with a NOAEL of 10mg/kg/day, the above worst case assumption of a systemic exposure of 7µg/kg/day (from consumer products) would represent a safety margin (MOS) of about 1500.”

180. If it is not the worst case scenario, and if one is simply using the systemic exposure for a 60-kg person who uses shampoo on daily basis, which is 0.36µg/kg/day, the MOS would be about 31,000.[84]

181. Based on the analysis, NICNAS concluded that “[the] presence of 1,4-dioxane (up to 30 ppm) as an impurity in consumer products is therefore not considered to pose a significant health risk to the general public”.[85]  In its final recommendation, NICNAS stated that “[in] the protection of public health, it is recommended that levels of 1,4-dioxane in consumer products be limited to 100 ppm.”[86]

182. Besides NICNAS, the EU Bureau is also a proponent for the threshold approach.  First, the EU Report stated that as 1,4-dioxane is considered to be a non-genotoxic carcinogen, a threshold approach is appropriate and justified.[87] It specifically noted that:

“Although there are some indication that 1,4-dioxane may be weakly genotoxic, 1,4-dioxane is considered a non-genotoxic compound based on the total weight of evidence. This is further supported by the absence of DNA-adducts at hepatoxic doses.”[88]

“The mechanism behind the organ-specific toxicity and carcinogenic effects of 1,4-dioxane has not yet been elucidated. 1,4-dioxane is considered as a non-genotoxic compound.”[89]

183. Then, it went on and discussed the MOS:

“The margin of safety (MOSs) between the inhalation exposure estimates (0.013 mg/m3… for scenario I[90]… respectively), and the NOAEL of 400 mg/m3 are all >>10,000. The MOSs between the dermal exposure estimates (0.03 mg/cm3… for scenario I… respectively) and the calculated dermal NOAEL of 20 mg/kg bw/day are far greater than 1000. When comparing the oral NOAEL with the total internal doses for scenario I (0.92 µg/kg bw/day)… the MOS-values are all >>3,000.

Taking into account intra- and inter-species differences, the non-genotoxic properties of the substances and the use of NOAELs from chronic studies, these MOSs indicate no concern from consumers by inhalation and dermal exposure...

Remark: Even when the announced reduction measures were not taken or were not effective, the total internal doses for scenario I (5.53 µg/kg bw/day)… would result in MOS-values ≥1,000 when compared to the oral NOAEL of 10 mg/kg bw/day. Hence, also in this very worst case, there would be no concern for consumers after inhalation and dermal exposure…”[91]

184. One should note that EU Bureau assumed a 1,4-dioxane concentration of 50 mg/kg in shampoo (resulting in a total internal dose of 0.92 µg/kg bw/day), and 300 mg/kg (resulting in a total internal dose of 5.53 µg/kg bw/day) as a very worst case.[92]  Despite that, EU Bureau was of the view that there would be no concern for consumers even if the concentration of 1,4-dioxane is up to 300 mg/kg, i.e. 300 ppm.

185. The threshold approach was also adopted in the Health Canada Report (March 2010).  In the report, the adoption of the threshold approach was explained at various places including the following:[93]

“Based principally on the weight of evidence-based assessments of several international agencies … … and available information … … The collective evidence indicates that 1.4-dioxane is not a mutagen and exhibits weak clastogenicity in some assays, but not others, at high exposure levels often associated with cytotoxicity. Consideration of the available information regarding genotoxicity, and conclusions of other agencies, indicate that 1,4-dioxane is not likely to be genotoxic. … … Although the mode of induction of tumours is not fully elucidated, the tumours observed are not considered to have resulted from direction interaction with genetic material. Therefore a threshold approach is used to characterize risk to human health.”

186. On exposure scenario, a number of consumer products were examined.  Pertinent to this case is the shampoo scenario.  Based on a number of reference materials, shampoos selling in Canada contained a range of 0.05 up to 45.5 ppm.  The maximum concentration of 45.5 ppm was therefore used in the study to derive at the exposure estimates.[94]

187. Exposure to female adult was used in the study as women were considered to be the most exposure group.  The total exposure to 1,4-dioxane through the use of shampoo was set out in Table 7b.[95] The MOS was also set out in a table in Appendix 5.[96]  The MOS was between 8,000 and 13,300.  According to the Health Canada Report, the MOS for intake from combined inhalation and exposure during use of consumer products (including shampoo) was considered adequately protective and the approach taken by them was conservative.[97]

188. Health Canada thus concluded that:

“The margins between upper-bound estimates of exposure from environmental media and use of consumer products, taking into consideration multiple product use scenarios and levels associated with effects in experimental animals are considered to be adequately protective to account for uncertainties in human health risk assessment for both cancer and non-cancer effects.

On the basis of the adequacy of the margins between conservative estimates of exposure to 1,4-dioxane and critical effect levels in experimental animals, it is concluded that 1,4-dioxane is not entering the environment in quantity or concentration or under conditions that constitute or may constitute a danger in Canada to human life or health.”[98]

189. Dr Parent adopts the assessments made by NICNAS, EU Bureau and Health Canada.  He is therefore of the opinion that BaWang Shampoo Products certainly constitute no real risk of harm to humans, whether the concentration of 1,4-dioxane in the shampoo is 6.4, 10 or 27 ppm.  The MOS is large enough to be safe.

190. Pausing here, these 3 reports were the most up-to-date, authoritative and relevant studies relating to the presence of 1,4-dioxane in consumer products including shampoo at the time of the publication of the Article.  According to these reports, shampoo containing 6.4, 10 or even 27 ppm is considered safe and is not posing any health risk to consumers.

191. I then turn to the methodology adopted by Dr Sawyer in assessing the health risk of 1,4-dioxane in BaWang Shampoo Products.

192. Dr Sawyer disagrees that there is a threshold dose for 1,4-dioxane in consumer products under which there is no health risk to human.  Instead, he uses the linear extrapolation approach to assess the risk, which is a default option used by USEPA in assessing the health risk to human in case of the lack of available epidemiological study data on human.

193. Put it simply, scientists would collect data from animal studies.  In the case that scientists can only obtain a LOEAL from the experiments, due to the limitations of animal studies, they would not be able to know the effect on the animal (if any) if the dosages are to be reduced much further.  In such case, scientists would use the LOEAL and to draw a linear line from the LOEAL point down to zero dosage in order to assess the risk on human.  The approach of “linear extrapolation down to zero” is therefore a mathematical construct with a view to estimate the dose-response at such low dose levels not observable from experiments.  The extrapolation all the way to zero is based on an assumption that there is a “linear relationship” between dosage and response.  In other words, it differs from the threshold approach in that it assumes that there is no “threshold dose” under which a substance would produce no effect, and hence harm.

194. In conducting the risk assessment, Dr Sawyer uses the USEPA cancer slope factor derived from the Kano’s Study.  Then it is multiplied with the lifetime average daily dose, which then gives the risk per million people.

195. The Kano’s Study was a report building on what had been observed in the Yamazaki’s Study in 1994.  In particular, Kano studied the responses at lower dosage situation using additional statistics to do such further studying.  In the Yamazaki’s Study, there was no denotation as to whether particular results were statistically significant or not.

196. Male rats and mice and female rats and mice were tested in the Kano’s Study.  According to Dr Sawyer, it is a strict rule applied internationally that scientists, in performing a toxicology study to examine different species of animals, would use the most sensitive species in making a toxicological risk assessment.  In Kano’s Study, the data relating to female mice, which was the most sensitive species, were used to derive the cancer slope factor.

197. The incidence of hepatocellular adenomas or carcinomas in the female mice in the Kano’s Study was statistically significant at all doses (namely 500 ppm, 2,000 ppm and 8,000 ppm) according to the various tests.  In the opinion of Dr Sawyer, the lowest dose of 500 ppm for female mice, the incidence of liver tumours (namely 35 out of 50 female mice) was statistically significant.

198. Due to the limitations in the experiment, only LOAEL was found in the case of the female mice, namely 66mg/kg/day (i.e. 500 ppm).  No NOAEL was found.  As mentioned above, the presence of a NOAEL may support the existence of a threshold.  But as no NOAEL was found, Dr Sawyer (and indeed USEPA[99]) used the LOAEL for female mice (i.e. 500 ppm) to extrapolate the risk in a linear manner down to zero.  That gives us the cancer slope factor.

199. Dr Sawyer performs a risk assessment based on the cancer slope factor adopted by USEPA.  It is not disputed that USEPA adopts a cancer slope factor for 1,4-dioxane of 0.1 (mg/kg-day).[100]  Risk for dermal exposure to 1,4-dioxane can then be evaluated by multiplying the cancer slope factor with the dose.  Apart from the risk associated with the dermal contact with 1,4-dioxane, there is also risk resulting from inhalation of the substance. The calculation of the dose involves a complicated formula involving both the inhalation risk and dermal risk, and NICNAS Report and EU Report adopted slightly different figures for calculating dose resulting from the use of shampoo. After performing the calculation, Dr Sawyer evaluates the risks from using shampoo containing 1,4-dioxane as follows (using the internal dosage levels adopted in the EU Report[101]):

(i)  the risk from using shampoo containing 6.44 ppm 1,4-dioxnae results in a 1.2 x 10 level, i.e. 12 new cancer diagnoses per year per population of 1 million;

(ii)  the risk from using shampoo containing 10 ppm 1,4-dioxnae results in a 1.8 x 10 level, i.e. 18 new cancer diagnoses per year per population of 1 million; and

(iii)  the risk from using shampoo containing 27 ppm 1,4-dioxnae results in a 4.9 x 10 level, i.e. 49 new cancer diagnoses per year per population of 1 million.

200. The Plaintiffs dispute the methodology of Dr Sawyer but not his calculation.

201. According to Dr Sawyer, it is generally accepted amongst toxicologists that de minimis benchmark level for cancer risk in the United States is 1 x 10^-6.  In other words, the risk of cancer is considered negligible only when posing less than one in a million lifetime risk to the most exposed individual.  Hence, based on the cancer slope factor adopted by USEPA for 1,4-dioxane, the use of shampoo containing 10 or 27 ppm of 1.4-dioxane materially increases the risk of persons in the general population having cancer, and that the risk is many times greater than the generally accepted de minimis risk level benchmark.

202. The linear extrapolation approach has been adopted by some of the regulatory authorities in setting the safety limit of 1,4-dioxane.

203. USEPA set a safety level for drinking water containing 0.7 μg of 1,4-dioxane per day and be at the de minimis cancer risk level.  According to Dr Sawyer, the amount of 1,4-dioxane that enters a person’s body by using shampoo containing 6.44 ppm or 10 ppm would be 8.3 μg/day or 13 μg/day respectively, over 10 to 20 times the abovementioned drinking water level of 0.7 μg/day.

204. USFDA has set the limit of 10 ppm for 1,4-dioxane in glycerides and polyglycerides for use in products such as dietary supplements.[102] In setting such limit, USFDA calculated the risk estimate adopting the linear extrapolation approach.

205. According to Dr. Sawyer, some countries are not even willing to adopt the risk assessment.  For example, Germany considers there to be no safe level of exposure to 1,4-dioxane and resists the use of risk assessment.  In Canada, 1,4-dioxane is prohibited at any level in cosmetics.[103]

206. Further, Dr Sawyer opines that there is a particular vulnerability of populations in Southeast Asia to liver cancer from the high prevalence of hepatitis B in this part of the world.  Dr Sawyer has not taken such factor into account in his assessment, and Mr Yu therefore submits that his assessment is an underestimate.

207. According to Dr Sawyer, the linear extrapolation approach is the “state-of-the-art” methodology in assessing health risk of possible carcinogen on human.

208. Based on the aforesaid, it is Dr Sawyer’s opinion “to reasonable toxicological certainty that the BaWang shampoo presents as a genuine public health risk”.

(e)  Is 1,4-dioxane genotoxic?

209. So which is the better methodology in assessing the cancer risk?  It is common ground that scientists may take into account, inter alia, the following factors in determining which methodology to adopt in assessing the risk:

(i)  whether the substance is genotoxic; and

(j)  whether the MOA of the substance in causing cancer is known.

210. I first deal with the issue as to whether 1,4-dioxane is genotoxic.

211. Genotoxins cause genetic damage or mutations by binding to or damaging DNA sequencing.

212. Both experts agree in Toxicology Joint Expert Report that if 1,4-dioxane is not genotoxic, there exists a probable threshold of intake by human beings below which there would be no carcinogenic risk.[104]  As 1,4-dioxane is considered not as a genotoxin, Dr Parent, as well as many other regulatory agencies including NICNAS, EU Bureau and Health Canada, adopt a threshold approach in assessing the cancer risk to humans.  On the other hand, if 1,4-dioxane is genotoxic, it may be justified to adopt the linear extrapolation approach which is the one adopted by Dr Sawyer and USEPA.  Hence, one of the main battlefield in the present case is whether 1,4-dioxane is genotoxic.

213. In his final submissions, Mr Yu has kept on asking the question as to why there is a threshold.  To answer his question, it is the scientists who say so.  Based on the various reasons given in their reports, the scientists of NICNAS, EU Bureau and Health Canada accepted that there is a threshold in the case that the chemical is not a genotoxin.  That is also the agreement between both toxicology experts.[105]

214. As to whether 1,4-dioxane is genotoxic, both experts agree in the Toxicology Joint Expert Report that:

(i)  there is no significant foundation for considering 1,4-dioxane as a genotoxin.  Most assays for genotoxicity for 1,4-dioxane have been negative, but the possibility that 1,4-dioxane is weakly genotoxic has not been eliminated;[106] and

(ii)  most of the international regulatory agencies, including EU Bureau, United States Agency for Toxic Substances and Disease Registry (“USATSDR”), Environment Canada, USEPA, United States National Toxicology Programme and the Australian authorities, do not consider 1,4-dioxane to be genotoxic.[107]

215. In his oral testimony, Dr Sawyer expresses his view that 1,4-dioxane is, more probable than not, a genotoxic substance.  He positively asserts that 1,4-dioxane is “acting genotoxic, either as a direct or indirect genotoxin”.  Furthermore, threshold approach should only be used if it can be proved for certain that 1,4-dioxane is not genotoxic.

216. I myself have great reservation about the reliability of such evidence.

217. First, Dr Sawyer’s assertion is a departure from his own opinion contained in the Toxicology Joint Expert Report.  In such Report, both experts agreed that: “[the] lack of significant genotoxicity along with cytotoxicity observed at dosing levels that induce tumours support the view that 1,4-dioxane acts via an unknown, or indirect, non-genotoxic mechanism”[108].  Now Dr Sawyer changes his stance and says that 1,4-dioxane is acting genotoxic.  Dr Sawyer explains that he had overlooked the aforesaid point in the Toxicology Joint Expert Report. Despite such purported explanation, whether 1,4-dioxane is genotoxic is an important factor in deciding which risk assessment model to adopt, and so I find it extremely strange that Dr Sawyer had missed such crucial point before signing the Toxicology Joint Expert Report.

218. Further, Dr Sawyer does not seek to resile from the other agreed points in the Toxicology Joint Expert Report, including that 1,4-dioxane as a “weakly genotoxic agent” is just a possibility that has not been eliminated.  However, those studies which indicated positive results all involved high dosing levels which produced cytotoxicity in many cases. Hence, those studies suggesting that 1,4-dioxane may be a weak genotoxic agent have been discounted in favour of more reliable tests.[109]  These points of agreement between the experts are clearly inconsistent with the latest stance of Dr Sawyer.

219. In his submission, Mr Yu argues that Dr Sawyer has not actually changed his stance on the issue.  Dr Sawyer’s view is that it is uncertain and unproven that 1,4-dioxance is genotoxic.  However, scientists are concerned with certainty whereas the court would adopt a different burden of proof.  Taking into account studies such as Kitchin & Brown’s study (1990)(“Kitchin & Brown’s Study”)[110] and Roy’s study (2005)(“Roy’s Study”)[111], it is more probable than not that 1,4-dioxane is a genotoxic substance.

220. In my judgment, that still does not resolve the inconsistency in Dr Sawyer’s evidence, and why Dr Sawyer suddenly changes his stance departing from the agreed opinions contained in the Toxicology Joint Expert Report. 

221. Further, Dr Sawyer’s opinion is contrary to the general consensus amongst scientific community.  In fact, there are abundant test results as contained in the available scientific literatures suggesting that 1,4-dioxane is non-genotoxic.  Whether one looks at the NICNAS Report or the EU Report, these studies have been fully considered and adumbrated.  A more updated summary can be found in the Health Canada Report.  Interpretation of these studies were summarized by the team of scientists in Health Canada in this way:[112]

“Genotoxicity of 1,4-dioxane has been assessed in a range of in vitro and in vivo assays. All tests for mutagenicity were negative, including those in a variety of bacterial, yeast and mammalian cells, dominant lethal mutation assay in mice and recessive lethal mutation assay in Drosophila … In addition, all mutagenicity tests with one of the metabolites of 1,4-dioxane, 1,4-dioxan-2-one, were negative … For clastogenicity investigations, including chromosomal aberration, micronuclei induction and sister chromatid exchange, the in vitro results were principally negative, with one weak positive result, and the results of the in vivo micronuclei induction assays in both CD-1 and C57BL/6 mice were mixed … 1,4-dioxane was positive in some assays, but not in others, for effects on deoxyribonuclei (DNA), such as DNA strand breaks, enhanced DNA repair processes or cell proliferation (measured as replicative DNA synthesis) and cell transformation. However, these were typically significant only at higher doses or following prolonged exposure and often in the presence of cytotoxicity. … The European Commission has concluded that the total weight of evidence indicates that 1,4-dioxane is a non-genotoxic compound (EURAR 2002). The Agency for Toxic Substances and Disease Registry (ATSDR) has indicated that “collectively, the information available suggests that 1,4-dioxane is a non-genotoxic compound, or at best, a weakly gentotoxic compound” (ATSDR 2007). The Government of Australia has concluded that “overall, the weight of evidence from in vitro and in vivo tests indicates that 1,4-dioxane is unlikely to be a mutagen” (NICNAS 1998).

Although potential mode of action of carcinogenicity of 1,4-dioxane have been examined by other agencies, these have not been fully elucidated, as data on dose-response and temporal progression with which to characterize and/or identify the key events in the processes of 1,4-dioxane-induced tumour formation and thus support any of the hypothesized carcinogenic modes of action are insufficient, inconsistent or not available. However, the collective evidence indicates that 1,4-dioxane is not genotoxic. Accordingly, although the mode of induction of tumours is not fully elucidated, the tumours observed are not considered to have resulted from direct interaction with genetic material (NICNAS; EURAR 2002; ATSDR 2007; VCCEP 2007)… Additionally, 1,4-dioxane is not a complete carcinogen, as it exhibited only tumour promotion activity, not tumour initiation activity…”

222. Then in setting out its conclusion under “Characterization of Risk to Human Health”, Health Canada said:[113]

“Based principally on the weight of evidence-based assessments of several international agencies (International Agency for Research on Cancer, European Union, US Environmental Protection Agency and US National Toxicology Programme) and available information … The collective evidence indicates that 1,4-dioxane is not a mutagen and exhibits weak clastogenicity in some assays, but not others, at high exposure levels often associated with cytotoxicity. Consideration of the available information regarding genotoxicity, and conclusions of other agencies, indicate 1,4-dioxane is not likely to be genotoxic… Although the mode of induction of tumour is not fully elucidated, the tumours observed are not considered to have resulted from direct interaction with genetic material. Therefore a threshold approach is used to characterize risk to human health.”

223. These conclusions were not simply drawn by the team of scientists compiling the Health Canada Report, they were built upon consistent “weighing of evidence” conducted by other teams of scientists engaged in the earlier NICNAS and EU Bureau’s assessments.  Apart from NICNAS and EU Bureau, USATSDR is also of the view that, collectively, the information available suggests that 1,4-dioxane is a non-genotoxic compound.[114]

224. The scientists writing the Health Canada Report had not only considered earlier conclusions or points-of-view drawn by their peers, they had also taken into account updated scientific researches on the subject before drawing to their own valued judgment.  In particular, they had taken into consideration:

(i)  Kitchin and Brown’s Study;

(ii)  Kasai’s study (2009)(“Kansai’s Study”)[115]; and

(iii)  a paper by USEPA on 1,4-dioxane (CAS RN123-91-1) in 1990 (cited May 2009).

225. Items (i) and (ii) are two of the three so-called “new studies” which Dr. Sawyer relies on to suggest that the risk assessments by NICNAS, EU and Health Canada are outdated.  The third of the so-called “new studies” is Kano’s Study.  This was not available to Health Canada at the time of their assessment done in 2010.

226. The Health Canada’s assessment (which confirmed the assessments of NICNAS and EU Bureau) supports Dr. Parent’s proposition that 1,4-dioxane is non-genotoxic or at most weakly genotoxic at very high doses.

227. In fact, this was the same conclusion drawn in the USEPA’s Toxicological Review (Aug 2010).  By this time, USEPA had extensively considered the Kano’s Study.[116]  USEPA concluded :[117]

“Overall, the available literature indicates that, 1,4-dioxane is nongenotoxic or weakly genotoxic.”

228. This same conclusion on genotoxicity of 1,4-dioxane was repeated in the most updated USEPA Toxicological Review (2013):[118]

“In the large majority of in vitro systems, 1,4-dioxane was not genotoxic. Where a positive genotoxic response was observed, it was generally observed in the presence of toxicity. Similarly, 1,4-dioxane was not genotoxic in half of the available in vivo studies. … Overall, the available literature indicates that 1,4-dioxane is nongenotoxic or weakly genotoxic.”

229. From these materials, the “new studies” relied on by Dr Sawyer do not change anyone’s overall “weighing of evidence” that 1,4-dioxane is “non-genotoxic or weakly genotoxic”.  This remains the most updated assessment even by USEPA which Dr. Sawyer relies on.  I therefore agree with Mr Pow that there is no basis for Dr Sawyer to suggest that NICNAS, EU Bureau and Health Canada were “out-dated” in their assessment of genotoxicity of 1,4-dioxane.  The original “agreement between experts” is actually consistent with the most updated description in USEPA’s Toxicological Review (2013).  Dr. Sawyer’s attempt to resile from this agreement is therefore contradicted and rendered unreliable by these literatures.

230. In re-examination, Dr. Sawyer tries to justify his “out of the norm” opinion by resorting to an argument that more weight should be attached to in vivo studies[119] (as opposed to in vitro studies[120]) because “that allows us to see the actual full gamut of metabolites and how it acts”.  Further, another possible reason for the mixed results on the question of genotoxicity is that a substance can be genotoxic in some species or some organs of some species, but not genotoxic in other animals or tested under different conditions.

231. However, such argument does not sit well with the conclusions set out in USEPA’s Toxicological Review (2013), a paper relied on by Dr Sawyer himself.  It is clear from this paper that in drawing the conclusion as set out above, all studies (whether in vitro or in vivo) had been fully taken into account.[121]  The conclusion was clearly stated as arriving from the “overall available literatures”.

232. Dr Sawyer’s view is also challenged by Dr. Parent in two aspects.  First, there are advantages in in vitro testing, because it can strip away any defence capability of the animal and subject the cells specifically to the carcinogen or proposed carcinogen or mutagen, which cannot be achieved by way of in vivo testing.  Second, in vitro assays can be “reliable information” because in conducting in vitro tests, scientists would do it “with and without a liver fraction to make sure [they] aren’t missing anything”.

233. In short, Dr. Parent is of the view that although the positive findings in a micronucleus test cannot be ignored, in view of the overwhelming evidence to the contrary, the classification of 1,4-dioxane as (at most) a weak genotoxin is appropriate.  Dr. Parent’s conclusion is consistent with the conclusion drawn by USEPA in its latest 2013 Toxicological Review.  It is also consistent with the general consensus of scientists involved in the NICNAS, EU and Health Canada’s assessments.

234. At the trial, there is some debate as to the meaning of “weakly genotoxic”.  According to Dr Parent, 1,4-dioxane is regarded as “weakly genotoxic” because it is genotoxic only at high doses and in one particular type of assay, involving DNA damage.  This means that the possibility of 1,4-dioxane exhibiting genotoxic property at high dosages (comparable to those shown in animal studies) has not been eliminated.  Yet, such situation has no relevance to the present case, as the amount of 1,4-dioxane absorbed by human in using shampoo is much smaller.

235. I accept Dr Parent’s view in this regard. Firstly, it is in line with the agreement made by the experts in the Toxicology Joint Expert Report.[122]  Secondly, as agreed by the experts, most of the health and regulatory agencies in the world do not consider 1,4-dioxane to be genotoxic.[123]  Hence I do not accept that 1,4-dioxane is, more probable than not, a genotoxin.  In the case that 1,4-dioxane is not genotoxic, as agreed by the experts, there exists a probable threshold below which there would be no carcinogenic risk to human.

(f)  MOA (Mode of action)

236. Another factor in determining which risk assessment model to adopt is the MOA of 1,4-dioxane in causing cancer.

237. Dr Parent first seemed to suggest that the MOA of 1,4-dioxane is known.  Dr Parent opines that the results of carcinogenicity found in animal studies were because the animals were dosed at levels beyond their metabolic capability.  The MOA for liver cancer development has therefore been suggested to be one based on saturation of metabolic capability and hepatotoxicity as precursors to cancer development.  In his reports, Dr Parent has emphasised many times that the doses of 1,4-dioxane administered in animal carcinogenicity studies were very high, involving “saturation metabolism” where the metabolic capability and immune systems of the animal were overwhelmed resulting in organ damage and exhalation of vapours of unmetabolized 1,4-dioxane producing lesions in the nasal passages of test animals.  This is very different from the exposure of human to 1,4-dioxane through the use of shampoos.

238. According to Dr Parent, there was a recent study by Dourson and other scientists (“Dourson’s Study”)[124] in 2014, which involved a blind re-read of mouse liver slides from the 1978 National Cancer Institute bioassay on 1,4-dioxane in drinking water, suggesting that the MOA of 1,4-dioxane is saturation of metabolic capability.

239. According to the Dourson’s Study, while 1,4-dioxane does not produce mutations, at very high doses it produces a regenerative hyperplasia which evokes endogenous mutations resulting in tumours. It therefore indicates that at very high dosing levels where the metabolic capability of the animal is exceeded, 1,4-dioxane kills liver cells resulting in a repair process that may stimulate mutations related to this toxicity. This indicates a non-genotoxic MOA with a threshold dose.[125]

240. The Defendant seeks to undermine the significance of this study because it was funded by a chemical corporation which had been actively involved in litigations in the United States in relation to groundwater clean-up that was contaminated with 1,4-dioxane and thus had a conflict of interest.  As discussed below, this challenge now becomes academic because of the admission made by Dr Parent in his cross-examination.

241. When Dr Parent is cross-examined at the trial on the results of the various animal studies, including Yamazaki, Kansai and Kano’s Studies, Dr Parent reluctantly has to admit that cancer can develop even before there is metabolic saturation to the cells or cytotoxicity (i.e. damage to the cell).  In fact, USEPA has repeatedly emphasised that studies such as the Kano’s Study suggest that cell proliferation can occur in the absence of liver cytotoxicity, and so cytotoxicity is not a required precursor event for the inducement of liver tumours.[126]  Even in the reports relied on by Dr Parent himself, including the NICNAS Report[127], the EU Report[128] and the Health Canada Report[129], the scientists accepted that the MOA of 1,4-dioxane is unknown.  A number of MOAs (including saturation metabolism and cytotoxicity) have been suggested, but none of which has been proved to be the confirmed cause.

242. I agree with Mr Yu that Dr Parent has made a mistake in suggesting that the MOA of 1,4-dioxane is one of saturation of metabolism.  One of the important assumptions in Dr Parent’s expert reports is that 1,4-dioxane requires saturated metabolic pathways to produce hepatocellular malignancies or cancers, which may not be correct in view of the results in some of the animal studies.

243. Dr Parent is now 80 years of age.  He is very keen to establish his credibility as an expert witness, sometimes to the extent of making unwarranted attack on the credential of Dr Sawyer.  However, that does not mean that the court should reject his evidence all together, in particular most of his opinion is supported by the scientists conducting the assessments in the NICNAS, EU and Health Canada Reports.

244. As I see it, both experts are keen to establish the credibility of their risk assessments, as a result they may sometimes lose their objectivity in looking at the supporting materials.  One should therefore approach their evidence with caution.  In my judgment, the best way to judge the credibility of their assessments is to test their evidence against the opinions and findings made by other scientists in the toxicology literature.  That is what I intend to do, and that is also the reason as to why I reject Dr Sawyer’s evidence that 1,4-dioxane is more probable than not a genotoxin.

245. The MOA of 1,4-dioxane is unknown[130], and that is where the scientists start to diverge in assessing the cancer risk of 1,4-dioxane.

246. The scientists conducting the assessments in the NICNAS, EU and Health Canada Reports are of the view that, though the MOA of 1,4-dioxane is not fully elucidated, a threshold approach should be adopted.  In making such decision, they have taken into account that 1,4-dioxane is not considered as a genotoxin and such substance exhibited weak clastogenicity in some assays, but not others, at high exposure levels often associated with cytotoxicity.  In working out the safe threshold, the scientists had given a sufficient wide MOS in the exercise which I have already explained in the earlier part of this Judgment.  That is also the view of Dr Parent.

247. On the other hand, Dr Sawyer is of the opinion that, since the MOA of 1,4-dioxane is unknown, it would be more appropriate to adopt the linear extrapolation approach to assess the risk.  This is also the default option adopted by USEPA.

(g)  The policy behind the default option and its limitations

248. In considering whether the linear extrapolation model truly and factually reflects the cancer risk of 1,4-dioxane, one have to understand the policy behind the adoption of such model as the default option.

249. First, the purpose of USEPA doing all the studying of scientific data, coming up with quantitative assessment of risk, is to provide building blocks of information so that other risk management agencies or risk managers can get hold of such information and make their own assessment in given circumstances. The risk assessment exercise undertaken by USEPA is only supposed to be guidelines and it has never been intended to be statements binding on regulatory authorities. The regulatory authorities have to make their own judgment and assessment. This can be seen in various passages in the USEPA’s Guidelines for Carcinogen Risk Assessment (2005) (“USEPA Guidelines”):

“Risk management applies directives in statutes, which may require consideration of potential risk or solely hazard or exposure potential, along with social, economic, technical, and other factors in decision making. Risk assessments may be used to support decisions, but in order to maintain their integrity as decision-making tools, they are not influenced by consideration of the social or economic consequences of regulatory action.”[131]

“These cancer guidelines do not suggest that all of the kinds of data covered here will need to be available or used for either assessment or decision-making. The level of detail of an assessment is a matter of Agency management discretion regarding applicable decision-making needs…”[132]

“The extent of health protection provided to the public ultimately depends upon what risk managers decide is the appropriate course of regulatory action.”[133]

250. Second, the primary goal of USEPA is to protect human health in general, and it explicitly acknowledges that it tends to take an over-cautious attitude in its assessment. Given this attitude, USEPA adopts the linear extrapolation model as its default option in risk assessment, whenever the MOA of a particular substance is unknown or not wholly ascertained scientifically.  In other words, if USEPA sees certainty from scientific studies, then they will address the calculation of risk in a particular way.  But in case of insufficient information or the existence of uncertainty, they will take a default approach or default option: “better err on the safe side”.  This policy was stated in USEPA’s Guidelines:[134]

“As an increasing understanding of carcinogenesis is becoming available, these cancer guidelines adopt a view of default options that is consistent with EPA’s mission to protect human health while adhering to the tenets of sound science. Rather than viewing default options as the starting point from which departures may be justified by new scientific information, these cancer guidelines view a critical analysis of all of the available information that is relevant to assessing the carcinogenic risk as the starting point from which a default option may be invoked if needed to address uncertainty or the absence of critical information. Preference is given to using information that has been peer reviewed, e.g. reported in peer-reviewed scientific journals. The primary goal of EPA actions is protection of human health; accordingly, as an Agency policy, risk assessment procedures, including default options that are used in the absence of scientific data to the contrary, should be health protective (U.S. EPA, 1999b).”

251. More importantly, it was explicitly stated: “In the absence of sufficiently, scientifically justifiable mode of action information, [USEPA] generally takes public health-protective, default positions regarding the interpretation of toxicologic and epidemiologic data: animal tumour findings are judged to be relevant to humans, and cancer risks are assumed to conform with low dose linearity.”[135] This shows that in an USEPA’s assessment, it is the policy to adopt the default position whenever they consider that the MOA is still an open question.[136]

252. In the case of 1,4-dioxane specifically, USEPA made it clear in the Toxicological Review (2013) that linear extrapolation approach was used because the MOA is not conclusive:

“The linear approach is recommended if the mode of action of carcinogenicity is not understood (U.S. EPA, 2005a). In the case of 1,4-dioxane, the mode of carcinogenic action for liver tumors is not conclusive. Therefore, a linear low-dose extrapolation approach was used to estimate human carcinogenic risk associated with 1,4-dioxane oral exposure.”[137]

253. As mentioned above, even in the Toxicological Review (2013), USEPA concluded (similar to NICNAS, EU Bureau and Health Canada) that 1,4-dioxane is “non-genotoxic or weakly genotoxic at high dosage”.  The above quotation makes clear that USEPA adopted the linear extrapolation approach in the case of 1,4-dioxane not because USEPA considered it to be genotoxic, but because the MOA for liver tumours is not conclusive.  It is their policy, out of abundance of caution, that the linear extrapolation approach is used to assess human carcinogenic risk.  It is then for individual risk management agency to make judgment on their assessment of the risk that 1,4-dioxane may pose in any given exposure environment.

254. Third, given this over-cautious and conservative attitude, USEPA’s estimates of risk are more likely to overstate than understate hazard or risk, which is a matter acknowledged by USEPA itself:

“Use of health protective risk assessment procedures as described in these cancer guidelines means that estimates, while uncertain, are more likely to overstate than understate hazard and/or risk.”[138]

255. Fourth, as Dr. Parent repeatedly says, USEPA’s risk calculation was based on many assumptions.  This is in fact recognised by USEPA as shown in Table 5-12 of the Toxicological Review (2013) – “Summary of uncertainty in the 1,4-dioxane cancer risk estimation”.  These uncertainties include, amongst others, low-dose extrapolation procedure (if departure from USEPA’s default approach is justified, it could decrease or increase unit risk at an unknown extent) and human relevance of mouse tumour data (if rodent tumours proved not to be relevant to humans, unit risk would not apply (cancer slope would decrease)).  It shows that even USEPA recognised that its approach was based on many assumptions, which may be right or may be wrong.  USEPA does not treat its own assessment as empirical quantification of real risk.

256. Fifth, USEPA recognises that even in the choosing of default options, there are conflicting approaches and there is no consensus on a single approach.  They take a rather open attitude and welcome alternative procedures to shed light on any uncertainty in their assessment. It was said:[139]

“When there are alternative procedures having significant biological support, the Agency encourages assessments to be performed using these alternative procedures, if feasible, in order to shed light on the uncertainties in the assessment, recognizing the Agency may decide to give greater weight to one set of procedures than another in a specific assessment or management decisions.”

257. This shows that the scientists of USEPA do not suggest themselves or arrogate themselves to represent the only authoritative approach to be taken, and if there are alternative procedures that can be provided for comparison, they encourage and welcome whoever that is doing the risk management assessment to consider the other approaches.

258. In fact, USEPA invited external peer to comment on, amongst others, the weight of evidence on characterisation.[140]  The comments received were as follows:

(i)  3 reviewers clearly commented that the weight of evidence clearly supported the conclusion that a MOA could not be identified for any of the tumour sites;

(ii)  1 reviewer commented that there is inadequate evidence to support a specific MOA with any confidence and low-dose linear extrapolation is necessary, and USEPA should not rule out a metabolite as the toxic moiety; and

(iii)  2 reviewers commented that even though the MOA for 1,4-dioxane is not clear there is substantial evidence that the MOA is non-genotoxic; one of these reviewers also suggested that a nonlinear cancer risk assessment model should be adopted.

259. The response of USEPA was as follows:[141]

“[USEPA] agrees with the reviewer not to rule out a toxic metabolite as the toxic moiety. In Section 5.5.1.2 text is included relating that there is not enough information to determine whether the parent compound, its metabolite(s), or a combination is responsible for the observed toxicities following exposure to 1,4-dioxane… In accordance with USEPA’s Guidelines… the absence evidence for genotoxicity does not invoke the use of non-linear low-dose extrapolation, nor does it define an MOA. A non-linear low-dose extrapolation can be utilised when an MOA supporting a non-linear dose response is identified. For 1,4-dioxane this is not the case; a cancer MOA for any of the tumour types observed in animal models has not been elucidated. Therefore, as concluded in the Toxicological Review, the application of a non-linear low-dose extrapolation… was not supported.

Additional text has been added to Section 5.4.3.2 to relay the fact that several reviewers recommended that the MOA data support the use of a nonlinear extrapolation approach to estimate human carcinogenic risk associated with exposure to 1,4-dioxane and that such an approach should be presented in the Toxicological Review. … … …”

260. After considering the peer reviews, USEPA made the following conclusion:[142]

“[USEPA] concluded that the available information does not establish a plausible mode of action for 1,4-dioxane and data are insufficient to establish significant biological support for a nonlinear approach. [USEPA] determined that there are no data available to inform the low-dose region of the dose response, and thus, a non-linear approach was not included.”

261. Accordingly, it is clear that the USEPA did not categorically reject the threshold approach in assessing the risk of 1,4-dioxane.  It simply said that because there is no sufficient scientific evidence to conclude the MOA for carcinogenesis of 1,4-dioxane, it is their policy to resort to the default position of linear extrapolation.  Neither did USEPA suggest that toxicologically the linear extrapolation approach is to be preferred over the threshold approach.

262. Based on the aforesaid, it is clear that USEPA’s cancer risk assessment of 1,4-dioxane is at best an estimation of its potential cancer risk to human.  It is never meant to be an exact empirical quantification of risk.  The approach of “linear extrapolation down to zero” is therefore a mathematical construct with a view to estimate the dose-response at such low dose levels not observable from experiments.  Such estimation is more likely to overstate than understate the hazard or risk of 1,4-dioxane to human.  This is due to the policy of “public health protection”.

263. On extrapolation of low-dose response beyond the experimentally observable range, USEPA recognizes that there may be situations where there is no scientific consensus favouring one individual approach over the other.  In that case, assessment results derived from alternative approaches can and should be presented to the decision-maker.

264. This can be compared with the conclusions equally drawn by NICNAS, EU Bureau and Health Canada.  These 3 health agencies equally concluded that the MOA of 1,4-dixoane “is not elucidated”.  The difference between these 3 bodies on the one hand and USEPA on the other boils down to this.  NICNAS, EU Bureau and Health Canada considered that after weighing of all evidence, as 1,4-dioxane is not considered a mutagen or genotoxin (and at most a weak genotoxin at high dosage), their judgment is that a threshold approach is justified in estimating the low-dose response.  They form this judgment regardless of the fact that the MOA of 1,4-dioxane is not fully understood.  For USEPA, whilst it equally concluded that “[overall], the available literature indicates that 1,4-dioxane is nongenotoxic or weakly genotoxic”, USEPA considered that the MOA is currently not fully understood.  Hence, based on its own policy, the default option of linear extrapolation approach is adopted for its estimation of low-dose response.

265. I agree with Mr Pow that the above difference does not involve who is in the right and who is in the wrong.  They are both estimations at best, representing scientists’ venturing into the unknown.  It demonstrates a lack of consensus amongst the scientific community.

266. There was a similar debate in the American case of Whiting v Boston Edison Company[143], in which the court was asked to consider whether to adopt a threshold model or a linear non-threshold model concerning the assessment of risk of exposure to radiation vis-à-vis acute lymphocytic leukemia.  In that case, the plaintiff was required to show that the dose level of ionizing radiation received by a worker was a proximate cause of his illness and death.  The plaintiff sought to adduce expert evidence relying on the linear non-threshold model.  The District Judge accepted the defendant’s application to exclude such expert evidence.

267. This case illustrates the limitation of such kind of expert evidence.  The risk assessment models are hypothesises only.  In the judgment, District Judge Stearns said the following:

“What must be kept in mind is that both models, although radically different in their estimation of risk, are attempts to extrapolate from the known to the unknown, something that neither model does very well with ALL (acute lymphocytic leukemia)…

No testimony could better illustrate the difference between the quests for truth in law and science remarked on by the Supreme Court:

‘Scientific conclusions are subject to perpetual revision. Law, on the other hand, must resolve disputes finally and quickly. The scientific process is advanced by broad and wide-ranging consideration of a multitude of hypotheses, for those that are incorrect will eventually be shown to be so, and that in itself is an advance. Conjectures that are probably wrong are of little use, however, in the project of reaching a quick, final and binding legal judgment often of great consequence about a particular set of events in the past. … … …’

… … The linear non-threshold model cannot be falsified, nor can it be validated. … …  It is merely an hypothesis.  In sum, it has no capacity to be of assistance to a jury in resolving the ultimate issues in this case.”

268. The court is facing the same kind of challenge here.  1,4-dioxane is an animal carcinogen.  However, whether the level of 1,4-dioxane present in BaWang Shampoo Products would materially increase the risk of getting cancer in human is very much an unknown and uncharted area.  From the perspective of health protection, scientists would try their very best to use different models, which are based on different assumptions, to assess the health risk to human.  However, that remains a theoretical or mathematical risk, not a real or factual one based on empirical data.  As the Defendant has the burden of establishing the defence of justification, it may not be able to do so based on the expert evidence available to the court.

(h)  Are the assessments made by NICNAS, EU Bureau and Health Canada out-dated?

269. In trying to establish that the linear extrapolation approach is the “state-of-the-art” methodology in assessing the risk in respect of 1,4-dioxane, Dr Sawyer claims that, in view of the findings shown in some “new studies”, the threshold approach adopted by health agencies such as NICNAS, EU Bureau and Health Canada is “out-dated”.

270. The “new studies” relied upon by Dr Sawyer are: (i) the USEPA’s reassessment of Kitchin & Brown’s Study as discussed in USEPA’s Toxicological Review (2013); (ii) Kano’s Study; and (iii) Kasai’s Study.

271. On the issue of linear extrapolation, I agree with Mr Pow that these “new studies”, when properly understood, merely demonstrate that one of the hypotheses of 1,4-dioxane’s MOA, i.e. cytotoxicity, is not supported.  The result is at best a demonstration that the MOA of 1,4-dioxane is still not elucidated or ascertained.  In other words, 1,4-dioxane acts via a presently unknown mechanism.  It is not to say that assessment of cancer risk by the threshold approach is wrong.  It merely explains why USEPA continues to adopt its default option of linear extrapolation.

272. The Defendant relies heavily on the following passage in USEPA’s Toxicological Review (2013) to claim that USEPA considered the theory of cytotoxicity as a precursor to carcinogenesis has been negated by its reassessment of Kitchin & Brown’s Study:[144]

“Female Sprague Dawley rats (three to nine per group) were given 0, 168, 840, 2,550, or 4,200 mg/kg 1,4-dioxane (99% purity) by corn oil gavage in two doses at 21 and 4 hours prior to sacrifice (Kitchen and Brown, 1990)… The results of this study demonstrated that hepatic DNA damage can occur in the absence of significant cytotoxicity. Parameters associated with tumour promotion… were also elevated, suggesting that promotion may play a role in the carcinogenesis of 1,4-dioxane.”

273. However if one is to read the USEPA’s analysis in its entirety, USEPA was only of the view that the cytotoxicity theory may not be supported by the available studies, and the MOA for liver tumours is still unknown.  USEPA considered all the relevant studies including Kasai’s Study and Kano’s Study and concluded that the MOA leading to the development of cancer is uncertain.[145] Having taken into consideration all the available literature, USEPA stated its conclusion:[146]

“Liver – The available evidence in support of any hypothesised MOA for liver tumours is not conclusive. A MOA hypothesis involving 1,4-dioxane induced cell proliferation is possible but data are not available to support this hypothesis. Pharmacokinetic data suggest that clearance pathways were saturable and target organ toxicity occurs after metabolic saturation. Liver toxicity preceded tumour formation in one study (Kociba et al., 1974) and a regenerative response to tissue injury was demonstrated by histopathology. Tumour formation has also been observed in the absence of cytotoxicity (Kano et al., 2009; JBRC, 1998). Cell proliferation and tumour promotion have been shown to occur after prolonged exposure to 1,4-dioxane (Miyagawa et al., 1999; Uno et al., 1994; Goldsworthy et al., 1991; Lundberg et al., 1987; Bull et al., 1986; Stott et al., 1981; King et al., 1973).”

274. Therefore, it is most clear that after reviewing all the available studies in relation to several hypothesised MOAs for 1,4-dioxane, even though USEPA found that the cytotoxicity theory may not be supported in the light of conflicting evidence (which includes its reassessment of the Kitchin and Brown’s Study), it merely took the view that none of the hypothesised MOA for liver tumours is conclusive. In other words, theMOA still remains unknown.  But it did not outright reject the threshold theory.

275. This is precisely the reason why, basing on their policy, USEPA assumed human relevancy and continue to use the default linear approach to estimate human carcinogenic risk.  It was statedin the Toxicological Review (2013):

“Several hypothesised MOAs for 1,4-dioxane induced tumours in laboratory animals have been discussed along with the supporting evidence for each. Some mechanistic information is available to inform the MOA of the liver and nasal tumours but no information exists to inform the MOA of the other tumours types (Kano et al., 2009; Kasai et al., 2009; JBRC, 1998; Yamazaki et al. 1994). Human relevancy is assumed unless information indicates otherwise (U.S. EPA, 2005a).”[147]

276. As mentioned above, USEPA adopted the linear extrapolation approach (the default approach) because the MOA is not conclusive:

“The U.S. EPA Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a) recommend that the method used to characterise and quantify cancer risk from a chemical is determined by what is known about the mode of action of the carcinogen and the shape of the cancer dose-response curve. The linear approach is recommended if the mode of action of carcinogenicity is not understood (U.S. EPA, 2005a). In the case of 1,4-dioxane, the mode of carcinogenic action for liver tumours is not conclusive. Therefore, a linear low-dose extrapolation approach was used to estimate human carcinogenic risk associated with 1,4-dioxane oral exposure.”[148]

277. Hence, even after reviewing the Kitchin & Brown’s Study, USEPA did not claim that the MOA for liver tumours has been found or suggested that 1,4-dioxane is a genotoxic substance. Nor did it categorically reject the cytotoxicity theory or the threshold approach.  Rather, USEPA was just saying that there is insufficient evidence in support of the hypothesised MOA, and the MOA remains unknown.

278. Kano’s Study is another study heavily relied upon by Dr. Sawyer.  It showed that cancer (hepatocellular adenoma or carcinoma) developed in the group of female mice at the level of 500 ppm (equivalent to a dosage of 66 mg/kg).[149] Dr. Sawyer maintains that this result is a matter of concern, because the dosage was “small” and no NOAEL could be identified.  It showed some genotoxicity in respect of 1,4-dioxane.

279. However, as mentioned above, USEPA has not at all relied on the Kano’s Study as a ground to change their conclusion on the genotoxicity of 1,4-dioxane.

280. Furthermore according to Dr Parent, 66 mg/kg (500 ppm) is a very high dose level. It is about 184,000 times greater than the NICNAS intake figure at 30 ppm (i.e. 0.000358 mg/kg/day)[150], and 9,200 times greater than the NICNAS worst case scenario using 10 products (up to 7 µg/kg/day)[151].  If the 500 ppm dose for female mice is converted to human equivalent dose (i.e. 10 mg/kg/day[152]), 10 mg/kg/day is still 27,778 times[153] greater than the NICNAS intake figure at 30 ppm, and 1,429 times[154] greater than the NICNAS worst case scenario.

281. As such, the fact that cancer developed at the high dosage of 500 ppm is no indication of whether cancer would develop at the lower dosing level or whether there would be a NOAEL at the lower dosing levels.  There remains a large unknown area below 500 ppm.

282. Dr. Sawyer refers to the significant finding of the Peto test[155]. However, even if the dose response trend was significant from 500 ppm to 8000 ppm, it does not mean that there was a dose response from 0 to 500 ppm.  Again, given that 500 ppm is a very high dose compared to the amount contained in BaWang Shampoo Products, the significant finding of the Peto test cannot give us any realistic guidance in assessing risk in real life situations.

283. Dr. Sawyer also says that the animal data fitted into a log-logit curve, with no points deviating from it.  However, the fact remains that it is a mathematical construct.  There is no evidence to show any effect below 500 ppm.  So, even if the points above 500 ppm fit into any model or curve, such model or curve cannot shed any light on the real risk (as opposed to theoretical risk) in association with dosing level below 500 ppm.

284. Dr. Parent is asked if there is any evidence of saturation of metabolism in respect of the statistically significant finding for the female mice at 500 ppm.  Dr. Parent’s answer is that the Kano’s Study did not look for saturation of metabolism, and he maintains that the dosing level was so high that he would expect there was saturation of metabolism or cytotoxicity.  In such circumstances, the results in the Kano’s Study may cast doubt on whether metabolic saturation or cytotoxicity is the MOA for 1,4-dioxane, but that is not to say that the threshold approach is proven wrong or out-dated.

285. I also agree with the Plaintiffs’ case that the experiment result in the Kano’s Study does not support the Defendant’s contention that 1,4-dioxance is genotoxic.  First, there is a large unknown area at the low dosing level which no studies or data could shed light on.  Even if Dr. Parent cannot put forward any hypothesis of MOA that is well supported by evidence, it does not mean that 1,4-dioxane should be regarded as genotoxic.  This will be contrary to the “weighing of evidence” conducted by all health agencies even including USEPA.  At most, it merely means that 1,4-dioxane acts through an unknown mechanism, which is indeed the most logical conclusion in the present case.  Second, USEPA, after reviewing all the available literature (including the Kano’s Study), continued to opine that there may be insufficient evidence to conclude that cytotoxicity and saturation of metabolism are precursors to cancer.  However, USEPA’s conclusion was merely that the MOA is inconclusive.  Neither the Kano’s Study nor USEPA supports the proposition that 1,4-dioxane is (or is more likely than not to be) a genotoxic substance.

286. Therefore, the statistically significant result in female mice at 500 ppm is neither here nor there.  It merely shows that there is a large unknown area below 500 ppm, but it does not mean that all the points below 500 ppm were also significant.  In other words, the fact that there is no ascertained NOAEL in the female mice study at 500 ppm does not mean that there is no NOAEL below 500 ppm.  The dosing level below 500 ppm was simply not tested.  Hence, the Kano’s Study does not support that 1,4-dioxane is genotoxic.

287. In fact, while it is true that there was no ascertained NOAEL at this level in the mice study, there was a NOAEL in the rats study:

(i)  for both male and female rats, there was no statistically significant increase in hepatocellular adenoma or carcinoma at 200 ppm compared to control;[156]

(ii)  for male rats, while there were 3 cases of hepatocellular adenoma or carcinoma found in the control group, only 4 cases were found at 200 ppm, and the increase of 1 case was not found to be statistically significant; and

(iii)  for female rats, there were 3 cases of hepatocellular adenoma reported in the control group but only 1 case found in the 200 ppm group, showing that 1,4-dioxane at 200 ppm did not cause any increase in cancer risk.

288. Based on these results, the scientists in Kano’s Study identified the NOAEL to be 11 mg/kg/day (200 ppm) for hepatocellular tumours in male rats.[157] As mentioned earlier in this Judgment, the identification of a NOAEL may support the existence of a threshold.

289. Dr Sawyer suggests that we should ignore the data relating to the rats because it is a rule for scientists to focus on the data relating to the most sensitive species which was the mice for the experiment in the Kano’s Study.  I can perfectly understand the rationale for such rule.  From the health protection perspective, it would be more desirable, out of abundance of caution, to choose the data relating to the most sensitive species to assess the health risk.  It may be the case that the response of human being may be similar to that of the most sensitive species.

290. However as suggested by Dr. Parent, there is no reason why the rats study – in particular the result at 200 ppm – should be ignored, when the mice study cannot shed light on the low-dose effects.  Dr. Parent opines that the rats result was “a good piece of data”. Further, “the problem with the mouse study is that there is no no-effect level” and “there’s only two points”.  In other words, there is a vast unknown area for the low dose scenario below the LOAEL.  It may be the case that the female mice, like the rats, may have a NOAEL below 500 ppm, and so one should not ignore the NOAEL found in the rats study.  If the NOAEL is of no relevance, I do not think that the scientists of the Kano’s study would have taken the trouble of calculating the NOAEL for rats.

291. In fact, one peculiar point about the design of the Kano’s Study is that different dosing levels were used for mice and rats.  For mice, the lowest dosage was 500 ppm, and for rats, the lowest dosage was 200 ppm. In a situation where two species received the same dosage and showed different results, there is no dispute that one should take the more sensitive species.  But this is “one situation that we do not have”.  In this regard, Dr. Parent’s opinion is that we should also study the data with the rats where there was a no-effect level.

292. In the present case, where, due to the limitations of the experiment, the female mice study could not shed light on the dosing levels lower than 500 ppm, it is justified and not inconsistent with the general rule that the most sensitive species is to be taken into account, to refer to the rats study where the experiment went down to a lower dosing level.  In this regard, I agree with Dr Parent’s opinion.

293. This was exactly what the NICNAS did.  Since no NOAEL was identified in the female mice study, they derived a NOAEL from the rats study.[158] And as a matter of fact, the scientists in Kano’s Study did not ignore the rats result, as illustrated by the fact that they discussed in detail the rats result (together with the mice result) in their study.

294. I agree with Dr Parent and Mr Pow that the attitude of the scientists in the Kano’s Study actually supports the threshold theory as a viable means of assessing cancer risk of 1,4-dioxane.  If one compares the human equivalent dose of the 200 ppm dose for male rats, i.e. 3.1 mg/kg/day, to the NICNAS intake figure at 30 ppm, i.e. 0.00036 mg/kg/day, 3.1 mg/kg/day is 8,611 times greater than 0.00036 mg/kg/day.[159] This shows how large the safety margin is between the known or identified NOAEL in animal study converted to human equivalent dose and the NICNAS intake figure at 30 ppm (which is greater than the highest alleged amount found in BaWang Shampoo Products, i.e. 27 ppm).

295. In any event, even if we were to ignore the rats result as mentioned above, the Kano’s Study cannot be used to support the Defendant’s case that 1,4-dioxane is genotoxic, because the fact that cancer developed at the lowest dosing level tested in the female mice study (i.e. 500 ppm) (the human equivalent dose of which is 27,778 times greater than the NICNAS intake figure at 30 ppm) does not mean that cancer would likely develop at the lower dosing level.  As mentioned many times, the area below 500 ppm remains a large unknown area.

296. Finally, Dr Sawyer relies on Kasai’s Study which concerned inhalation toxicity of 1,4-dioxane.  It suggested that “Plasma levels of 1,4-dioxane increased linearly with an increase in the concentrations of exposure to 400 ppm and above”.  The Defendant argues that this indicates that the rat’s body was able to handle the chemical and metabolise it even up to the level of 2,330 mg/kg, and carcinogenesis can be initiated at levels below saturation of metabolism.

297. However, this study does not support that 1,4-dioxane is genotoxic which therefore justifies the adoption of the linear extrapolation approach.   First, Dr. Parent opines that the key event is not saturation of metabolism but cytotoxicity, which kills liver cells and leads to regeneration of cells.  Second, for the reasons already stated, even if the Defendant can show evidence contradicting the theory of saturation of metabolism, it does not necessarily follow that 1,4-dioxane is genotoxic.  Even if the theory of saturation of metabolism is contradicted by any studies, at its very best, it only means that the MOA is not elucidated and that 1,4-dioxane acts through an unknown mechanism.

298. For the above reasons, I accept that the re-assessment of Kitchin and Brown’s Study, the Kano’s Study and the Kasai’s Study cannot alter the “weight of evidence” long recognised by various agencies that 1,4-dioxane is a non-genotoxin, or at most a weak genotoxin (in the sense that it is genotoxic only at high dosing level).  This is the conclusion drawn by USEPA in 2013 and is consistent with the assessments and findings made by NICNAS, EU Bureau and Health Canada.  Even taken at the highest, these studies only show that 1,4-dioxane acts via an unknown mechanism.  There is no evidence suggesting that 1,4-dioxane is or more likely to be genotoxic.

299. Before leaving the topic as to whether the NICNAS, EU and Health Reports are outdated, I have to deal with one particular criticism by Dr Sawyer about the NICNAS Report.

300. The Defendant criticises the NICNAS Report on the ground that NICNAS did not convert the animal dose to human equivalent dose. However, different approaches demand different steps to be taken in order to make sense out of that approach.  In any event, NICNAS’s explanation of MOS clearly stated that intra-species variability would be taken into consideration before making a judgment as to whether the MOS is wide enough.

301. Instead of converting the animal dose into human equivalent dose, which was part of the methodology included in USEPA’s linear extrapolation approach, NICNAS’s approach was that they first calculated a MOS (margin of safety), and then it was for the experts to judge whether the margin was sufficiently wide to satisfy them that it was safe.[160] In this particular case concerning 1,4-dioxane, the experts were satisfied that the margin of 1,500 was large enough to be considered safe.  It is therefore wrong to suggest that NICNAS failed to take into account intra-species variability.

302. In any event, for demonstration purpose, Dr. Parent has already added in a factor of inter-species differences between human and animal, the safety margin would still be large enough (8,611 for normal scenario and 443[161] for worst case scenario).  Hence, I do not accept that the approach taken by NICNAS is flawed.

(i)  WHO’s and HKWSD’s safety level of 1,4-dioxane in drinking water and the “de minimis” risk level advocated by Dr Sawyer

303. In order to test the accuracy of Dr Sawyer’s approach, one may also compare his safety level with the current WHO’s standard and the HKWSD’s standard in respect of the amount of 1,4-dioxane in drinking water.  After making such comparison, there is no serious dispute that using BaWang Shampoo Products is safer than drinking the water with the level of 1,4-dioxane at the current standard recommended by WHO.

304. The current WHO guideline value is 50 µg/litre.  The WHO is of the view that this guideline value represents the concentration of constituents in drinking water that will not result in any significant health risk to a person weighing 60 kg over a lifetime consumption of 2 litres per day for 70 years.[162]

305. Assuming a water consumption rate of 2 litres/day, the daily intake would be 100 µg.  On the other hand, the daily intake of shampoo containing 30 ppm of 1,4-dioxane of a 60 kg human (using the NICNAS figure) would just be 0.36 µg/kg/day × 60 kg = 21.6 µg (4.6 times less than the level of intake recommended by WHO).  Thus, it is clear that Bawang Shampoo Products, even if they contain up to 27 ppm of 1,4-dioxane as alleged by the Defendant, are considered safe.

306. If one applies Dr. Sawyer’s linear extrapolation approach and USEPA’s cancer slope factor, a concentration of 50 µg/litre will generate 140 cancers in a million.  This will be 140 times above the so-called “de minimis” level advocated by Dr. Sawyer.[163] Comparing this to the risk level of shampoo containing 27 ppm of 1,4-dioxane calculated by Dr. Sawyer (i.e. 49 persons in a million), BaWang shampoo is 3 times safer than the WHO-recommended water.  If Dr Sawyer’s opinion is correct, it would be more dangerous to drink WHO-recommended water than to use the Plaintiffs’ products.

307. In fact, USEPA was well aware of this WHO’s standard in 2013 when it prepared the Toxicological Review (2013).  The WHO guideline value of 50 µg/litre as of 2005 was mentioned in Appendix H of the Review under the title “Assessments by other National and International Health Agencies”.  Furthermore, WHO also put forward new edition of guideline value in 2011 which is the current guideline adopted by the HKWSD as of 2014.[164] The same value of 50 µg/litre has been maintained by WHO throughout.  Dr. Sawyer again says that the WHO’s guideline value is “outdated”.  WHO certainly does not think so.  There is no reason to suspect that WHO was not aware of the materials included in USEPA’s Toxicological Review (2013), in particular USEPA has its own safety level for drinking water calculated by reference to its de minimis cancer risk level.[165]

308. Given that the WHO’s guideline value of 50 µg/litre, which is authoritative and is widely followed around the world, is 140 times above the one in a million risk level, if USEPA found that it was erroneous or it would present any harm to the general public, it is reasonable to expect that USEPA would voice out their concern in the Toxicological Review (2013).  But 2 years have passed since USEPA published its Toxicological Review (2013), and USEPA has not issued any warning as regards the WHO’s guideline value, nor has the WHO revised its guideline value since 2014. The WHO guideline value remains to be 50 µg/litre.[166]

309. As mentioned earlier, USEPA never arrogates itself to be the only guideline that makes people safe.  It only provides guidelines and building blocks of information so that other risk management agencies can get hold of such information and make their own assessment in given circumstances.  This is exactly what happened in the present case: USEPA provided guidelines and information for other agencies to make their own risk assessment, and the WHO provided its own judgment and assessment.

310. This brings out another question: Does the “de minimis” risk level advocated by Dr Sawyer represents a real risk in practical terms or theoretical risk?  As explained by Dr. Sawyer himself, setting the “de minimis” risk level merely means that anything less than it is negligible. But non-negligible risk does not necessarily mean real risk to the public.  Dr. Sawyer mentions that liver cancer has a background rate of 7.9 per 100,000 person-years in the United States.[167]  Against this background rate, it is difficult to understand why an addition of 0.1 per 100,000 person-years (i.e. 1 per 1,000,000) would be significantly increasing the risk in real and practical terms.

311. In fact, USEPA itself did not set any particular risk limit.  They just set a range, 1 in 10,000, 1 in 100,000, and 1 in 1,000,000.[168] The range of benchmarks are for the risk managers to make their own decision.  USEPA does not say that any level above 1 in 1,000,000 is necessarily unsafe.

312. Further, there is some material in the Kano’s Study which seems to contradict Dr. Sawyer’s advocated “de minimis” benchmark.  In Kano’s Study, the writers referred to WHO’s guideline value of 50 µg/litre and the linear extrapolation of USEPA, which came up to 30 µg/litre, carrying a risk of 1 in 100,000 persons.  The writers considered that both levels are safe for drinking water.[169]

313. Dr Sawyer is asked about such contradiction.  All he can say is that the authors of the Kano’s Study are Japanese and he does not know how familiar they are with the “de minimis” risk level.  This is hardly a satisfactory explanation, given the fact the authors of the Kano’s Study are toxicologists conducting a study on the carcinogenicity of 1,4-dioxane.  In my judgment, this only shows that different scientists and health agencies around the world may have different benchmarks as to whether a substance would pose a health risk to the public in real and practical terms.

314. Besides the WHO, reference can be made to standard of water quality in Hong Kong.  According to the article “Drinking water quality for period of October 2013 and 2014” published by the Water Science Division of the HKWSD (“the HKWSD’s Article”), the minimum, maximum and average amount of 1,4-dioxane in Hong Kong drinking water was no more than 12.5 µg/litre between 2013 and 2014.  According to Dr. Sawyer’s risk calculation, it would represent a cancer risk up to 35.7 cases of cancer per year in a million.[170]

315. Although Dr. Sawyer explains that 12.5 µg/litre was the lowest point on the instrument calibration, he accepts that in any event it means that so long as the amount of 1,4-dioxane is below 12.5 µg/litre, HKWSD considers that “they are doing a good job and it’s one of the safest drinking water”.  In other words, HKWSD considers that water containing 1,4-dioxane which would, according to the linear extrapolation approach, give rise to 35.7 cases of cancer per year in a million, is safe.  In fact, the HKWSD’s Article actually claimed that “Hong Kong enjoys one of the safest water supplies in the world.”

316. Also, as HKWSD adopts the WHO’s 2011 guideline value, i.e. 50 µg/litre, it must follow that HKWSD accepts that there is no risk to human health to consume 2 litres of water a day with that concentration of 1,4-dioxane, which involves intake of 3 times more than what one can possibly obtain from the daily use of BaWang Shampoo Products (even assuming that they contain 27 ppm of 1,4-dioxane as alleged by the Defendant).

317. Yet according to the linear extrapolation model adopted by Dr Sawyer, the maximum safety limit of 1,4-dioxane in drinking water is 0.35 µg/litre, taking into account the risk benchmark of one in a million.[171]

318. Moreover, HKWSD accepts that drinking water is safe even if there are occasional deviations above the WHO’s guideline value, i.e. 50µg/litre.  The HKWSD’s Article stated that “occasional deviations above the WHO GVs do not mean that the water is unsuitable for consumption.  Large safety margins have been allowed for in the derivation of the GVs”.  It is thus the opinion and advice of the government to the public, that in deriving the guideline value of 50 µg/litre, a wide margin of safety has already been taken into account.

319. Furthermore, if Dr Sawyer’s risk assessment is correct, shampoo containing even a minute level of 1,4-dioxane would be unsafe.  Shampoo containing 6.44 ppm of 1,4-dioxane is certainly unsafe.  Even according to the Defendant’s case, it seems that some Middle East countries have adopted the most stringent control by limiting the level of 1, 4-dioxane in shampoo to less than 10 ppm, and yet this level, according to Dr Sawyer, is many times over the “safety level”.  If Dr Sawyer is correct, a lot of shampoos available in the market are unsafe for human consumption.

(iii)  Conclusion on toxicology evidence and the defence of justification

320. Hence, what is the conclusion from all the above analysis?  First, we know that 1,4-dioxane is a confirmed animal carcinogen.  Second, 1,4-dioxane is a possible human carcinogen.  However, the actual effect of the substance on humans is unknown because of the lack of empirical data.  It is so classified mainly because of the confirmed results on animals, and a lot of animal carcinogens have been eventually confirmed as human carcinogen.  Third, even if 1,4-dioxane is a confirmed human carcinogen (processed meat is now classified as human carcinogen), that does not necessarily mean that the intake of any quantity of 1,4-dioxane would materially increase the risk of cancer.  It may or may not have a safety threshold dose.  Fourth, experiments on animals were usually performed with prescriptions of relatively high doses, and based on the results of the animal studies, scientists are using different models to assess the effect of 1,4-dioxane on humans with lower doses.  These different models are based on different assumptions.  Some of these assumptions are: (i) there is a possible safety threshold if the substance is not a genotoxin; (ii) there is a linear relationship between dose and risk below LOAEL and so the risk of cancer can be extrapolated in a linear relationship below LOAEL down to zero dose.  Fifth, policy considerations would affect the choice in the use of different assumptions and models.  For example, for health protection purposes and out of abundance of caution, even if it has not been proven to be factually correct, it would be prudent to adopt a linear extrapolation approach in assessing the health risk of 1,4-dioxane if the MOA is unknown.  Sixth, most of the scientists do not regard 1,4-dioxane as a genotoxin.  Seventh, the MOA of 1,4-dioxane in causing cancer in animals is unknown.  Eighth, there are two respectable groups of scientists taking different views as to whether the level of 1,4-dioxane in BaWang Shampoo Products, whether it is 6.4, 10 or 27 ppm, would pose a health risk to consumers, with each group of scientists adopting different risk assessment models and hypothesises.  Ninth, if the more conservative group of scientists (including Dr Sawyer) are correct, even minute quantity of 1,4-dioxane in shampoo would pose a health risk to consumers, and a lot of materials for human consumption, including shampoo and drinking water, may be unsafe.  Tenth, different health agencies and risk managers may have different risk benchmarks in assessing the real or practical risk of possible harmful substance to the public.  Eleventh, there is relatively little statutory control around the world over the level of 1,4-dioxane in shampoo.

321. In establishing the defence of justification, it is the burden on the defendant to establish the truth of the alleged libel.  As it is still unclear in the science world as to whether the level of 1,4-dioxane in BaWang Shampoo Products would materially increase the risk of getting cancer or other illnesses, the Defendant has simply failed to discharge the burden of proving the truth of the Words.

322. In his submission, Mr Yu relies on McGhee v National Coal Board[172] and Lee Kin-kai v Ocean Tramping Co Ltd[173] and reminds me that there is distinction between the task of scientists and that of the court.  Whilst scientists are only concerned with certainty, the court has to make findings based on the balance of probabilities.  Hence, even if the scientists cannot prove to their satisfaction whether regular use of BaWang Shampoo Products would materially increase the risk of cancer, the court is still obliged and should make a finding, on the balance of probabilities, as to whether such Products are safe.  In this regard, since: (i) 1,4-dioxane is a confirmed animal carcinogen and a lot of animal carcinogens are eventually confirmed to be human carcinogen; and (ii) animals’ physiology, cell structure and organs are similar to those of humans, it is more probable than not that 1,4-dioxane is also harmful to human health.

323. But that is not the proper question.  The proper question should be: whether regular and continuous use of shampoo containing 10 or even 27 ppm of 1,4-dioxane is harmful to human health?  The dose is an important consideration.  So far as this particular question is concerned, the scientists cannot provide a satisfactory answer.  It may be the case that further researches in the future may shed light on this question, but the answer is unknown at present.  Any finding by the court can only be a guess.  In my judgment, it would be quite inappropriate and irresponsible for the court to rely on risk model or hypothesis to make such a finding even on the balance of probabilities.

324. Mr Yu also submits that because the Article involved a matter of public interest, that is whether BaWang Shampoo Products are safe for the public and the general consumers, the Defendant should be allowed to write on the subject.  Relying on European human right authorities such as Lingens v Austria[174] and Thorgeir Thorgeirson v Iceland[175], it would be unfair to place a burden on the Defendant to prove the impossible.  It would affect the freedom of speech and the right of the public to know.

325. In my judgment, there should be a clear distinction between the defence of justification and the defence of publication in the public interest.  Whilst the court will certainly take into account civil right considerations in determining the defence of publication in the public interest, it would not be right to relax the burden of proof in respect of the defence of justification.  Ultimately, the defence of justification only concerns the truthfulness of the alleged libel, and so the normal rule in respect of burden of proof should apply.  If the defendant cannot discharge the burden, he can still pray in aid of the defence of publication in the public interest to justify the publication of the alleged defamatory materials.  To allow civil right considerations to disturb the burden of proof for the defence of justification is, in my judgment, just confusing the two distinct defences in the law of defamation.  It would also create great confusion and uncertainty in the law.

326. Further, so far as the defence of justification is concerned, there is no unfairness to a person to bear responsibility if he disregards a controversy between scientists and proceeds to make a positive assertion about the safety of a product with no proper scientific basis.  One must bear in mind that the journalist can always pray in aid of the defence of publication in the public interest if he has satisfied the necessary requirements.  Finally, the two cases relied on by Mr Yu relate to criminal proceedings and so they are not applicable in the present case.

327. Mr Yu also seeks to rely on s 26 of the Defamation Ordinance (Cap 21) which reads as follows:

“In an action for libel or slander in respect of words containing 2 or more distinct charges against the plaintiff, a defence of justification shall not fail by reason only that the truth of every charge is not proved if the words not proved to be true do not materially injure the plaintiff’s reputation having regard to the truth of the remaining charges.”

328. According to him, it is a complete defence to a claim for libel if the defendant proves that the imputation that is being sued upon is substantially true.  In order to succeed on the defence of justification, the defendant would need only to prove that “the main charge, or gist, of the libel” is true, and does not need to justify statements or comments which do not add to the sting of the charge.  In other words, the important question is whether the core of the libel is accurate, and not to be distracted by inaccuracies around the edge.

329. In the present case, the Defendant is only able to prove that BaWang Shampoo Products did contain 1,4-dioxane which is a confirmed animal carcinogen or a possible human carcinogen. The Defendant is not able to establish, on the balance of probabilities, that regular and continuous use of BaWang Shampoo Products, with the concentration of 1,4-dioxane of 6.44, 10 or even 27 ppm, would materially increase the risk of developing cancer.  Hence, the Defendant is not able to prove that the main charge of the libel is true.  Further, I accept that the Words do carry the various meanings as pleaded by Plaintiffs in the Re-Re-Amended Statement of Claim.  Some of the defamatory stings are imputations directed against the Plaintiffs themselves as corporations.  As discussed above, the proof of the various matters included in the Lucas-Box meanings are not sufficient to meet the defamatory stings pleaded by the Plaintiffs, and hence the Defendant cannot rely on the protection under s 26 of the Defamation Ordinance.

330. For the above reasons, the defence of justification fails.

IV: DEFENCE OF HONEST COMMENT

331. I then turn to the defence of honest comment which is not the focus of the parties in this case.  According to the Plaintiffs, honest comment is the most insubstantial of the defences raised by the Defendant. 

332. The Defendant seeks to establish this defence in respect of the following 3 “comments”:

(1) "...兩星期後，化驗報告出爐，陳先生一看之下頓時又驚又怒，「霸王咁大間公司，洗頭水居然都有二噁烷!」"

(Two weeks later, the examination report was ready.  Mr Chan felt scared and angry upon reading it, “Bawang is such a big company and yet its shampoo contains 1, 4-dioxane) (“Comment (1)”);

(2) "結果，飄柔及澳雪均含微量二噁烷，故他現時選用沒有被驗出該物質的O'Naomi 洗頭水，「既然人哋都可以無，咁點解霸王咁大牌子會有呢?」”

(The results, [of the examinations] showed that Rejoice and Acene both contained traces of 1,4-dioxane.  So he now uses O’Naomi shampoo, in which no 1,4-dioxane was detected.  Other [products] do not have [1,4-dioxane], why was it found in a big brand like BaWang?)(“Comment (2)”; and

(3) "其實多個國家如歐盟、中國及台灣均禁止生產商在個人護理產品及化妝品添加二噁烷，惟若是在製造過程中令產品受到微量污染，則可以接受，但各地均沒界定何謂微量。

霸王便以此作擋箭牌，堅拒認錯，表示產品按照國內現行法律、法規及標準之要求規範生產。"

(In fact, a number of countries such as the EU, China and Taiwan all prohibited manufacturers to add 1,4-dioxane into any personal skin care products or cosmetics, but it is acceptable if the product is contaminated with traces of [1,4-dioxane] in the course of manufacturing.  However, there is no definition for “traces”.

BaWang thus used this as a shield and firmly refused to admit any fault.  They indicated that their products were produced according to the current laws, regulations and standards of China.)(Comment (3))

(i)  Undisputed legal principles

333. The legal principles for establishing a defence of honest comment are trite after Cheng v Tse Wai Chun[176] and Spiller v Joseph[177]. The ingredients of the defence are:-

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

(b)  the comment must be recognisable as comment, as distinct from an imputation of fact;

(c)  the comment must be based on facts which are true or protected by privilege;

(d)  the comment must explicitly or implicitly indicate, at least in general terms, the facts on which it is based; and

(e)  the comment must be one which could have been made by an honest person, however prejudiced he might be, and however exaggerated or obstinate his views.

(ii)  The “three comments” are statements of fact

334. The comment must be recognisable as comment as distinct from an imputation of fact.  As stated in Gatley on Libel and Slander[178], a comment is “something which is or can reasonably be inferred to be a deduction, inference, conclusion, criticism, remark, observation etc”.

335. Comment (1) is allegedly a quote from the dubious “Mr. Chan”.  It is clearly an imputation of fact as opposed to comment. The essence of the statement suggested that BaWang Shampoo Products contained 1,4-dioxane. The use of a quotation, or an exclamation mark, does not change the nature of the statement.  The reference to BaWang being a “big” company does not change the nature of the statement which is one of fact: that BaWang Shampoo Products contained 1,4-dioxane.

336. By the same token, Comment (2) is also a statement of fact, i.e. BaWang Shampoo Products contained 1,4-dioxane whilst shampoos of some other brands could be free of 1,4-dioxane.  The fact that the statement was put in a rhetorical question asked by “Mr. Chan” does not change its nature.

337. Comment (3) is a plain statement describing the attitude of the Plaintiffs.  It is also a statement of fact. The first paragraph is purely descriptive of the rules and regulations around the world concerning the presence of 1,4-dioxane in personal care products and cosmetics. The second paragraph states that BaWang uses the lack of definition of “微量”  (trace amount) in the regulations as a shield and refuses to apologize for its fault.  There is not a shred of “deduction, inference, conclusion, criticism, remark, observation” or opinion in Comment (3).  In such case, all 3 “comments” are statements or imputations of fact and so the defence of fair comment is not available to the Defendant.  This is already sufficient to dispose of the defence.

(iii)  “Comments” not based on true facts

338. Furthermore, the so-called “comments” are not based on true facts.

339. Comments (1) and (2) were made on the factual basis that shampoos of O’Naomi, Watsons & Manning did not contain 1,4-dioxane.[179]

340. However, as I will develop in the latter part of this Judgment on responsible journalism[180], these samples were all supplied by the dubious “Mr. Chan”.  Being a possible competitor who was making a story to attack the Plaintiffs, I have reason to believe that “Mr. Chan” might have tampered with the samples which he gave to SGS for testing.  In my judgment, it is dangerous for the court to rely on the test results of the samples submitted by “Mr Chan”, and so there is no factual basis for Comments (1) and (2) to base on.

341. Comment (3) imputes that the Plaintiffs had done something wrong: they were hiding behind regulatory lacuna or loopholes and refused to acknowledge or apologise for the wrong they committed.  The existence of “fault” on the part of the Plaintiffs was the essential factual basis of the “comment”, which would in turn depend on the fact whether BaWang Shampoo Products do pose a risk of harm to the health of consumers. As already discussed under the defence of justification, the Defendant has failed to discharge the burden of proving such fundamental fact.  As the fact upon which the “comment” was based was untrue, the defence is not available for such “comment”.  The Plaintiffs had committed no fault or wrong, and there was nothing from which the Plaintiffs sought to hide behind.

342. Further, the Defendant had omitted the reference to the safety limit proposed by the scientists compiling the NICNAS Report, which was the most relevant scientific report about the safety level of 1,4-dioxane in shampoo.  That was the safety standard adopted by the Australian health and regulatory authorities, Mainland SFDA, the Hong Kong Customs and Taiwanese health authorities and relied on by the Plaintiffs.  Yet the Defendant claimed in Comment (3) that there was no definition about “traces” and proceeded to allege that the Plaintiffs were hiding from their wrong.  Such “comments” were therefore made not based on true facts.

(iv)  The first two “comments” do not meet the defamatory stings in the Article

343. Further, I must point out that the first two “comments” do not meet the defamatory stings in the Article. Comments (1) and (2) are taken from “Mr. Chan” and are confined to the suggestion that unlike other brands of shampoo, such a big shampoo brand like BaWang contained 1,4-dioxane.  In isolation, these two comments perhaps convey a message that BaWang Shampoo Products were inferior in quality to that of 3 other brands.  This is however not the Plaintiffs’ complaint.  The complaints are those stings discussed in the earlier part of this Judgment conveyed by the defamatory meanings of the Words.  As establishing these two comments as honest comments does not meet any of such stings of the libel, the defence of honest comment would not be available at least in respect of the first two comments.

(v)  Malice in publishing the “comments”

344. Based on my findings above, it is not necessary for me to deal with the issue as to whether the “comments” were actuated by malice.  In any event, I will deal with malice in the latter parts of this Judgment both on the issues of responsible journalism and malicious falsehood. Although I find that the Defendant’s reporting fell below the standard of responsible journalism, I have some doubt as to whether the “comments”, if properly established, were tainted with malice.  Hence, I do not reject the defence on such ground.

345. For the above reasons, the defence of honest comment also fails.

V:  DEFENCE OF PUBLICATION IN THE PUBLIC INTEREST

346. The defence of publication in the public interest is another main battlefield between the parties.  According to the Defendant, the Article involved a piece of investigative journalism relating to a matter of public interest.  In order to protect the freedom of the speech and the press and the right of the public to know, the Defendant should be allowed to rely on such defence to defeat the Plaintiffs’ claim even if it fails to justify the truth of the Words.

347. There has been considerable development in the law about this particular defence in recent years.  In a way, the law has been “liberalised”[181] or “constitutionalised”[182] from a human right perspective.  As a result, the notion of “responsible journalism” has been developed.

(i)  The development of the principle of “responsible journalism”

348. In Jigme Tsewang Athoup v Brightec Ltd[183], which is a case concerning the introduction of the defence of reportage in Hong Kong, I have traced the development of the notion of responsible journalism.

349. It first started with the principles laid down in the landmark decision of the House of Lords in Reynolds v Times Newspaper Ltd[184]. In that case, the House of Lords “established a new variant of qualified privilege in which less emphasis was placed on the traditional, reciprocal duty and interest test, and more on the question of whether the publication was on a matter of public interest and whether it was the product of responsible journalism (with the issue of malice being subsumed within this latter element)”[185].  According to the Law Lords, the duty and interest test can be satisfied if the public is entitled to know the particular information being published subject to 2 essential prerequisites.  The first is that the article as a whole must be in the public interest.  The second prerequisite is responsible journalism.  Whether the article is of value to the public depends upon its quality as well as its subject matter and the value of the article to the public must be tested against a standard of responsible journalism.  Responsible journalism is the point at which a fair balance can be held between freedom of expression on matters of public concern and the reputation of the individual harmed by that disclosure.[186]

350. The court, not the editor, must decide whether the particular material is privileged because of its value to the public but due weight must be given to editorial judgment.[187]

351. In Reynolds, Lord Nicholls set out a non-exhaustive list of circumstances which would be relevant to the issue of whether the standard of responsible journalism had been met in a given case. According to Lord Nicholls, the list is not exhaustive, and the weight to be given to these and any other relevant factors will vary from case to case.[188]

352. The 10 listed factors are:

(i)  The seriousness of the allegation.  The more serious the charge, the more the public is misinformed and the individual harmed, if the allegation is not true.

(ii)  The nature of the information, and the extent to which the subject matter is a matter of public concern.

(iii)  The source of the information.  Some informants have no direct knowledge of the events.  Some have their own axes to grind, or are being paid for their stories.

(iv)  The steps taken to verify the information.

(v)  The status of the information.  The allegation may have already been the subject of an investigation which commands respect.

(vi)  The urgency of the matter.  News is often a perishable commodity.

(vii)  Whether comment was sought from the plaintiff. He may have information others do not possess or have not disclosed.  An approach to the plaintiff will not always be necessary.

(viii)  Whether the article contained the gist of the plaintiff’s side of the story.

(ix)  The tone of the article.  A newspaper can raise queries or call for an investigation.  It need not adopt allegations as statement of fact.

(x)  The circumstances of the publication, including the timing.

353. In Jameel (Mohammed) v Wall Street Journal Europe SPRL[189], the House of Lords emphasised that the “Nicholls factors” must be approached in a practical and flexible manner with due deference to editorial discretion.  The list certainly does not set up a series of hurdles to be negotiated by a publisher before he can successfully rely on qualified privilege as a defence.  Baroness Hale also said:[190]

“It should by now be entirely clear that the Reynolds defence is a ‘different jurisprudential creature’ from the law of privilege, although it is a natural development of that law. It springs from the general obligation of the press, media and other publishers to communicate important information upon matters of general public interest and the general right of the public to receive such information. It is not helpful to analyse the particular case in terms of a specific duty and a specific right to know. That can, as experience since Reynolds has shown, very easily lead to a narrow and rigid approach which defeats its object. In truth, it is a defence of publication in the public interest.”

(ii)  Human rights jurisprudence

354. As mentioned by Ward LJ in Roberts v Gable[191]and the learned authors in Gatley[192], the European Convention for the Protection of Human Rights and Fundamental Freedoms (“the European Convention”), which was incorporated into English law by the Human Rights Act 1998, played a major formative role in the decisions of both the Court of Appeal and the House of Lords in Reynolds v Times Newspaper Ltd[193].

355. Article 10 of the European Convention provides that everyone has the right to freedom of expression.  However, the exercise of such freedom may be subject to, inter alia, such restrictions as are prescribed by the law and are necessary in a democratic society for the protection of the reputations or rights of others.

356. The European Court of Human Rights has repeatedly emphasised that a liberal approach should be adopted in cases concerning the freedom of speech.  In The Sunday Times v The United Kingdom[194], the Court said:

“The court is faced not with a choice between two conflicting principles, but with a principle of freedom of expression that is subject to a number of exceptions which must be narrowly interpreted.” [195]

357. The Court also pointed out that:

“It is not sufficient that the interference involved belongs to that class of the exceptions listed in article 10(2) which has been invoked; neither is it sufficient that the interference was imposed because its subject-matter fell within a particular category or was caught by a legal rule formulated in general or absolute terms: the Court has to be satisfied that the interference was necessary having regard to the facts and circumstances prevailing in the specific case before it.”[196]

358. A vital and distinct aspect of the right is the freedom to receive information and ideas.[197]  In The Observer and the Guardian v United Kingdom[198], the European Court of Human Rights said:

“Not only does the press have the task of imparting such information and ideas: the public also has a right to receive them. Were it otherwise, the press would be unable to play its vital role of `public watchdog’.”[199]

359. The English courts have also emphasised the importance of the freedom of expression.  In Loutchansky v Times Newspapers Ltd. & others (No.4 and 5)[200], the English Court of Appeal quoted with approval the following passage in Al-Fagih v HH Saudi Research and Marketing (HK) Ltd[201]:

“This approach [referring to Reynolds and the ten factors identified by Lord Nicholls] reflects the jurisprudence if the European Court of Human Rights under article 10 of the Convention for the Protection of Human Rights and Fundamental Freedoms and is designed to enable a proper balance to be struck between on the one hand the cardinal importance of freedom of expression by the media on all matters of public concern, and on the other hand the right of an individual to his good reputation. Neither right is absolute but the former, particularly in the field of political discussion, is of a higher order, a constitutional right of vital importance to the proper functioning of a democratic society. That is why ‘any curtailment of freedom of expression must be convincingly established by a compelling countervailing consideration, and the means employed must be proportionate to the end sought to be achieved’ (per Lord Nicholls at 200F), and why ‘Any lingering doubts [as to how the balance should be struck] should be resolved in favour of publication’ (per Lord Nicholls, at p 205F).”[202]

360. The law of defamation is therefore concerned with striking a balance between the freedom of expression and information on the one hand and the protection of the honour and reputation of individuals on the other.[203] Although the former is a constitutional right of cardinal importance, neither right is absolute.

361. Hong Kong has similar human rights legislations.

362. Article 27 of the Basic Law provides that Hong Kong residents shall have, inter alia, freedom of speech, of the press and of publication.

363. Article 16 of the Hong Kong Bill of Rights Ordinance (Cap 383)(“HKBOR Ordinance”), which is equivalent to Article 19 of the International Covenant on Civil and Political Rights, provides that, inter alia, everyone shall have the right to freedom of expression, including the freedom to seek, receive and impart information, provided that the exercise of such right may be subject to certain restrictions as are provided by law and are necessary for respect of the rights or reputations of others.

364. In Cheng & Anr v Tse Wai Chun[204], Li CJ said the following:

“The freedom of speech (or the freedom of expression) is a freedom that is essential to Hong Kong’s civil society. It is constitutionally guaranteed by the Basic Law (art. 27). The right of fair comment is a most important element in the freedom of speech.

In a society which greatly values the freedom of speech and safeguards it by constitutional guarantee, it is right that the courts, when considering and developing the common law, should not adopt a narrow approach to the defence of fair comment. See Eastern Express Publisher Ltd & Anr v Mo Man Ching & Anr (1999) 2 HKCFAR 264 at p. 278.  The courts should adopt a generous approach so that the right of fair comment on matters of public interest is maintained in its full vigour.”

365. In Jigme Tsewang Athoup v Brightec Ltd[205], I made the remark that although the dicta of Li CJ were made in the context of the defence of fair comment, the same rationale and liberal approach should apply when considering the defence of Reynolds privilege or publication in the public interest.

366. The law of defamation in Hong Kong must therefore be understood and applied in a way consistent with the Basic Law and the HKBOR Ordinance.  In determining whether the press is liable for a libel, the court has to consider whether it is necessary for respect of the right or reputation of other to prohibit the publication of the alleged libel to the public which may have the right to receive them.

367. The word “necessary” should be given its ordinary meaning. The requirement also involves the application of a proportionality test.  The restriction should be rationally connected with one or more of the legitimate purposes and the means used to impair the right must be no more than is necessary to accomplish the legitimate purpose in question.[206]

368. However, the “subject matter” of the publication may be of public interest and concern is not a warrant for the inclusion of all sorts of defamatory statements, otherwise it would just be a blanket licence to the journalists to publish whatever they like so long as the subject matter of the publication somehow relates to matter of public interest, and they do not need to exercise any care in their reporting.  This cannot be in the public good.

369. In Jameel (Mohammed) v. Wall Street Journal Europe SPRL, Lord Hoffmann said:[207]

“If the article as a whole concerned a matter of public interest, the next question is whether the inclusion of the defamatory statement was justifiable. The fact that the material was of public interest does not allow the newspaper to drag in damaging allegations which serve no public purpose. They must be part of the story. And the more serious the allegation, the more important it is that it should make a real contribution to the public interest element in the article.”

370. Further, the learned authors of Gatley on Libel and Slander noted that “[a] defendant who makes a specific charge… will not be allowed to dress up his defence with a plea of Reynolds privilege simply because the background to the charge involves a matter of public interest.”[208] A passage in Miller v Associated Newspaper Ltd was cited:[209]

“there is a danger that any plea citing a few generalities about the duty of the media to be a public watchdog will be allowed to pass muster and thus to prolong and complicate unnecessarily a significant number of libel actions in which qualified privilege has no legitimate role to play.”

371. In considering whether it is part of the story to include the defamatory statements, Lord Hoffman said the following in Jameel:[210]

“… … But whereas the question of whether the story as a whole was a matter of public interest must be decided by the judge without regard to what the editor’s view may have been, the question of whether the defamatory statement should have been included is often a matter of how the story should have been presented. And in that question, allowance must be made for editorial judgment. If the article as a whole is in the public interest, opinions may reasonably differ over which details are needed to convey the general message. The fact that the judge, with the advantage of leisure and hindsight, might have made a different editorial decision should not destroy the defence, That would make the publication of articles which are, ex hypothesi, in the public interest, too risky and would discourage investigative reporting.”

372. The introduction of the defence of publication in the public interest was, therefore, a response by the courts to apply the law of defamation in accordance with the human rights legislations and jurisprudence.  The yardstick for the defence is the notion of responsible journalism.  Although the court should apply such notion liberally and generously in accordance with the human right jurisprudence, such right is not absolute and the court has to strike a balance between the freedom of expression and the right to protection of the reputation of individuals.  This is not an easy exercise.

373. Before I deal with the principle of responsible journalism, I have to address one more human right argument advanced by Mr Yu. According to him, freedom of expression needs to be balanced against the freedom of private life or right to the private life or right to the protection of reputation as provided for in, for example, Article 14 of the HKBOR Ordinance.  In the present case, the right to reputation is not engaged because there is a distinction between an individual’s right to reputation and a company’s reputation, the latter is not protected under human right legislations.

374. I disagree.  If Mr Yu’s argument is correct, there would be two different standards to apply, one for a claimant who is a natural person and another one for a claimant which is a corporation, in the balancing exercise involving the right to freedom of expression and the right to protection as to reputation.  This simply cannot be right.  Furthermore, when Article 10 of the European Convention, Article 27 of the Basic Law and Article 16 of the HKBOR Ordinance refer to the restriction of the freedom of expression, they describe the restrictions “as are provided by law and are necessary for respect of the rights or reputations of others”, and “others” should, in my judgment, include both natural persons and corporations.  A lot of the rights under human right legislations are equally enjoyed by corporations, for example, equality before the courts and right to fair and public hearing, and so I do not see a distinction between the claimant being a natural person or a corporation.

(iii)  The standard of responsible journalism

375. How should the court then set the standard of responsible journalism? 

376. Firstly, the standard of responsible journalism is an objective one to be decided by the court and not the journalist, but due weight must be given to editorial judgment.

377. Second, the standard should neither be too low nor too high.  In Loutchansky v Times Newspaper Ltd (Nos 4 and 5)[211], it was held that while the free flow of information should not be compromised and matters of public interest should not be prevented from coming to light so that the standard for responsible journalism should not be set too high, the standard should also not be too low as that would encourage the proliferation of defamatory materials which cannot be in the public good.  The English Court of Appeal observed:[212]

“(ii) Setting the standard of journalistic responsibility too low would inevitably encourage too great a readiness to publish defamatory matter. Journalists should be rigorous, not lax, in their approach. It is in the interests of the public as well as the defamed individual that, wherever possible, truths and not untruths should be told. This in the interests of the media too: once untruths can be published with impunity, the public will cease to believe any communications, true or false.

(iii)  Setting the standard too high, however, would be no less damaging to society.  This would deter newspapers from discharging their proper function of keeping the public informed.  When determining in respect of any given article whether or not it should attract qualified privilege, the court must bear in mind the likely impact of its ruling not only upon the case in hand but also upon the media’s practices generally.  Qualified privilege ordinarily falls to be judged as a preliminary issue and before, therefore, the truth or falsity of the communication is established.  The question to be posed is accordingly whether it was in the public interest to publish the article, true or false, rather than whether it was in the public interest to publish the untruth.  Even, moreover, when the untruth of the article is established (or when, as here, it is not formally disputed), it is important to remember that the defence of qualified privilege tolerates factual inaccuracy for two purposes: first so as not to deter the publication sued upon (which might have been true); and secondly so as not to deter future publications of truthful information.”

378. In the past, the courts have always cautioned about the danger of setting the standard too high.  There is a concern that a high standard would discourage the press from discharging its function as a watchdog in a free society, and as a result the general public would suffer because the right to know would be curtailed.

379. However, little has been said about setting the standard too low.  In my judgment, the public also has an expectation for the court not to set the standard too low, otherwise it would encourage the press not to exercise any care in making their reports. With the result that the press would lose its objectivity, with selective reporting of materials which are favourable to make up a story.  Once the press loses its credibility, the public will cease to believe in any of their reports, whether true or false. This is also a legitimate concern of the public, and the responsible journalist should therefore try to ensure that what is published is accurate and fit for publication.

380. As Lord Hobhouse said in Reynolds:[213]

“The liberty to communicate (and receive) information has a similar place in a free society but it is important always to remember that it is the communication of information not misinformation which is the subject of this liberty. There is no human right to disseminate information that is not true. No public interest is served by publishing or communicating misinformation. The working of a democratic society depends on the members of that society, being informed not misinformed. Misleading people and the purveying as facts statements which are not true is destructive of the democratic society and should form no part of such a society. There is no duty to publish what is not true: there is no interest in being misinformed. These are general propositions going far beyond the mere protection of reputations.”

381. Lord Bingham also noted in Jameel that:[214]

“The rationale of this test is, as I understand, that there is no duty to publish and the public have no interest to read material which the publisher has not taken reasonable steps to verify. As Lord Hobhouse observed with characteristic pungency, at p 238, "No public interest is served by publishing or communicating misinformation." But the publisher is protected if he has taken such steps as a responsible journalist would take to try and ensure that what is published is accurate and fit for publication.”

382. No public interest is served by communicating misinformation, and the dicta of Lord Hobhouse highlight the harm that can be caused by misinformation in a free society.  In fact, we are now living in the internet stage and watchdogs are everywhere.  We are flooded with information and misinformation every day, and on many occasions the public has no way to ascertain the truth of the information fed to them.  People act on those information, whether true or false, and one should not undermine the harm and damage that can be caused by misinformation.

383. Information from the press is different from information available casually in the internet.  Because of its function as a watchdog in the society, the public would expect the press to carry out some basic duties in verifying the truth of the information published by them.  This is especially the case when the public is presented with a piece of investigative journalism dealing with scientific issues.  The public would expect that the press has carried out the relevant investigative and research works, and the conclusion published in the article is one substantiated with reasonable materials.  Due to these assumption and expectation, such kind of report would carry much weight in the eyes of the public.  It is therefore appropriate to ask the journalists to do their works responsibly, especially when they knew full well that the report would seriously affect the reputation and business of others.

384. Mr Pow has referred me to the Journalists’ Code of Professional Ethnics published by the Hong Kong Press Council (“the Code of Ethics”) which requires journalists to, inter alia:

“(i) Journalists should handle news information with an attitude of seeking truth, fairness, objectivity, impartiality and comprehensiveness. Journalists should strive to ensure accuracy of their reports. They should not mislead the public be quoting out of context, distorting facts or twisting original meaning.

(i) Whenever proven inaccuracy, misleading facts or figures, or distortion with earliest opportunity to reply, and corrections should be made promptly.

.. … …”

385. However, the Code of Ethics is not law and so it should not be construed as such.  If one were to apply the Code of Ethics strictly, it would certainly discourage journalists to make any reports on matters of public concern which would not be in public good.  On the other hand, I would imagine every journalist would know about their duty to be objective and not to mislead the public by telling the half truth.  When the press is exercising its power and influence as now commonly referred to as the “fourth estate (or fourth power)” in every civilised society, there must be corresponding duty and standard of responsible journalism that they have to observe. Journalists are also professionals.  If they perform their works irresponsibly which seriously affect the reputations of others, they cannot always put forward freedom of speech as an excuse.

386. It is also true that the law recognizes that journalists, in the exercise of their freedom of expression, should be permitted a degree of exaggeration. In Turcu v News Group Newspapers Ltd[215], Eady J commented, in respect of the article complained of, that there “may be a good deal of sloppiness and inaccuracy in what was published”[216], but nevertheless went on to say:

“108… journalists, in the exercise of their rights to freedom of expression, need to be permitted a degree of exaggeration even in the context of factual assertions (not only when making comments or voicing their opinions)…

… … …

111.  In deciding whether any given libel is substantially true, the court will have well in mind the requirement to allow for exaggeration, at the margins, and have regard in that context also to proportionality. In other words, one needs to consider whether the sting of a libel has been established having regard to its overall gravity and the relative significance of any elements of inaccuracy or exaggeration. Provided these criteria are applied, and the defence would otherwise succeed, it is no part of the court's function to penalise a defendant for sloppy journalism — still less for tastelessness of style…”

387. Similarly, the European Court of Human Rights stated in Bladet Tromsø and Stensaaas v Norway:[217]

“One factor of particular importance for the Court’s determination in the present case is the essential function the press fulfils in a democratic society. Although the press must not overstep certain bounds, in particular in respect of the reputation and rights of others and the need to prevent the disclosure of confidential information, its duty is nevertheless to impart – in a manner consistent with its obligations and responsibilities – information and ideas on all matters of public interest. In addition, the Court is mindful of the fact that journalistic freedom also covers possible recourse to a degree of exaggeration, or even provocation. In cases such as the present one, the national margin of appreciation is circumscribed by the interest of democratic society in enabling the press to exercise its vital role of “public watchdog” in imparting information of serious public concern.”

388. On the other hand, the courts have also emphasised that exaggeration is not without bound.  In the present case, the subject matter of the Article relates to the safety of a product which involves considerable scientific researches on the effect of 1,4-dioxane on animals and humans.  The issue here is whether the Defendant should be allowed to exaggerate the health risk of BaWang Shampoo Products, which is very much a scientific issue, without presenting the contrary scientific materials to the readers. As a result, the readers only got the half truth and received the message that the scientific community had conclusively come to the conclusion that BaWang Shampoo Products were unsafe.  In particular, the Defendant knew full well that the Article would have a significant impact on the reputation and business of the Plaintiffs.

(iv)  Events leading to the publication of the Article

389. Did the Defendant meet the standard of responsible journalism?

390. In his submission, Mr Pow has conducted a microscopic analysis of every step taken by Madam Lam and the editorial staff of the Magazine leading to the publication of the Article.  It is the Plaintiffs’ case that Madam Lam and the staff had lost their objectively since the commencement of the investigation, and as a result the final Article was a piece of biased and defamatory report targeting the Plaintiffs.

391. It is easy to make these criticisms with the benefit of hindsight.  But despite the generous allowance given to journalists as to how to present their reports, a lot of the complaints lodged by the Plaintiffs do have merit.  As I see it, the main “mistakes” committed by the Defendant are that the Article made a positive and serious assertion that regular use of BaWang Shampoo Products was harmful to human health without adequately telling the readers that: (i) there were responsible scientific researches showing the contrary; and (ii) there were responsible regulatory agencies in other scientifically advanced countries taking the view that the level of 1,4-dioxane found in BaWang Shampoo Products was safe.  The Defendant had failed to verify the source of the scientific materials behind the so-called “10-ppm safety limit”.  The Article had exaggerated, in a sensational manner, the health risk associated with the use of the Plaintiffs’ shampoos.  Further, it served no public interest for the Defendant to include a lot of other unnecessary defamatory accusations against the Plaintiffs, for example, the Plaintiffs were irresponsible, incompetent and greedy commercial organisations which lacked commercial morality, and they used inferior raw materials to produce BaWang Shampoo Products with mercenary motive to save costs.

392. The Defendant published the Article knowing full well that they were making one of the most serious allegations against the Plaintiffs and their products.  The Article was a featured article in the Magazine intended to create an impact on the Plaintiffs and their consumers, which is best illustrated by the sudden drop in the share price of the 1^st Plaintiff within a few hours after the publication of the Article.  In such circumstances, the Defendant should have borne a greater responsibility and exercised more care in ensuring that the contents of the Article were accurate and not biased.  Unfortunately, the Defendant had failed to meet such standard and I do not accept that the Article was a product of responsible journalism. 

393. The Defendant had also committed other “errors”, such as failing to verify the credibility of the complainant and the test results provided by him and acting oppressively against the Plaintiffs when it demanded for a response from them.  Although each of these individual “errors” may not by itself be sufficient to conclude that the Defendant’s reporting fell below the standard of responsible journalism, the court will certainly take all these factors into account in determining whether the standard has been met in the present case.  Unfortunately, the Defendant has failed to meet such standard.

394. To make good these observations, I need to deal with the evidence leading to the final publication of the Article.

(a)  How did the investigation begin?

395. It began with a written complaint by one “Mr Chan” to the Complaint Section of the Magazine on 16 April 2010.  The complaint record showed that “Mr Chan” had taken a bottle of BaWang shampoo to SGS for testing.  “Mr Chan” suggested that the tested content of 1,4-dioxane exceeded “the American standard by 17 ppm”.  “Mr Chan” also called 1,4-dioxane a carcinogen and stressed his worry that Bawang Shampoo Products were harmful to health.

396. This complaint was followed by “Mr. Chan’s” email dated 16 April 2010 again stressing his worry about BaWang Shampoo Products posing a health risk.  On 17 April 2010, by another email, “Mr. Chan” attached a copy of the SGS report.  On 23 April 2010, “Mr. Chan” approached another staff of the Magazine and sent her the SGS report again.  Then on 29 April 2010, “Mr. Chan” sent another email threatening to go to Oriental News and other media if he did “not hear any news or action from NEXT Media”.  It is very clear that “Mr. Chan” was a particularly keen complainant.

397. The matter eventually reached the attention of Madam Lam on about 22 June 2010.  Madam Lam admits that she had suspicion over “Mr Chan” at that time because: (i) he was particularly keen for the Magazine to report on his complaint; (ii) he spent money in sending the shampoos for professional testing of 1,4-dioxane; and (iii) he seemed to have knowledge as to the permissible limit of 1,4-dioxane.

398. At the trial, Madam Lam admits that she did not feel safe to rely on the SGS report sent by “Mr Chan” under his emails.

(b)  How was the investigation conducted?

399. According to Madam Lam, prior to meeting and interviewing “Mr Chan” on 29 June 2010, she had undertaken some research on 1,4-dioxane.  The materials obtained by her included, inter alia: (i) an article from an American campaign group called “Campaign for Safe Cosmetics” concerning failures of listing 1,4-dioxane in product labels; (ii) an article on 1,4-dioxane from “Skin Deep: Cosmetic Safety Review”; (iii) an article of Oriental Daily dated 17 March 2010; (iv) an article of Sing Tao Daily dated 22 March 2009; and (v) a public statement in the website of the Cosmetic & Perfumery Association of Hong Kong Ltd (“the HK Cosmetic Association”) informing consumers that they need not worry about the presence of small quantity of 1,4-dioxane in cosmetics.

400. The main focus of the research was to ascertain the harmful effect of 1,4-dioxane on humans (if any) and the safety level of such substance in shampoo (if any).  From these materials, Madam Lam should have known the following, inter alia:

(i)  There was concern about the presence of 1,4-dioxane in consumer products in the United States but there was no express reference to a safety level;

(ii)  A variety of international agencies classified 1,4-dioxane differently in relation to “carcinogenicity” of 1,4-dioxane. Some clearly considered that there was insufficient evidence of carcinogenicity to humans.  Madam Lam admits in cross-examination that she understood that there were all together “4 schools of thought” and that different responsible agencies were taking different viewpoints. Madam Lam must have appreciated that she was researching into a scientific topic in controversy;

(iii)  According to the Mainland SFDA, trace amount of 1,4-dioxane in cosmetic would not have effect on human health, and there was no regulatory limit of 1,4-dioxane set in the Mainland;

(iv)  Professor Lam opined in some previous interviews that according to USEPA, concentration of 1,4-dioxane in drinking water should not be above 4 ppm.  The amount of 1,4-dioxane tested by the Mainland authority in shower lotion was below this guideline.  Furthermore, it is not for ingestion and so the public should not panic;

(v)  According to a spokesman from Hong Kong Customs, Hong Kong authorities would consult the Mainland and EU safety standards in relation to presence of 1,4-dioxane in shampoo. However, Madam Lam could not find any materials on the EU standard and she dismissed the standard adopted by the Mainland SFDA as unreliable;

(vi)  EU has prohibited the use of 1,4-dioxane in consumer products. USFDA has not set any safety guideline but opined that a trace amount of 1,4-dioxane is harmless and poses no hazard to consumers;

(vii)  1,4-dioxane is not an ingredient intentionally added to any product but can be present in trace quantities in some raw materials. Unintentional presence of trace amount is allowed under EU Cosmetics Directive. Similar allowance is found in many other countries or territories; and

(viii)  Department of Health and Ageing of Australian Government held the opinion that the presence of trace 1,4-dioxane not more than 100 ppm in consumer products is safe (an internet link to NICNAS was provided).  This was the most relevant scientific research before Madam Lam about the safety level of 1,4-dioxane in shampoo, and the NICNAS Report was endorsed by the HK Cosmetic Association.

401. Madam Lam testifies that prior to 29 June 2010, she talked with Professor Lam over the phone.  According to her notes, there were references to “OSHA”, “Air 100 ppm  8-hours work”, “US FDA”  “10 ppm” “food and drug”, “20-30 ppm, long term use will affect liver and kidney”, and “animal experiments shows that it causes liver cancer”.

402. In cross-examination, Madam Lam testifies that “100 ppm” was explained by Professor Lam as relating to working environment.  She was not sure whether it related to Occupational Safety and Health Administration (OSHA), which had in fact set a “permissible exposure limit” of 100 ppm.  Professor Lam also explained about why there were different schools of thought on “carcinogenicity” of 1,4-dioxane.  It was due to the fact that cancer-producing effect only came from animal studies and there was no experiment to prove that it would cause cancer to humans.  According to the testimony of Professor Lam, he would never have commented on carcinogenic effect of 1,4-dioxane on humans because there is no direct evidence of such effect.  He says that he should have done the same to Madam Lam’s enquiry.  I accept his evidence in this regard.

403. With reference to USFDA’s 10-ppm limit, Madam Lam understood Professor Lam as referring to the limit set by USFDA for food and drug.  Professor Lam testifies that when he mentioned “10-ppm limit”, he emphasized to Madam Lam that it referred to food and drug.

404. As for reference to “20-30 ppm”, Madam Lam understood Professor Lam as expressing his own opinion that at this concentration, it just meant that it would affect the liver and kidney.  Professor Lam was not saying that it would cause cancer.  According to Professor Lam, he definitely did not mention exposure of “20-30 ppm” to humans.  He only remembers mentioning “10 ppm”.  Professor Lam admits that a product either below or up to 10 ppm was not a matter of serious concern.  It was only when he was told about a test result of over 20 ppm that caused him concern.

405. Professor Lam does not know about the NICNAS or the EU Report.  Madam Lam had not referred him or asked him to comment on these reports.

406. Madam Lam then conducted an interview with “Mr Chan” on 29 June 2010.  During the interview, “Mr Chan” showed her even more SGS test reports on other brands of shampoos.  “Mr Chan” appeared even more keen than ever.  Madam Lam also accepts that “Mr Chan’s” conduct as “rare” for a complainant.  However, after “Mr Chan” talked about his family history of contacting cancer, it suddenly cleared her suspicion about “Mr Chan”. According to her, a normal person would not lie about the death of his family members.  Madam Lam did not ask “Mr Chan” for his name card to verify his occupation or contact. 

407. Whether because of suspicion or the need to obtain independent evidence, a decision was then made for the Defendant to conduct independent testing of BaWang Shampoo Products.  Madam Lam then purchased 3 bottles of BaWang Shampoo Products and delivered the same to SGS for testing.

408. After meeting “Mr Chan”, Madam Lam conducted another informal interview with Dr Lau over the phone.  According to her notes, Dr Lau mentioned the following points:

(i)  the presence of 3.5 ppm of 1,4-dioxane in shower lotion is acceptable;

(ii)  in the United States, consumer products should not contain more than 20 ppm of 1,4-dioxane;

(iii)  below 10 ppm, “definitely no problem” (“一定無事”);

(iv)  between 10 to 20 ppm, “there is a chance that something will begin to happen if use for long term” (“長期用有機會開始有事”); and

(v)  between 20 to 30 ppm, “long term use may possibly cause cancer” (“長期使用有可能致癌”).

409. The notes clearly show that when Dr. Lau talked about “possibility of causing cancer”, he mentioned that it was based on animal experiments.  Madam Lam actually understood Dr. Lau as meaning that: “It had been confirmed through animal experiments that it will cause cancer.  So for long term use of 20 to 30 ppm, it is possible to cause cancer”.

410. The evidence of Dr Lau is more emphatic.  He points out that the “doctor” (“醫生”) mentioned in the opening paragraph of the Article should not be a reference to him because up to now, he does not think 1,4-dioxane would affect the kidney and he did not emphasize cancer to the reporter.  Dr Lau then mentions about “the danger for media to take words out of context”.  He says that if he had mentioned that 1,4-dioxane may cause cancer, he would have said together with the words “just supported by some experiments on animals”.  When he mentioned between 20 to 30 ppm, “long term use may possibly cause cancer” (“長期使用有可能致”), he said “for experiments on animals, it could be proved, but concerning humans, it was deduced from experiments on animals”.  He is certain that he would have mentioned “animal experiments” (“動物實驗”).

411. I accept the evidence of Dr Lau without any reservation.  His evidence highlights the danger of such kind of casual telephone interview. Although Dr Lau cannot remember everything said by him during the interview, he seems to suggest that his reference to cancer was completely taken out of context.  Further, the reference to “doctor” (“醫生”) in the opening paragraph of the Article could only have been a reference to Dr. Lau.  He was the only medical doctor interviewed.  It is thus clear from Dr. Lau’s evidence that he was misquoted in the Article.  Finally, Madam Lam had not referred Dr Lau to the NICNAS or EU Report for comment.

412. There was also another interesting reference in Madam Lam’s notes about “the costs may be increased by 5 to 6 times”.  In the Article, Dr Lau had been quoted for saying that the costs would be increased by 5 to 6 times if it had to remove all the contaminant of 1,4-dioxane in the product.  Dr Lau cannot remember whether he expressly mentioned the figures of 5 to 6 times, and so I have to accept Madam Lam’s evidence that Dr Lau might have endorsed such figures.  However, Dr Lau is only a medical practitioner with some knowledge in toxicology, and so I am quite certain that it was Madam Lam who brought up this subject.  Obviously, Dr Lau himself was not suggesting that the producer of the shampoos was using cheap or inferior raw materials in the manufacturing process with a mercenary motive to save costs.

413. Then on 8 July 2010, Madam Lam conducted further research on the topic.  7 articles and documents were printed out.  Apart from one of them, all these articles and documents were related to the corporate information about BaWang and its listing.  Madam Lam admits that she researched into the corporate information of BaWang because she was instructed to do so by Mr Cheung Kim Hung (張劍虹)(“Mr Cheung”) who was the content advisor of Book A of the Magazine.

414. The article which was unrelated to the corporate information about BaWang was from an American campaign group called “Campaign for Safe Cosmetics”.  It stated that in 2000, USFDA recommended that cosmetic products should not contain 1,4-dioxane at concentration greater than 10 ppm.

415. Then on 9 July 2010, Madam Lam printed out a news article from Bloomberg about a lawsuit instigated in United States based on “California Proposition 65”.  It explained that under “Proposition 65”, consumer products that contain toxic levels of 1,4-dioxane must have warning labels stating that they may cause cancer.  The article however did not say what amount to “toxic level”.  It did not mention “10 ppm” at all.

416. It was on 12 July 2010 that SGS provided its test results to Madam Lam.  The 3 samples of BaWang Shampoo Products bought by the Defendant independently and sent for testing were all tested with 10 ppm concentration of 1,4-dioxane.  Noting the great disparity in the test results as compared with the earlier one of 27 ppm supplied by “Mr Chan”, Madam Lam called SGS to make enquiry.  This showed that Madam Lam felt concerned about the disparity.  Yet, upon being told by SGS that they could not discuss the earlier report with her because she was not the client, Madam Lam put a sudden halt to her enquiry.  She had not sought the consent of “Mr. Chan” to allow SGS to provide explanation.

417. In fact, had Madam Lam studied and compared the reports in detail, she would have noticed that there were different descriptions of the Sample Receiving Condition in the two reports: “unopened bottle” in respect of the samples submitted by “Mr Chan” and “sealed bottle” for the samples sent by the Defendant for testing.

418. Shortly before the publication of the Article, Madam Lam contacted “Mr Chan” again, and the latter supplied to her the test reports of the other brands of shampoos which were alleged not to have contained any 1,4-dioxane and the receipts for purchasing the BaWang, Rejoice and Acene shampoos.

419. On 12 July 2010, Madam Lam printed out a number of research materials.  According to some of the news reports contained in these materials, Qatar banned the sale and use of shampoos containing a concentration of 1,4-dioxane higher than 10 ppm.  These reports also mentioned that cosmetic products and shampoos containing more than 10 ppm of 1,4-dioxane can cause cancer.  Madam Lam admits at the trial that she did not known how the Qatar Government came about that particular standard.  Interestingly enough, the Plaintiffs started to export BaWang Shampoo Products, without any changes in the formula, for sale in the Middle East countries about 6 months after the publication of the Article, and there was no complaint about the quality of such Products.

420. On 13 July 2010, Madam Lam also printed out other materials.  One of such materials was an article from “Campaign for Safe Cosmetics”.  It mentioned 10 ppm as the “actionable level under Proposition 65”, but the Proposition was about the putting of warning label as clearly shown in a writ of an action in a Californian court obtained by Madam Lam.  There was also an announcement from USFDA in July 2007.  It stated that USFDA had not established or recommended a specific limit on the level of 1,4-dioxane in cosmetics.  If USFDA were to determine that a health hazard exists, it would advise the industry and the public.

421. There were also other research materials found out by Madam Lam which bore no print-out dates, including:

(i)  an article with the title “California Prop 65 Settlement Updates: Part 6”, indicating that Proposition 65 was about warning label and there were consent agreements made by consumer products manufacturers voluntarily agreeing to reduce the level of 1,4-dioxane in their products to no more than 10 ppm;

(ii)  articles mentioning that Procter & Gamble (“P & G”) Herbal Shampoos were tested to have contained 24 ppm of 1,4-dioxane, and P & G agreed to reformulate its herbal shampoos to reduce the levels of 1,4-dioxane to a level of 10 ppm or below, tough it emphasized that it was not a regulatory requirement;

(iii)  articles explaining the nature of 1,4-dioxane, stating that the substance is reasonably anticipated to be a human carcinogen on sufficient evidence of carcinogenicity in experimental animals, but with inadequate evidence for carcinogenicity in humans;

(vi)  NICNAS Report; and

(vii)  a publication by the health authority in Taiwan in April 2009 referring to the NICNAS standard of 100 ppm.

422. During cross-examination, Madam Lam also admits that, prior to 13 July 2010, she had the announcement of the Mainland SFDA dated 21 March 2009 and had read it more than once.  The announcement addressed specifically the safety level of 1,4-dioxane in shampoo products.  According to the announcement:

(i)  it was prohibited to include 1,4-dioxance as a raw material for the production of cosmetic products;

(ii)  if 1,4-dioxane was produced as an inevitable by-product in the manufacturing process, the producer had to comply with the relevant standards and regulations and to ensure that the normal, reasonable and foreseeable use of the cosmetic products would not be hazardous to human health;

(iii)  Mainland SFDA had referred to a number of standards, in particular the NICNAS safety level in consumer products; and

(iv)  Mainland SFDA considered that the satisfactory level of 1,4-dioxane in consumer products (excluding food and drugs) should not exceed 30 ppm, but products containing less than 100 ppm should be regarded as toxicologically acceptable.

(c)  How was the Article drafted and edited?

423. Madam Lam prepared the first working draft of the Article (“the Working Draft”) on or about 12 July 2010[218], which was subsequently submitted to the editorial meeting for discussion.

424. The first editorial meeting which discussed the story about BaWang Shampoo Products was held on 12 July 2010 with the following persons present: Mr Lee Chi Ho (chief editor)(“Mr Lee”), Mr Yeung Wai Hong (publisher), Ms Louise Wong (deputy chief editor), Mr Choy Shun Chiu (person-in-charge of the personal finance section of Book A) and Mr Cheung.

425. Without first consulting Madam Lam or studying the research materials supplied by her, the editorial board decided to change the title of the Working Draft of the Article from a more neutral title of “霸王含致癌物” (BaWang contains carcinogenic substance) to a more charging one “霸王致癌” (BaWang causes cancer).[219] Both Mr Lee and Mr Cheung cannot recall whose idea it was to change the title.

426. The editorial board also made another significant change to the Working Draft: the quotation of Mr Chan “但依家十支內地洗頭水居然有九支致癌” (9 out of 10 brands of Mainland shampoos would cause cancer) was deleted.  The effect of this deletion was that BaWang was targeted and singled out as the shampoo causing cancer.

427. After the editorial board had made all the changes to the Working Draft including the ones mentioned above, the design draft of the Article was then produced including the draft layout of the Article.  Madam Lam and the editorial board then made some final changes and the end product appeared in the Magazine distributed in the market on 14 July 2010.

(d)  How did the Defendant seek the reply or verification from the BaWang Group?

428. One day before the publication of the Article, Madam Lam contacted the Plaintiffs through their public relation company, Porda International (Finance) PR Group (“Porda”), at around 11 am on 13 July 2010 and demanded a response before 3 pm on the same day.

429. Despite the limited time available, the BaWang Group (through Porda) submitted the written response at around 4 pm in the afternoon. The Plaintiffs’ reply contained, inter alia, the following:

(i)  a declaration that BaWang Shampoo Products were safe and the level of 1,4-dioxane in the Products was far below the concentration that would cause harm to human beings;

(ii)  a declaration that BaWang Shampoo Products were produced strictly in accordance with the legal requirements in the Mainland and they complied with all the quality and safety standards in the Mainland and Hong Kong;

(iii)  the Australian health authority considered that the recommended level of 1,4-dioxane in consumer products should be 30 ppm, but products containing less than 100 ppm should be regarded as toxicologically acceptable;

(iv)  the cosmetics monitored by USEPA from 1992 to 1997 were found to have contained as much as 79 ppm of 1,4-dioxane, but USEPA did not consider that would pose a health risk to consumers; and

(v)  OSHA prescribed that workers, based on working time of 8 hours per day, should not be exposed to air with a concentration of more than 100 ppm of 1,4-dioxane.

430. The Defendant had also sought reply from P & G (the producer of “Rejoice” shampoo) and Acene, the two other brands of shampoo found to have contained 1,4-dioxane according to the test reports submitted by “Mr Chan”.  P & G gave a similar reply, referring the Defendant to the WHO’s standard on drinking water, EU’s allowance for the presence of traces of 1,4-dioxane in consumer products as contaminant, the USFDA’s safety limit on food and drug, and the safety limit endorsed by the Australian, Mainland and Taiwanese health authorities.  According to Madam Lam, she also talked with a staff of P & G over the phone.  During the conversation, the staff at first said that the level of 1,4-dioxane in Rejoice shampoo would not be too much, and that it was unavoidable.  However, when Madam Lam confronted him that some other brands of shampoos were tested not to have contained any 1,4-dioxane, that person replied that P & G would do further research to see how to reduce the level of 1,4-dioxane in its products.  Acene gave a shorter reply, stating that 1,4-dioxane was a by-product of a raw material known as AES and there was no known method of removing all the 1,4-dioxane in the end product.  Acene had complied with the strictest safety standard and it would try to improve the formula of the shampoo reducing the use of AES and sourcing AES with less concentration of 1,4-dioxane.  There was also a reference to the safety standard endorsed by the Mainland SFDA.

431. In the meantime, another reporter Mr Kwan Hui Fai (關曉輝)(“Mr Kwan”) and photographer Mr Kwan Wing Ho attended the head-office of the BaWang’ Group in Guangzhou without any prior appointment and demanded response from senior management.  By that time, Madam Wan was not in the office.  An urgent telephone conference was arranged for Mr Kwan to speak to Madam Wan and meeting was arranged for Mr Kwan to meet Madam Yu Xueling (余雪玲)(“Madam Yu”), the Secretary General of the Guangdong Chamber of Daily Used Chemicals (廣東省日化商會) which is a trade association established in 1999 by enterprises in the business of daily used chemicals.

432. During these telephone conference and meeting, both Madam Wan and Madam Yu assured Mr Kwan that any 1,4-dioxane found in the BaWang Shampoo Products was only of very small quantity and such Products would not cause any harm to human body.  Madam Yu also told Mr Kwan that the level of 1,4-dioxane present in BaWang Shampoo Products was lower than the limit of 100 ppm regarded by the Mainland authority to be safe. 

433. Interestingly enough, Mr Kwan cannot remember whether he had been briefed by Madam Lam about the details of her research including the different safety limits relating to 1,4-dioxane in shampoo.  It is fair to say that he had done very little preparation to familiarise himself with the subject before confronting the Plaintiffs in Guangzhou.  After returning to Hong Kong, Mr Kwan was not in any way involved in editing the Article and the reply from Madam Yu was not included in the Article.

434. On 13 July 2010, Madam Lam had also approached Professor Lam for comment on the reply from Madam Wan.  According to Madam Lam, Professor Lam agreed that the presence of 10 ppm of 1,4-dioxane was already “on the margin of being dangerous” (危險邊緣) and 20 ppm was absolutely excessive.  However, the phrase “on the margin of being dangerous” had already appeared in Madam Lam’s Working Draft before 13 July 2010, and so I have reason to believe that the phrase originated from question from Madam Lam, though Professor Lam might have answered yes to the question.  Again, Madam Lam had not referred Professor Lam to the NICNAS Report or its recommended safety level of 100 ppm, nor did she ask him for comment on such safety standard.

435. The Magazine containing the Article was then published the following morning on 14 July 2010.  The share price of the 1^st Plaintiff fell by about 14% by noon, and the trading of the shares was suspended in the afternoon on the same day.

(e)  The dubious background of “Mr Chan”

436. After the publication of the Article, the BaWang Group received an anonymous email regarding the complainant “Mr Chan”.  According to such email, the complainant “Mr Chan” was either one of the two shareholders, both with the surname “Chan”, of a company known as Sun Lok Wah International Co Ltd (“Sun Lok Wah”).  This company was the manufacturer of O’Naomi brand of Chinese herbal shampoo and the OEM manufacturer of Mannings and Watson brands of Chinese herbal shampoo.  According to the test results provided by “Mr Chan” as mentioned in the Article, shampoos of these 3 brands were allegedly found not to have contained any 1,4-dioxane.

437. Further, advertisements of O’Naomi shampoos appeared in newspapers and magazines shortly after the publication of the Article.  A SGS report allegedly showing that O’Naomi shampoo did not contain 1,4-dioxane appeared in the advertisements.  Interestingly enough, the number of that report was the same as that of the report supplied earlier to Madam Lam.  Subsequently in the 28 July 2010 issue of the Eastweek Magazine, there was a report about the Chan’s brothers and their relationship with Sun Lok Wah.  The report also stated that the person suspected to be the complainant was evasive in a confrontation with the reporter of Eastweek Magazine near Sun Lok Wah’s office. After the publication of the Article, Madam Lam tried to contact “Mr Chan” but without any success.

438. The company chop of Sun Lok Wah appeared on an application form dated 21 April 2010 relating to one of the SGS test reports obtained by “Mr Chan”.  Similarly, another application form dated 22 April 2010 showed that the relevant sample was picked up from Sun Lok Wah for testing.

439. I have reminded myself that there is no direct evidence to show that the complainant, “Mr Chan”, was one of the directors or shareholders of Sun Lok Wah.  Neither is there any positive evidence to show that Sun Lok Wah was the OEM manufacturer for Watsons and Mannings shampoos.  On the other hand, it is very difficult to explain why the SGS report supplied by “Mr Chan”[220] to the Defendant earlier also appeared in the advertisement of O’Naomi shampoo, and why Sun Lok Wah was involved in the sample submitted by “Mr Chan” to SGS for testing.  Coupled with the fact that the Defendant could not contact “Mr Chan” after the publication of the Article, I have reason to believe that the complainant was related to Sun Lok Wah who was a competitor of the Plaintiffs.

(v)  Whether the Defendant’s reporting fell below the standard of responsible journalism?

440. Based on the evidence in this case, I conclude that the Defendant’s reporting fell below the standard of responsible journalism.

(a)  Serious nature of the allegations targeting the Plaintiffs

441. First, as admitted by the Defendant’s witnesses, the Article involved serious allegations against the Plaintiffs and they knew full well that the Article would have a serious impact on their reputation and business. The effect of the Article can be demonstrated by what happened to the share price of the 1^st Plaintiff shortly after the publication of the Article.

442. Further, although the Article also referred to the shampoos of Rejoice and Acene as also containing 1,4-dioxane, it is fair to say that the Article was targeted specifically against the Plaintiffs.  As mentioned above, the quotation of “Mr Chan” about 9 out of 10 Mainland shampoos causing cancer was removed from the Working Draft.  Although Acene and Rejoice, in particular the latter, made similar replies to the Defendant’s enquiries, the response of BaWang was singled out as suggesting culpability on the part of the Plaintiffs of refusing to admit guilt.  Further, the Article contained a lot of corporate information about the Plaintiffs and their founders.  Based on all these factors, the readers would have obtained the message that the Plaintiffs (and not the others) were irresponsible, incompetent and greedy commercial organisations which lacked commercial morality, and they used inferior raw materials to produce BaWang Shampoo Products with mercenary motive to save costs.

443. Journalists are certainly allowed to target a piece of investigative journalism against a particular party and the court would be reluctant to intervene as how the journalist would present a particular story to the readers.  However, seriousness of the allegation is one of the ten factors listed out by Lord Nicholls in deciding whether the standard of responsible journalism is met in a given case.  If the journalist is reporting a story containing very serious allegations targeting a particular party, there should be a greater responsibility on the journalist to ensure that: (i) the contents and allegations in the report are true; (ii) the contents of the article are objective and unbiased; and (iii) the target’s side of the story is presented adequately to the readers.  For the reasons given below, the Defendant has failed to meet such standard.

(b)  Serious flaws in the conclusion reached by the Defendant as to the safety level of 1,4-dioxane in shampoo and its failure to verify the scientific basis for the advocated safety limit

444. There are serious flaws in the conclusion reached by Madam Lam and the editorial staff of the Defendant as to the safety level of 1,4-dioxane in shampoo.  In her investigation, Madam Lam came to the conclusion that 10 ppm was the only correct safety limit, and that was the main if not the only reason as to why she reported that BaWang Shampoo Products were unsafe.

445. During her cross-examination, Madam Lam admits that, during her research, she was aware of the distinction between “confirmed animal carcinogenic effect of 1,4-dioxane” and “no sufficient evidence and hence only a possibility of causing cancer to human”, and there were 4 different schools of thought on carcinogenicity of 1,4-dioxane.  Further, she knew full well that whether BaWang Shampoo Products were harmful would depend on how much 1,4-dioxane they contained.  The debate is on what is a safe level.  She also knew about the 100-ppm standard endorsed by NICNAS, HK Cosmetic Association, Mainland SFDA and Taiwanese health authority.

446. When asked as to why she came to the conclusion that 10 ppm is the correct safety limit, she mainly relies on: (i) the articles referring to actionable level for Proposition 65; (ii) the article referring to the USFDA’s recommendation that cosmetic should not contain more than 10 ppm of 1,4-dioxane; (iii) news reports referring to the 10-ppm limit in Qatar and other Middle East countries; (iv) news reports that big manufacturers of shampoo agreed to reformulate their products in order to reduce 1,4-dioxane to a level of below 10 ppm; and (v) news reports about the public concern relating to the presence of 1,4-dioxane in the baby bathing lotion of Johnson & Johnson up to the level of 3.27 ppm; and (vi) the advice or opinion given by the two experts, namely Professor Lam and Dr Lau.

447. As mentioned earlier, the reference to 10 ppm in the context of Proposition 65 does not really suggest a safety limit.  It refers to a duty to put warning on product labels.  The Proposition has nothing to do with the safety level of 1,4-dioxane.  As to the recommendation made by USFDA, it remains a recommendation only and the regulatory authority in the United States does not see any need to impose a statutory safety limit in respect of 1,4-dioxane in consumer products.[221]

448. As to the safety standard in Middle East countries, Madam Lam admits that she did know how these countries came to such standard.  Madam Lam is then asked as to why she preferred such standard over that of NICNAS (which she admittedly had no reason to doubt), she resorts to her extraordinary “Pizza Theory”: if someone says pizza is not good for your health and they put such belief into action by not eating pizza, then pizza must be unhealthy.  In other words, she considers a theory or a standard to be accurate if the persons proposing such theory or standard put it into action or practice.  As the Middle East countries have imposed such standard and put it into action, such standard must be correct.

449. With the greatest respect, this is a wholly illogical, arbitrary, and unscientific approach.  Madam Lam was actually handling a complicated scientific issue about the possible harmful effect of 1,4-dioxane and its safety level in consumer products including shampoo.  She did not care to find out the scientific basis behind the “10-ppm safety standard”.  Further, instead of reporting to the readers about the controversy relating to the safety level, she made a positive assertion that BaWang Shampoo Products containing 10 ppm were unsafe. As I have mentioned above, because of the role of the press, the readers would expect that the journalists would have reached their conclusions, in particular on scientific and technical matters, in an objective way substantiated with at least reasonable materials.  The public has no interest in being informed of such unscientifically and arbitrarily drawn conclusions.

450. Madam Lam admits making a “personal judgment” to prefer relying on the standard adopted by the Qatar Government.  When asked as to the reason for such preference, she says:

“Based on all these studies and researches, from the point of view of the public, it’s safe. But to myself, myself only, I don’t want it contains it. Of course, it’s my own thinking only.”

451. I can understand her sentiment.  1,4-dioxane is a possible human carcinogen.  Even in the absence of scientific proof as to how much 1,4-dioxane would affect human health, it would be arguable, from the health protection perspective, to avoid the substance all together, and one should not criticise Madam Lam if she sought to lodge a campaign to reduce the level of 1,4-dioxane in consumer products which may be a matter involving great public interest.

452. However, that is not the theme or the content of the Article.  The Article contained a positive and serious assertion that using BaWang Shampoo Products was harmful to human health.  The Article targeted BaWang, alleging that the Plaintiffs were irresponsible, incompetent and greedy commercial organisations which lacked commercial morality, and they used inferior raw materials to produce shampoos with mercenary motive to save costs. In view of these serious charges in the Article, Madam Lam had to ensure, so far as it is reasonable, that the scientific community had regarded shampoo containing 10 or 27 ppm as unsafe.  Or at the very least, if the scientific community had no consensus on the issue, she should have presented to the readers both sides of the coin.  As a responsible journalist, Madam Lam was not justified to force her own obstinate view onto readers, as if that was the concluded view reached by the scientists.

453. Madam Lam knew full well about the NICNAS Report and its recommended safety level.  At the time of the publication of the Article, it was the most relevant scientific study on the safety level of 1,4-dioxane in specifically consumer products including shampoos.  At the time of her research, she should have known that the intake of 1,4-dioxane into the body by ingestion and through the use of shampoo are different.  Intake through the use of shampoo by dermal contact and inhalation would be more indirect than ingestion.[222] USFDA only has a limit of 10 ppm on food and drug, and so the safety level of 1,4-dioxane in shampoo should have been much higher.  Madam Lam did not seek further clarification as to whether the 10-ppm limit on food and drug also applied to shampoo.  Madam Lam is a lay person and not a scientist, and that was more so as to why she was not justified or qualified to disregard a most relevant scientific study made by a responsible health agency in a scientifically advanced country.

454. The Mainland SFDA had expressly endorsed the NICNAS safety limit, but Madam Lam discarded the Mainland standard as unreliable.  When Madam Lam is questioned about her suspicion over the announcement of the Mainland SFDA, her answers show that she confused the concepts of “forbidding to use 1,4-dioxane in manufacturing process” as opposed to “safe level of presence as impurity”.  However, I agree with Mr Pow that a careful and objective reading of the Mainland SFDA’s announcement would not give rise to such doubt.  The stance of Mainland SFDA was based on NICNAS which Madam Lam admits she had no reason to doubt.  The same stance was also adopted by the Hong Kong Cosmetic Association, the Hong Kong Customs and the Taiwanese health authority.

455. Madam Lam also relied on some newspaper reports that there was public concern about the presence of 1,4-dioxane in Johnson & Johnson’s baby bathing lotion up to the level of 3.27 ppm.  However, that should not be regarded as the safety level recognised by scientists.  As shown in the newspaper reports obtained by Madam Lam, when some of the experts (including Professor Lam) were interviewed by local reporters about the Johnson & Johnson’s incident, they were of the opinion that the level of 1,4-dioxane in the baby bathing lotion was not a matter of concern.  This is the case even if babies may accidentally swallow some of the bathing lotion which may drastically increase the intake of the substance.  In the casual telephone interview with Dr Lau, he also told Madam Lam that bathing lotion containing 3.5 ppm of 1,4-dioxane should be regarded as safe.  Hence, these newspaper reports did not tell the readers very much as to what was the safety level backed up by science.

456. Madam Lam relies heavily on the advice and opinion given by the two experts over some hastily and informally conducted telephone interviews.

457. I first start with the advice given by Professor Lam.  The only safety level supplied by Professor Lam which was backed up by statutory regulation was 10 ppm put forward by USFDA.  However, that limit only applied to food and drug which would be consumed by direct ingestion.  As shown in the NICNAS Report, intake by direct ingestion was different from the dose that would be absorbed by humans through the regular use of shampoos.[223] Further, as admitted by Madam Lam, Professor Lam had told her that whether 1,4-dioxane is actually carcinogenic to humans is still a subject of scientific debate and there are different schools of thought amongst international agencies.  Even if Professor Lam opined that there is a “possibility of harm to kidney and liver between the range of 20-30 ppm” (there being a conflict of evidence), it was Professor Lam’s personal opinion.  He did not quote any authority to support his view.  He did not elaborate on how serious that “possibility” would be.  In my judgment, without verifying the source of Professor Lam’s view, Madam Lam had failed in her duty as a responsible journalist to verify the accuracy of the materials, in particular she was fully aware of contrary view stated in the NICNAS Report.

458. I then turn to the advice given by Dr. Lau.  Dr Lau was of the view that below 10 ppm is definitely safe.  For 10 to 20 ppm, in the case of long-term use, there is a chance to begin to have problem.  It would be unsafe if the level is over 20 ppm.

459. Dr Lau apparently put forward another standard for consumer products, i.e. 20 ppm, but Madam Lam failed to verify the source of such standard.  Even up to this stage, the court has no idea where this figure came from.  Further, Dr Lau’s comment about cancer had been taken out of context. According to him, he does not think 1,4-dioxane would affect the kidney and he did not emphasize cancer to the reporter.  If he had mentioned that 1,4-dioxane may cause cancer, he would have said together with the words “just supported by some experiments on animals”.  In other words, the effect of the substance on humans is still uncertain.  That is also supported by the words “動物實驗” (animal experiments) written by Madam Lam in her notes.

460. More importantly, it is difficult to understand why Madam Lam had not referred the NICNAS Report to Professor Lam or Dr Lau for comment. Professor Lam has expressly told the court that he did not know about such report.  Madam Lam was fully aware of the conflicting 100-ppm standard recommended by NICNAS.  Since Madam Lam had not confronted either expert with this information for their comment, this conflict was unresolved.  At least, she would not have any basis to prefer one over the other.  The Article involved a serious allegation against the Plaintiffs, Madam Lam, as a responsible and competent journalist, should have done more to seek proper clarification from the two experts.  Even Mr Yu cannot provide the court with a satisfactory answer for such serious omission.

461. Mr Pow also complains that Madam Lam had failed to investigate into the EU safety standard that the Hong Kong Customs said they would equally consult.  Even Madam Lam’s own supervisor, Mr Cheung, had advised her to check the EU standard.  As mentioned earlier in this Judgment[224], shampoo containing as much as 50 ppm was not considered as a matter of concern by the EU Bureau. In her testimony, Madam Lam says she had tried to find the EU standard from the internet but she could not find it.  Mr Pow argues that, according to the Health Canada Report published in March 2010, the EU chemical substances information system would have been available in a specified website in 2009, and so Madam Lam should have been able to find out more about the EU standard from this particular website.  On the other hand, Mr Yu submits that the Plaintiffs cannot now run such argument because it has not been put to Madam Lam during her cross-examination that the EU standard could be found in such website.

462. It is not necessary for me to resolve this particular factual issue.  During her research, there is no dispute that Madam Lam came to know about the NICNAS standard.  Hence, whether she came to know or should have known about the EU standard becomes quite academic. The material issue should therefore be, even if Madam Lam was only aware of the Australian NICNAS standard, whether she should have sought further clarification about such standard from the two experts, and whether she should have included the NICNAS standard in the Article.

463. It is well settled that a responsible journalist would have to verify the source of the information in order to ensure that the report is accurate.  According to Gatley[225], the reliability of the source from which the published material was obtained as it appeared to the defendant at the time, and the steps that were taken to verify that material, play a significant role in deciding whether there is privilege.  Whether the journalist has taken reasonable steps to verify the material has to be judged objectively.[226] In Flood v Times Newspapers Ltd[227], Lord Mance opined that “it will not be, or is unlikely to be, in the public interest to publish material which has not been subject of responsible journalistic enquiry and consideration”.

464. The materials obtained by Madam Lam during her online research were mostly of newspaper articles (as opposed to academic or research papers) and did not include any authoritative statement confirming that the presence of 10 ppm of 1,4-dioxane in shampoo was unsafe. Quite on the contrary, there was an authoritative report by NICNAS suggesting a safety level of 100 ppm.  When Madam Lam was making enquiries from the two experts, she did not verify the sources of the views given by the experts. Neither did she seek further clarification from the experts about the NICNAS Report which was the only authoritative source from scientific community directly on the subject which was available to Madam Lam by that time.  In such circumstances, Madam Lam had only made perfunctory inquiries about the subject which fell below the standard of responsible journalism.

465. Mr Yu emphasises to me that Madam Lam is not a scientist. With the benefit of hindsight and the luxury of time now available, it would be easy for the Plaintiffs to make these criticisms against her judgment. However, in order to protect the role of the press as a public watchdog, generous allowance should be given to journalists as to how to present their stories in particular on matters involving public interest.  The proper question to be asked is therefore not whether Madam Lam’s judgment is right or not, but whether her judgment is one that is entitled to be made by a journalist.  Even if the court disagrees with her judgment, Mr Yu submits that it is still a judgment entitled to be made by her as a journalist.

466. This is a very attractive argument. However, it was clear to Madam Lam and the Defendant that the Article dealt with some technical and scientific matters in controversy between scientists. It was not a story about a political matter involving different viewpoints. Knowing full well that the story was to contain serious allegations against the Plaintiffs which would have a grave impact on their reputation and business, there should be greater responsibility on the Defendant to ascertain the truth before making a positive allegation that 10 ppm was the safety limit accepted by scientists and BaWang Shampoo Products were unsafe.  This is not a case that the Defendant did not know about the contrary standard, but the Defendant chose to ignore that standard and presented the half truth to the readers.  To me, that was not responsible journalism.  Even taking into account the generous allowance given to reporters in conducting their works, there must be a standard which the court and indeed the public would expect the press to maintain.  Hence in my judgment, that was not a judgment entitled to be made by a responsible journalist.  To lower the standard in such circumstances would encourage the press to be biased or to adopt personal, arbitrary or irrational approach in making reports on scientific issues.  This would not be in the public good.

(c)  Failure to present fairly or adequately the contrary scientific view and the reply of the Plaintiffs

467. Even if Madam Lam was entitled to make the judgment that 10 ppm was the safety limit, Madam Lam or the editorial board of the Defendant, as a responsible journalist, should have presented in the Article fairly the contrary expert views expressed by other responsible scientists and health agencies.  Further, the Defendant had also failed to present adequately the Plaintiffs’ side of the story and their reply.

468. It had been said clearly in Reynolds v Times Newspapers Ltd that where the defendant already has the claimant’s version of events, failure to publish that will almost inevitably count against the defendant:[228]

“A failure to report the other side will often be evidence tending to show that the occasion ought not to be protected by qualified privilege. But it would not necessarily always be so, e.g. when the victim’s explanation is unintelligible or plain nonsense.”

469. Lord Nicholls explained the requirement as follows:[229]

“It goes without saying that a journalist is entitled and bound to reach his own conclusions and to express them honestly and fearlessly. He is entitled to disbelieve and refute explanations given. But this cannot be a good reason for omitting, from a hard-hitting article making serious allegations against a named individual, all mention of that person's own explanation. Particularly so, when the press offices had told Mr Ruddock that Mr Reynolds was not giving interviews but would be saying all he had to say in the Dáil. His statement in the Dáil was his answer to the allegations. An article omitting all reference to this statement could not be a fair and accurate report of proceedings in the Dáil. Such an article would be misleading as a report. This article is not defended as a report, but it was misleading nonetheless. By omitting Mr Reynolds's explanation English readers were left to suppose that, so far, Mr Reynolds had offered no explanation. Further, it is elementary fairness that, in the normal course, a serious charge should be accompanied by the gist of any explanation already given. An article which fails to do so faces an uphill task in claiming privilege if the allegation proves to be false and the unreported explanation proves to be true.”

470. Lord Nicholls’ dicta also highlight the harm that can be caused by misleading or half-truth story.

471. First, I have to emphasise that, at the time of the publication of the Article, the NICNAS Report was the most relevant scientific study on the safety of 1,4-dioxane in consumer products including shampoo.  In fact, all the available scientific studies by that time, including NICNAS, EU and Health Canada Reports, though Madam Lam was not aware of the latter two, suggested that the level of 1,4-dioxane in BaWang Shampoo Products, whether it was 10 or 27 ppm, was within the safety level recommended by these responsible health agencies.  Contrary to the picture portrayed in the Article that 10 ppm was the safety limit accepted conclusively by the scientific community, the scientific materials available at that particular time were all one-sided pointing the contrary.  Hence, even if Madam Lam might disagree with the higher standard, of which she had no reason to doubt, she should have included such standard in the Article.

472. Even Mr Cheung agrees that the Defendant had to be objective, and adequate opportunity should be given to the BaWang Group to reply to the allegations.  In his witness statement, he stated that he was “very concern to make sure that the article that we published was fairly presented and in particular, that BaWang were given the opportunity to comment/respond to the SGS investigation results”.

473. At the trial, I have actually asked Mr Cheung whether the Defendant should have included the safety limit recommended by NICNAS in the Article.  Mr Cheung’s answer is interesting.  He says that it would be better to include the NICNAS’s safety standard, but he does not see any problem with such omission.  These are the words from an experienced journalist.  I also find it surprising that Mr Cheung still gives this answer now, despite the fact that he should have known there would be great controversy in this trial between the toxicology experts about the proper safety standard.

474. I do not think any journalists can deny that one of their important responsibilities is to be objective.  If journalists are allowed to be biased or prejudiced, or to omit to put forward fairly the opposite view of their story or to present the readers with the half truth, that would destroy the confidence of the public in the press, and the readers would tend to disbelieve any materials contained in their reports, whether true or false.

475. As pointed out by Lord Nicholls in the dicta above, whether the omission to present the contrary view is material would also have to be weighed against the seriousness of the allegations against the person or entity being targeted.  Journalists should be given generous allowance as to how to present their stories.  But if the journalists knew full well that their serious allegations would gravely affect the reputation and business of others, they should try to handle the report in a fair manner and to present adequately the reply and explanation given by the person or entity being targeted.  It would be dangerous to allow journalists just to include the materials favourable to the story without letting the readers to know the whole truth.

476. In my judgment, the Plaintiffs’ replies were not fairly or accurately presented in the Article at all.  Most importantly, the Defendant failed to mention the NICNAS’s standard of 30 and 100 ppm which was relied upon by the Plaintiffs and the Mainland SFDA.  In the Article, the Plaintiffs’ response had been portrayed as a groundless or bare assertion.  The readers were prevented from knowing that in fact, the Plaintiffs’ stance of BaWang Shampoo Products being safe was supported by a scientific study conducted by a responsible team of scientists; published by a responsible health agency; endorsed by the government of a scientifically advanced country.  This is particularly important because the Article also accused the Plaintiffs of hiding behind international regulatory lacuna or loopholes. Readers were also prevented from knowing that regulatory agency of their own country (i.e. the Mainland) had also relied on and referred to this scientific study in its public announcement.

477. In the Article, Madam Lam had written that there was no definition of what were minute traces of 1,4-dioxane around the world (可惜各地均沒有界定何為微量). This is certainly a misleading if not wrong statement.  By avoiding the reference to the viewpoints of NICNAS, EU and the Mainland SFDA’s etc., such statement caused readers to be misinformed in that there were no other expositions of safety level other than the 10 ppm advocated by Madam Lam.  The express omission to such important standard can only mean that Madam Lam had tried to portray that BaWang Shampoo Products were dangerous products.  I agree with Mr Pow that it was a classic example of selective, biased and imbalanced reporting.

(d)  Inclusion of unnecessary defamatory accusations against the Plaintiffs

478. As mentioned above, the “subject matter” of the publication may be of public interest and concern is not a warrant for the inclusion of all sorts of defamatory accusations targeting a particular party, otherwise it would just be a blanket licence to journalists to publish whatever they like and they do not need to exercise any care in their reporting.  This cannot be in the public interest.

479. The Article went beyond raising the awareness of the public about the presence of 1,4-dioxane in shampoos.  In discussing the meanings of the Words in the earlier part of this Judgment, I have already found that the Words carry the defamatory stings that the Plaintiffs were irresponsible, incompetent and greedy commercial organisations which lacked commercial morality, and they used inferior raw materials to produce shampoos with mercenary motive to save costs.  In the trial bundle, there are a lot of news reports in other free societies discussing the public concern of 1,4-dioxane in shampoos.  However, targeting these sorts of defamatory accusations against a particular commercial entity is uncommon.

480. As demonstrated by the case law, the court should respect the editorial judgment as to how to present a particular story to the readers.  But there must be a limit that has to be observed by responsible journalists, otherwise they can include all sorts of unnecessary remarks in the report putting forward the excuse that they are part of the stories.

481. Mr Yu submits that it was justified for the Defendant to target the Plaintiffs.  First, BaWang Shampoo Products contained the highest level of 1,4-dioxane as compared with the other samples of shampoos submitted by “Mr Chan” to SGS for testing.  Second, BaWang Shampoo Products, being advertised as a natural product containing Chinese herbal, were popular in the market by that time, and so the readers should be told about the presence of 1,4-dioxane in their shampoos.

482. I agree that the public should be told about the presence of 1,4-dioxane in BaWang Shampoo Products.  But that was only part of the story.  The public should also be told, as part of the Plaintiffs’ reply, that the most relevant scientific study available at that time showed that the level of 1,4-dioxane present in BaWang Shampoo Products was below the recommended safety level.  By accusing the Plaintiffs of being irresponsible and evasive and in the wrong for refusing to admit guilt, the readers should be told at the same time that the Plaintiffs’ stance was backed up by responsible scientific study and not just bare allegation.  It would be irresponsible to force the others to accept something that the Plaintiffs reasonably believe to be true.

483. Even worse, the Defendant accused the Plaintiffs for using inferior raw materials for maximising profits.  Such inferential statement was wholly speculative.  At the material time, Madam Lam had no factual basis for making such speculation.  I agree with Mr Pow that the purported reliance on the general comments of the two experts as the basis for drawing such conclusion was plain non sequitur.  In particular, Dr Lau is just a medical practitioner and so I do not accept that he would have volunteered such information to Madam Lam.  On the other hand, Madam Lam ignored her research which should have revealed that 1,4-dioxane was commonly present in shampoos and cosmetics as a manufacturing by-product.  At the time of the report, the use of inferior raw materials to save costs could only be a speculation.  The Defendant simply had no basis to refer this cause as a distinct probability which was the meaning of the Article.  The public has no interest in receiving such groundless insinuation.

484. If journalists are allowed to make these sorts of accusations freely without the readers being told the whole truth, there is real danger that the press is able to manipulate public opinion against a particular entity by selecting only facts that are favourable to the story.  Again, this would not be in the public interest for the readers to receive such kind of disinformation.

485. According to the evidence of Madam Lam and Mr Cheung, the contents of the Article might be different if BaWang Group had agreed to reduce the level of 1,4-dioxane in their shampoos, rather to insist that their shampoos were safe.  The Defendant seems to rely on such matter to justify why it had targeted the Plaintiffs.

486. Again, this was another adverse judgment passed on the Plaintiffs as to how they handled the matter.  But should the readers be told the whole picture so that they could make a proper judgment themselves?  If the defamatory accusations made by the Defendant against the Plaintiffs were correct (which I do not find it to be the case), those accusations were equally applicable to the other manufacturers whose shampoos were found to have contained 1,4-dioxane.  The fact that these manufacturers made a promise to reduce the level of 1,4-dioxane in their shampoos, after the matter had been exposed by the press, cannot change the fact that they were also irresponsible commercial entities and they used inferior raw materials to produce shampoos with a view to save costs, if that was indeed the case.  Hence, the Defendant’s explanation for targeting the Plaintiffs is hardly convincing.

487. In addition, Mr Cheung had already asked Madam Lam to collect information about the BaWang Group on 8 July 2010 before they obtained the independent testing reports from SGS and the replies from various shampoo producers.  Coupled with the fact that the Defendant did not conduct independent testing of other shampoos in the market, I have reason to believe that the Defendant had already targeted the BaWang Group as irresponsible commercial entity irrespective of its reply.

488. For the above reasons, it served no public interest for Defendant to include the unnecessary defamatory accusations targeting the BaWang Group in the Article.

(e)  Failure to verify the credibility of the information provided by the complainant

489. In the Article, Madam Lam referred to the test result of the BaWang shampoo submitted by “Mr Chan” to SGS for testing (“the Dubious Sample”), i.e. 27 ppm.  The test result was also relied upon by the Defendant to support that BaWang Shampoo Products were unsafe.  As a responsible journalist, the Defendant therefore had a duty to verify the genuineness of such information.

490. According to Madam Lam, 10 ppm was the safety limit and so one can argue that BaWang Shampoo Products independently tested by the Defendant might not be unsafe because it was right on the upper limit of the safety level.  Even according to the opinion of Dr Lau, it was just at the lowest end of the range of “possible beginning to have problem after long term use”.  According to Professor Lam, a product either below or up to 10 ppm was not a matter of serious concern.  It was only when he was told about a test result of over 20 ppm that caused him concern. Hence, the inclusion of the test result of the Dubious Sample would serve the purpose of showing that BaWang Shampoo Products were clearly unsafe because it was well above the 10-ppm safety limit. 

491. As admitted by Madam Lam herself, she had some suspicion about “Mr Chan” as a complainant.  “Mr Chan” was a particular keen complainant.  He paid for the tests and he was well aware of the “safety limit” of 10 ppm.  Yet Madam Lam did not verify “Mr Chan’s” background and she turned from “suspicion” to “trust” not on the basis of any objectively verifiable facts or information.  In a way, Madam Lam had adopted a rather naïve and unprofessional approach in her investigation.

492. There is great disparity in the test results between the Dubious Sample and those purchased by the Defendant.  If the safety limit was indeed 10 ppm as advocated by Madam Lam, such disparity was crucial in determining whether BaWang Shampoo Products were unsafe.  After noting such disparity, Madam Lam contacted SGS for enquiry.  After she was told that SGS could not discuss the test report of the Dubious Sample because the Defendant was not the direct client, she just left the matter at that.  Madam Lam also failed to notice the difference in the descriptions of the Sample Receiving Condition in the reports.

493. During cross-examination, Madam Lam was asked why, having obtained her own test reports of the 3 BaWang shampoo bottles that she felt sure about their veracity, which were consistently tested of 10 ppm, she continued to make use of “Mr Chan’s” 27-ppm test report in her Article.  Madam Lam explains that it was because the subject matter of the Article was “Mr Chan’s” complaint, she had to write everything he provided upon doing the test and the report.

494. However, since the Article involved serious allegations, Madam Lam was under a greater responsibility to verify the information because she was relying on such test result to show that BaWang Shampoo Products were unsafe.  The difference in the descriptions of the Sample Receiving Condition (unopened bottle and sealed bottle) might have alerted Madam Lam as to the possibility of tampering with the sample supplied by “Mr Chan”, and yet she made no enquiry with SGS about the difference between the two descriptions.

495. Even if Madam Lam wanted to keep the complaint by “Mr Chan” as a background to the Defendant’s investigation, there was no need to refer to something that she felt unsafe to rely on.  She could have proceeded with the Article on the basis of 10 ppm tested out of BaWang Shampoo Products.  In a way, Madam Lam included the 27-ppm test result in order to run the story that the Plaintiffs were clearly the parties to be blamed and to project BaWang Shampoo Products as containing way beyond what she advocated as the safety limit.

496. Further, the test result of the Dubious Sample was an important basis for the comments given by the two experts Professor Lam and Dr Lau.  As mentioned above, based on their advice about the safety limit of 1,4-dioxane, it is very arguable that the presence of 10 ppm of 1,4-dioxane in shampoo was not a matter of serious concern.

497. Mr Pow is therefore justified in saying that Madam Lam was only making “half-hearted” attempt in her investigation.  This would be a factor weighting adversely against the Defendant as to whether it had discharged the duty as a responsible journalist.

498. There is also an issue as to whether the court has to make a factual finding as to whether the Dubious Sample had indeed been tampered with.  My answer is no.  As mentioned above, such factual dispute would not be relevant to the defence of justification.[230] So far as the defence of publication in the public interest is concerned, the issues are whether, taking into account the dubious features which should have alerted Madam Lam about the veracity of the Dubious Sample, Madam Lam should have included the 27-ppm test result in the Article or used such test result as a basis to conclude that BaWang Shampoo Products were unsafe.  The determination of this factual dispute is therefore not necessary.

499. If I am wrong on such issue and I have to make a factual finding in this regard, I would find that, on the balance of probabilities, the Dubious Sample had been tampered with.  The only person who can give direct evidence on this question is “Mr Chan” himself. Because he could not be located after the incident, it would be quite impossible for both sides to call him to testify at the trial.  The court therefore has to rely on the following circumstantial evidence.  First, the descriptions of the Sample Receiving Condition are different in the reports.  Second, the unusually high concentration of 1,4-dioxane in the Dubious Sample does not sit well with Madam Jiang’s evidence that each batch of ALS had been certified by Hunan Resun to have contained no more than 40 ppm of 1,4-dioxane (only 25% of it would go to the final product).  Third, there was a raid by the Mainland SFDA of the 2^nd Plaintiff’s production facility after the publication of the Article, and the BaWang Shampoo Products found therein were tested to have contained only about 6 ppm of 1,4-dioxane.  Fourth, BaWang Shampoo Products were quite successful in the market by that time, and “Mr Chan”, being a person related to the Plaintiffs’ competitor, had commercial reasons to sabotage the Plaintiffs’ business.  In particular, O’Naomi was readily prepared to lodge advertising campaign at about the same time of the publication of the Article, focusing on the absence of 1,4-dioxane in their products.  Fifth, “Mr Chan” was a keen, “knowledgeable” and suspicious complainant, and he could not be contacted by the Defendant after the publication of the Article.  All such circumstantial evidence leads me to the conclusion that, on the balance of probabilities, the Dubious Sample had been tampered with and it is also unsafe for the court to rely on the other test results supplied by “Mr Chan”.

(f)  Other considerations

500. The Plaintiffs are also relying on other complaints to claim that the Defendant’s works fell below the standard of responsible journalism.

501. First, they say that the way in which the Defendant sought reply from them was oppressive.  Given that there was no urgency in the publication of the Article, the Defendant gave very limited time (about 4 hours’ notice) to the Plaintiffs to respond before the publication of the Article, in particular the Defendant should have known that the subject involved considerable scientific researches and materials.  When the relevant management staff handling such kind of enquiry was not present in the office, the Plaintiffs were portrayed in the Article as being evasive in dealing with the enquiry.  The Article did not mention that the Plaintiffs had in fact arranged for the reporters to attend a telephone conference with Madam Wan and the reporters were allowed to take photographs in Madam Wan’s office.  Neither was the arranged visit to Madam Yu’s office mentioned.  By omitting these facts and together with the other defamatory stings, Mr Pow submits that the Article certainly suggested culpability on the part of the BaWang Group and it was trying to hide away from the journalists after their “fault” had been revealed.

502. There is some justification in the Plaintiffs’ complaint.  If the journalists are writing on some daily news, I can perfectly understand why very limited time is given to the interviewee for reply.  After all, news can be a perishable item.  However, the story here, if it could be stood up, would be of interest at any time, and would not become stale even if the Defendant took some time to verify the story in light of the explanation and assurance given by the Plaintiffs.  On the contrary, if the story was flawed, the Defendant must have appreciated that serious harm could be occasioned to the Plaintiffs.  As responsible journalist, the Defendant should have therefore handled the enquiry in a less oppressive manner, at least they should not have portrayed the Plaintiffs as being evasive.

503. Mr Cheung explains that it was not the policy of the Defendant to give ample time to interviewees to answer the enquiries, the reason being that “if a subject has notice of an investigation. The subject will then try and limit its exposure by speaking to other (“friendly”) publication to dilute the effect of what until then was an exclusive report”. Further, Mr Yu submits that there is nothing to indicate that the Plaintiffs’ reply could have been different if more time was given to them.

504. Mr Pow criticises the Defendant’s policy as wholly self-imposed with a selfish objective, and it had clearly placed its own interest and benefit above truthfulness and quality of its publication.

505. I can understand why the Defendant was seeking to guard jealously the exclusivity of the report, and there is also no evidence to suggest that the Plaintiffs’ reply could have been different given more time.  However, the Article certainly conveyed the message to the readers that the Plaintiffs were evasive when confronted by the reporters, which I do not find it to be the case.  Coupled with the other defamatory statements implicating culpability on the part of the Plaintiffs, it reinforced the readers’ impression that the Plaintiffs were irresponsible corporations lacking commercial morality.  This would be a factor for the court to take into account in concluding that the report in the Article was not a product of responsible journalism.

506. Secondly, Mr Pow complains that the editorial staff of the Defendant had deliberately changed the title of the Article from a more neutral title of “霸王含致癌物” (BaWang contains carcinogenic substance) to a more charging one “霸王致癌” (BaWang causes cancer) with a mercenary motive of attracting potential readers into buying the Magazine, and on this ground alone, the defence of publication in the public interest should be thrown out by the court.

507. Both Mr Lee and Mr Cheung cannot provide a satisfactory explanation as to why the editorial meeting decided to change the title of the Article without first consulting Madam Lam.  To me, it is obvious that the change of the title was made for the purpose of increasing the impact of the Article and to attract readers into buying the Magazine.  By itself, it should not be a matter of criticism.  Journalists should be given generous allowance as to how to present their stories, and there is no objection for editors to make their stories more appealing to readers.  Further, carcinogenic substance is in Chinese known as “致癌物” (cancer causing substance).  “Causing” (致) may denote that there is a possibility of getting cancer, not that it would definitely cause cancer.

508. After perusing the topic, the readers would certainly read the substance of the Article to ascertain the extent of the health risk of BaWang Shampoo Products.  In the Article itself, it was clearly suggested that the level of 1,4-dioxane in BaWang Shampoo Products was either in the margin of being dangerous (10 ppm for the shampoos independently tested by the Defendant) or well above the safety limit (27 ppm for the shampoo submitted by “Mr Chan” for testing).  Coupled with the wording of the title, the clear message conveyed to the readers was that it would be dangerous to use BaWang Shampoo Products because it materially increases the risk of getting cancer.  I have already discussed in length as to why I conclude that such report was not a product of responsible journalism and I do not want to repeat the reasons again. 

509. Thirdly, the Plaintiffs also complain that the Defendant had failed to put in place an effective system of verifying and cross-checking the work of Madam Lam.  Contrary to the evidence in their witness statements, it is clear that Mr Cheung and Mr Lee had not perused the research materials prepared by Madam Lam.  Neither did they pay a lot of attention to the formal reply made by the Plaintiffs on 13 July 2010.  To them, Madam Lam was a competent reporter and so they relied on her work and judgment.

510. So far as responsible journalism is concerned, the court will look at the work and the report made by the Defendant as a whole, and so for the purpose of determining whether the Defendant had met the standard, it is not necessary for the court to examine in details the internal communications between Madam Lam and the editorial staff of the Defendant, or how the editorial staff supervised Madam Lam in respect of her work.

511. There is also an issue about a typographical error in the Article.  In explaining the amount of 27 ppm, the Chinese description actually refers to 27 milllitre in every litre of shampoo.  It has exaggerated the actual amount by 1,000 times.  In my judgment, if the readers are reading the Article as a whole, those who notice the difference would understand that it is a typographical error.  For those who do not notice the difference, it would not matter at all.  Hence, such error is not a material consideration.

512. Before leaving this subject, I would have to address one more issue as to whether there was any sinister motive behind the report made by the Defendant in the Article.

513. Mr Pow argues that the only reasonable inference for the Defendant in not seeking further clarification about the NICNAS standard from the two experts or to put forward the contrary view in the Article was that the Defendant wanted to avoid the issue.  The presentation of the contrary view would lessen the impact of the Article, with the result it would reduce the attraction of the Article to the readers. Hence, the Defendant was reckless as to the truth of the contents in the report.

514. I have carefully considered the evidence of Madam Lam.  I do not accept that she had deliberately tried to harm the Plaintiffs with what she knew were untrue statements.  However, Madam Lam had adopted a naïve and unprofessional approach in reporting the story.  She simply trusted the story given by the complaint.  Influenced by her own personal view that 1,4-dioxane is an unpleasant substance, she carelessly came to the view that the NICNAS standard, which was endorsed by the regulatory authorities in Australia, Mainland, Taiwan and Hong Kong, must be incorrect.  She just relied on some newspaper reports and the opinion of the two experts without seeking further clarification as to the scientific basis behind the so called 10-ppm limit.  The editorial board just relied on the judgment and the work of Madam Lam and came to the view that 10 ppm was the safety limit.  As a result, the Defendant published the Article in a sensational manner resulting in serious damage to the reputation of the Plaintiffs.  Although there was no malice as such, the Defendant’s reporting still fell below the standard of responsible journalism.

(g)  Conclusion on responsible journalism

515. I do not accept that maintaining such standard of responsible journalism would discourage the press from discharging their function as a watchdog in a free society.  The Defendant was handling a piece of investigative journalism involving scientific issues about the possible harmful effect of 1,4-dioxane on humans and the safety level in shampoos.  The Defendant had ignored the only scientific study available to Madam Lam at that time showing that the 1,4-dioxane present in BaWang Shampoo Products was below the recommended safety level.  She failed to verify the source or the scientific basis for the support of the 10-ppm safety level.  Forgetting the important duty to be objective, she failed to let the readers to know that there was other standard proposed by a responsible team of scientists, and the Defendant forced its view onto the readers to the effect that 10 ppm was the only conclusive standard accepted by the scientific community.  She failed to verify the information supplied to her by the complainant.  Even worse, the Defendant included all sorts of unwarranted defamatory accusations against the Plaintiffs.  All these were done with the full knowledge that the serious allegations made in the Article would gravely affect the reputation and business of the Plaintiffs.

516. This case is not so much about the freedom of the expression or the press.  We all treasure these values.  The key question here is whether journalists should be held liable for libel if they are performing their works irresponsibly.  In my judgment, maintaining the standard of responsible journalism would not discourage the press from discharging its function as a watchdog in a free society, but rather it serves as a reminder to journalists that they have to be objective in performing their works, which is the cornerstone for the press in exerting its influence as the fourth power in every civilized society.  It also helps to maintain the confidence of the public in the professionalism of the press.

517. Despite that the court has to adopt a liberal approach in the application of the defence of publication in the public interest, I find that the Defendant had failed to meet the standard of responsible journalism.  Further, it would not offend the necessity and proportionality principle under the human right jurisprudence to make the Defendant liable for the defamation against the Plaintiffs.  It is necessary to protect the reputation of the Plaintiffs and to restrict the publication of the Words as the Article was not a product of responsible journalism.  Hence, I also reject such defence.

VI:   INJURIOUS FALSEHOOD

518. The parties have spent relatively little time in dealing with the claim on injurious falsehood.  In view of my ruling on the claim for defamation, this claim now becomes academic so far as liability is concerned.  However, whether the Plaintiffs can maintain the claim for injurious falsehood may have certain implications relating to quantum, and so I have to give my judgment on this particular claim as well.

519. It is trite that in order to establish a claim of malicious falsehood, the plaintiff must show the following:[231]

(i)  the defendant has published about the plaintiff words which are false;

(ii)  the defendant did so maliciously in that either:

(a)  he knew when he published the words that they were false; or

(b)  he was reckless as to whether they were true or not; or

(c)  he did so for the purpose not of advancing his own interest but injuring the plaintiff.

520. In my judgment, the Plaintiffs’ claim on injurious falsehood cannot succeed because of the following reasons: firstly, the Plaintiffs have failed to discharge the burden of proving the falsity of the statement; and secondly, the Defendant did not publish the statement maliciously.

(i)  Falsity of the statement

521. The statement in the Article relied upon by the Plaintiffs in support of such claim is that BaWang Shampoo Products would, even under normal and regular use, cause serious illnesses and even cancer.  Unlike the defence of justification, it would be the burden on the Plaintiffs to establish the falsity of such statement.

522. In dealing with the expert evidence on toxicology, I have already outlined the two schools of thought relating to the safety level of 1,4-dioxane in shampoo.  On the one hand, we have the safety levels endorsed by NICNAS, EU Bureau and Health Canada, suggesting that the level of 1,4-dioxane in BaWang Shampoo Products is safe. On the other hand, we have the more cautious approach adopted by USEPA and Dr Sawyer, suggesting very minute level of 1,4-dioxane in shampoo (even less than 1 ppm) is dangerous.  If the latter’s approach is correct, the world may have to consider to ban the presence of any 1,4-dioxane in consumer products, but the reality is that most of the responsible  regulatory agencies do not see the need to take any action in this regard.

523. Both groups of scientists use different risk models to assess the possible health risk of 1,4-dioxane in shampoo.  However, the risk worked out by them is very much a theoretical or mathematical risk, and not a real or factual one based on empirical data. These models are based on different assumptions which may be right or may be wrong.  At this stage, although a lot regulatory agencies around the world do not see the need to set a safety limit of 1,4-dioxane in shampoo, the Plaintiffs cannot, based on the existing scientific materials, prove the falsity of the statement on the balance of probabilities.  Hence, the Plaintiffs’ claim on injurious falsehood cannot succeed.

(ii)  Malice in publishing the statement

524. The Plaintiffs also claim that the Defendant published the statement maliciously, being at least reckless as to whether the statement was true or false.

525. In the earlier part of this Judgment on the defence of publication in the public interest, I have already discussed in length as to how Madam Lam conducted the research and how she and the editorial staff came to the conclusion that 10 ppm was the safety limit.  I accept Madam Lam’s evidence that she subjectively believed that 10 ppm was the safety limit.  The problem is that the way she came to such conclusion was flawed, naïve and unprofessional, and as a result her reporting fell below the standard expected from a responsible journalist.

526. Nevertheless, it is not equal to say that Madam Lam was reckless as to the truth of the statement.  One must bear in mind that “malice” and “responsible journalism” are two different concepts.  Obviously, if the journalist publishes a report maliciously, it would be extremely difficult for him to establish that he has met the standard of responsible journalism.  But even in the absence of malice, it would still be possible for the court to make a finding of not meeting the standard, having regard to Lord Nicholls’ 10 factors listed out in Reynolds.

527. Mr Pow also complains that the Defendant had deliberately published the Article in a serious and sensational manner with a mercenary motive of attracting more readers into buying the Magazine.  Again I do not accept that this factor alone is sufficient to establish that the Defendant published the statement maliciously.  Journalists should be allowed certain flexibility in presenting their stories to the readers, and there is no objection for the journalists to make their stories more appealing to the readers.  However, Madam Lam’s investigation was flawed.  Clouded by her own “personal judgment”, she chose to rely only on the materials supporting her story without taking the necessary steps to verify the scientific basis for the support of the “10-ppm safety limit”.  She believed that her judgment was correct, and the readers should only be told of what she considered to be more important and correct.  As a result, she did not inform the readers the contrary scientific view supporting the Plaintiffs’ explanation.  In my judgment, though the investigation by the Defendant was irresponsible, this is not sufficient to establish recklessness on the part of the Defendant.

528. For the above reasons, I find that the Defendant failed to meet the standard of responsible journalism but not that it published the statement maliciously.  The claim on injurious falsehood also fails on this ground.

VII  QUANTUM

529. The Plaintiffs succeed in the claim for defamation but not injurious falsehood.  I therefore proceed to assess the quantum of the Plaintiffs’ claim.

530. The issues relating to quantum are not much easier.  The bulk of the Plaintiffs’ claim relates to the loss of sales suffered by the Plaintiffs in another jurisdiction, i.e. the Mainland.  In most of the defamation cases in Hong Kong, the claimants would just focus on the losses suffered in the local jurisdiction.  The Plaintiffs’ claim therefore touches upon a lot of issues which have not been canvassed in the Hong Kong courts before.

531. I try to give a bird-eye view of the issues relating to quantum.

532. In this action, the Plaintiffs claim the following losses for defamation (and perhaps malicious falsehood) against the Defendant:

(i)  damage to the reputation of the Plaintiffs;

(ii)  loss of business and profits suffered by the Plaintiffs in the period from 15 July 2010 to 30 June 2011 (“the Relevant Period”) in the sum of RMB¥448,354,000;

(iii)  special damages suffered by the 1^st Plaintiff in the sum of HK$4,652.50 relating to the cost of issuing public announcement as required by the listing rules in Hong Kong:

(viii)  special damages suffered by the 2^nd Plaintiff, including the cost of handling media enquires and issuing press statements in the sum of RMB¥3,645,586, the cost of engaging media consultants and carrying out the advertising campaign to restore the image of BaWang Shampoo Products in the sum of $57,791,679.04, and the cost of sponsoring the Asian Games in 2010 to restore the image of BaWang Shampoo Products in the sum of RMB¥3,359,000; and

(ix)  exemplary damages.

533. The main part of the Plaintiffs’ claim relates to the loss of sales and profits relating to BaWang Shampoo Products during the Relevant Period.  In respect of such claim, the first dispute is about the calculation of the actual figures for the loss.  In this regard, both sides have called accounting experts to testify on the calculation.  The Plaintiffs’ expert is Mr Steven Skalak (“Mr Skalak”) and the Defendant’s expert is Mr Trevor Dick (“Mr Dick”).  There are disputes about the methodology and approach adopted by the respective experts in assessing the actual loss.

534. There is no dispute that both Mr Skalak and Mr Dick are qualified accounting experts.  However, both of them attempt to give expert evidence on market conditions relating to hair care products both in Hong Kong and the Mainland.  I myself have reservation as to whether they have expertise in this area.  However, some of their evidence is backed up by market data and so I would approach their evidence with such caveat in mind.

535. The second main issue relates to the claim for loss of business and profits caused by the “republications” by third parties in the Mainland.  Based on the query of the lack of juridical basis expressed by Eady J in Baturina v Times Newspapers[232], Mr Yu submits that the Plaintiffs should not be allowed to recover damages flowing from such republications in the Mainland and which are not the subject matters of the separate causes of action in the present claim.

536. At the trial, the parties have used the term “republications” loosely to cover any further republications which allegedly repeated the gist of the Words in the Article.  “Republications” do not mean exact replication of the original Article, as their contents are quite different from those in the original Article.

537. The third issue is very much a question of causation and remoteness of damages.  Most of the losses allegedly suffered by the Plaintiffs relate to the loss of sales in the Mainland.  The Magazine was published in Hong Kong, and it is common ground that the circulation of the Magazine in the Mainland was limited and restricted.  However, the 1,4-dioxane incident relating to the Plaintiffs were referred to in a lot of republications by third parties in the Mainland, mainly in the internet.  The contents of these republications were different from those in the original Article.  Hence, there is a serious dispute as to whether the loss of sales in the Mainland was caused by the publication of the original Article itself, and whether such loss was foreseeable at the time of the publication of the Article.

538. The fourth issue is a conflict of law issue.  It is the Defendant’s case that, under Mainland law, the Defendant would only be liable to the Plaintiffs for the loss caused by its own publication of the Words and not the republications by third parties.  Hence, the Plaintiffs should not be allowed to recover the loss of sales resulting from the republications in the Mainland under the rule of double actionability.  There are disputes as to what the Mainland law is and whether the rule of double actionability applies so as to prevent the Plaintiffs from recovering the loss of sales suffered in the Mainland.

539. The fifth issue is about the loss of sales and profits in Hong Kong.  It is common ground that the sales business of BaWang Shampoo Products in Hong Kong was conducted by the two subsidiaries of the 1^st Plaintiff in Hong Kong which are not parties to the proceedings.  Because of the no-reflective-loss principle, there is a dispute as to whether the existing Plaintiffs are able to recover the losses suffered by these two subsidiaries.

540. The sixth issue is how to assess the quantum of general damages.

541. The seventh issue is whether the facts of the present case justify an award of exemplary damages.

542. The last issue is a civil right issue.  Since the publication of the Article involved the right of the freedom of expression and the right of the public to know, the Defendant argues that the level of damages awarded in the present case should not be so disproportionately high resulting in the freedom of expression being unduly curtailed.  There is a dispute whether the assessment exercise should be influenced by such civil right consideration.

543. I will deal with these issues in turn.

(i)  The actual loss of sales and profits 

544. The Plaintiffs only claim for the loss of sales and profits relating to BaWang Shampoo Products (which included both the “BaWang” and “Royal Wind” brands shampoos) suffered by them in the period from 15 July 2010 (i.e. the date after the publication of the Article) to 30 June 2011 (“the Relevant Period”).  Hence, the Plaintiffs’ claim is only limited to a period of about 11½ months after the publication of the Article.

545. In this exercise, the court would try to assess the projected sales of BaWang Shampoo Products in the Relevant Period. Comparing that figure with the actual sales figure, the court can then work out the loss of profit in the Relevant Period.  Whether the Plaintiffs are able to recover all or any loss of profits because of conflict of law, causation, remoteness of damages or no-reflective-loss issues is a different matter which would be dealt with in the latter part of this Judgment.

546. Both experts have agreed that the loss of profits in respect of BaWang Shampoo Products in the Relevant Period can be calculated by reference to the following items of revenue and expenditure, but there is dispute as to how to arrive at the proper figure for each item. The respective calculation of the parties can be listed out as follows:

	Plaintiffs’ Expert RMB’000	Defendant’s Expert (Scenario A) RMB’000	Defendant’s Expert (Scenario B) RMB’000
Forecast revenue Forecast variable cost of sales	2,077,913 (678,162)	1,489,206 (628,303)	1,619,187 (683,102)
Forecast variable other operating expenses	(819,597)	(587,394)	(638,663)
Forecast profits (A)	580,154	273,509	297,422
Actual revenue Actual variable costs of sales Actual variable other operating expenses	606,165 (250,428) (239,091)	588,688 (250,083) (232,199)	588,688 (250,083) (232,199)
Actual profits (B)	116,646	106,406	106,406
1st Plaintiff’s lost profits before adjustments (A-B) Excessive sales returns Lost profits attributable to men’s series products	463,508 13,951 (28,952)	167,103 -   -	191,016 -   -
1st Plaintiff’s lost profits	448,507	167,103	191,016

547. The 1^st Plaintiff is the holding company of the BaWang Group.  The aforesaid calculation is made on the basis of the consolidated accounts of the BaWang Group.  In other words, those are the losses suffered by the BaWang Group as a whole.  Further, there are Scenarios A and B under the Defendant’s assessment.  Scenario A’s calculation is the one made by Mr Dick in his original report.  Scenario B’s calculation covers the scenario if the court were to adopt the growth rate of the general hair care products in Mainland during the Relevant Period in assessing the projected sales of BaWang Shampoo Products.  I will elaborate this point further in the latter part of this Judgment.

548. The first main dispute is the forecast revenue of BaWang Shampoo Products.  Mr Skalak arrives at the figure of RMB¥2,077,913,000, whereas Mr Dick works out a different figure of RMB¥1,489,206,000.

549. The actual sales figure of BaWang Shampoo Products in the immediate preceding 12 months before the publication of the Article (from 1 July 2009 to 30 June 2010) was RMB¥1,837,399,823.  As worked out by the experts themselves[233], the Plaintiffs’ estimate reflects a 18.1% annual increase in sales from the previous year, whereas the Defendant’s estimate indicates an annual drop of sales of 15.3%.

550. Mr Skalak works out the figure by relying on the forecast made by the management of the BaWang Group in May 2009 contained in a piece of paper (“the 2009 Forecast”), which the Plaintiffs say is the best piece of evidence because it was prepared before the publication of the Article.  Based on the 2009 Forecast, the implicit revenue growth rates in 2010 and 2011 were 35% and 20% respectively.[234]

551. On the other hand, Mr Dick complains that the 2009 Forecast did not provide a sufficiently reliable basis from which to derive an assessment of the forecast revenue during the Relevant Period.  The 2009 Forecast only compromised a single sheet of paper which was undated and contained no narrative, description or explanation of its contents.  It was not clear who prepared the 2009 Forecast or the purpose for the preparation of such Forecast. Further, there was no supporting documents or oral evidence from the officers who were involved in the budgeting purpose explaining the basis for arriving at such Forecast, or the assumptions or factors that were taken into account when the 2009 Forecast was prepared.  In such circumstances, it would be unsafe for the court to rely on such Forecast to assess the projected sales of BaWang Shampoo Products.

552. I agree with these criticisms.  Without any supporting evidence and material in explaining the basis for the calculation of the figures in the 2009 Forecast, the revenue forecast by the Plaintiffs was no more than a guess by them as to how the sales of BaWang Shampoo Products would perform in the future.  Further as pointed out by Mr Dick himself, one would expect a company, particularly a listed company like the 1^st Plaintiff, to continuously track its actual performance and to identify the reasons for deviations therefrom.  Yet no such information is available.  I am quite surprised that, bearing in mind the substantial amount of the claim, the Plaintiffs have not called any evidence from the management team of the Plaintiffs to explain the basis of their forecast, and hence it would not be safe for the court to rely on these figures without further supporting evidence.

553. The Plaintiffs seek to justify the accuracy of the 2009 Forecast by relying on the 2010 Annual Budget and the sales forecasts made by equity research analysts at Morgan Stanley[235], Bank of America Merrill Lynch and OSK in their reports.  Mr Skalak claims there these were reliable reports because they were published before the publication of the Article, and the forecasts contained in these reports matched with the figures contained in the 2009 Forecast.  Further, Mr Skalak suggests that the actual revenue in the 1^st half of the 2010 Annual Budget was approximately the same as that predicted in the 2009 Forecast.

554. On the other hand, there were also other equity analyst reports (one also by Morgan Stanley)[236] which suggested that the sales of the BaWang Shampoo Products in the 1^st half of 2010 (which was before the publication) fell short of their forecasts.  Though these reports were published after the publication of the Article, the analysts were discussing the sales performance of BaWang Shampoo Products before the publication, and they were of the view that the performance by that time was already unsatisfactory.  In any event, there was very little material to support how these forecasts were made by equity analysts.  The fact that these analysts could change their forecasts overnight also undermines the reliability of their forecasts.

555. Mr Skalak also seeks to argue that the actual sales result for the 1^st half of 2010 (as contained in the 2010 Annual Budget) approximates the 2009 Forecast.  However, since there was only one sales figure as the estimate for the whole year of 2010 in the 2009 Forecast and there is dispute between the experts as to how to split the sales between the 1^st and 2^nd halves of the year, the figures in the 2009 Forecast remain a bare speculation without any supporting basis. Finally, the 2010 Annual Budget was only for the year 2010 and did not contain any projection for 2011.  Given its limited coverage, it would not be safe for the court to rely on the 2010 Annual Budget as the yardstick to assess the projected sales of the Relevant Period.

556. As pointed out by Mr Dick, there are several factors which show that the figures in the 2009 Forecast were too aggressive and most likely overstated:[237]

(i)  BaWang’s business in the 1^st half of 2010 was slowing, which was a matter of concern for some equity analysts;

(ii)  the market for BaWang’s hair care products was becoming increasingly competitive;[238] and

(iii)  the prospects of the Royal Wind brand were uncertain and there was also danger of cannibalization[239] between the BaWang and Royal Wind brands.

557. According to the market data obtained by the experts, BaWang was facing tougher competition in the shampoo market.  First, the growth of the whole hair care products market in the Mainland had slowed down, with 30% growth in 2007, 17 % growth in 2009, but only 8% growth in 2009, and 9% growth in 2010 and 2011.[240] Second, the revenue generated by the BaWang’s established products grew by only 3% in 2009 and 5.7% in the 1^st half in 2010.[241]  Third, as admitted by Mr Skalak, BaWang had to rely on new products to maintain the growth of its sales, and this was why BaWang had introduced the Royal Wind shampoo.  However, the sales of the Royal Wind shampoo in the 1^st half of 2010 were unsatisfactory.  In the 1^st half of 2010, Royal Wind products experienced a substantial drop of around RMB¥ 90 million from the 2^nd half of 2009 to the 1^st half of 2010, which marks a decline of about 40%.[242]

558. According to Mr Dick, there are at least two reasons why Royal Wind brand faced difficulties.  Firstly, the Royal Wing brand competed in a less niche market than the BaWang brand.  Since Royal Wind brand competed in the anti-dandruff segment, it competed against some global giants such as P & G and Unilever.  Secondly, there would be cannibalisation between the two brands, i.e. BaWang and Royal Wind Brands.[243]

559. Mr Pow submits that the 2009 Forecast was prepared by the management team of BaWang who were indisputably in the best position to evaluate the trend of their own business.  Further, the 2009 Forecast was an internal document.  There was no motive for BaWang to exaggerate the figures in the Forecast.

560. Despite Mr Pow’s able submission, in the absence of further evidence to explain how the forecast was made, the 2009 Forecast remains a self-serving guess as to the future performance of BaWang Shampoo Products.  Further, there is no one from the management team to explain whether they had taken into account the negative market factors as outlined by Mr Dick in his reports and testimony.  In such circumstances, I do not accept that the 2009 Forecast provides a reliable basis to forecast the sales of BaWang Shampoos Products in the Relevant Period.

561. On the other hand, Mr Dick adopts a different methodology to assess the projected sales.  Mr Dick is of the view that the market share enjoyed by BaWang in 2010 would, but for the loss of consumer confidence, have been impacted in a similar fashion to those of the 2 comparable brands known as C-Bons and La Fang (“the 2 Comparables”).  He considers that the 2 Comparables are most similar to BaWang’s brands because they were manufacturers of domestic hair care products that were competing with the dominant hair care brands, and because of the comparable size of their market share prior to the publication of the Article.  Mr Dick is of the view that BaWang’s growth or reduction of the market would be the same as the average of the growth or reduction of the 2 Comparables.  Using such methodology, Mr Dick works out a figure of RMB¥1,489,206,000 which represents a 15.3% drop in projected sales.

562. I also have difficulty with Mr Dick’s projection. Though the sales growth of BaWang Shampoo Products might have slowed down before the publication of the Article, it is difficult to explain the reason why, where there had been an annual growth in the overall hair care products at the relevant time, there would have been a drop of 15.3% in sales had not Article not been published.  At the trial, no satisfactory explanation has been given by Mr Dick.

563. As I see it, there are 2 major flaws in the methodology adopted by him.  In working out the projected sales figures from 15 July 2010 to 30 June 2011, Mr Dick tries to work out the sales figures of the 2^nd half of 2010 and 1^st half of 2011 separately.  In his calculation, Mr Dick has used the sales figures in the previous year in working out a split ratio of the sales in the 1^st and 2^nd halves of the year, i.e. 39% and 61%.  He assumes that the same ratio would be applicable in the following year (which would be a notional percentage) and he then uses the real sales figure in the 1^st half of 2010 to work out the projected sales.  In using the notional figures and actual figures together, it bounds to produce an unrealistic result which shows a significant drop in sales despite the general growing trend (though at a slower rate) in the market in the Relevant Period.

564. One should be careful about the use of the split ratio between the 1^st and 2^nd halves of the year in assessing the projected sales.  There is simply no established pattern of sales between the two halves of the calendar year, and so I reject any methodology using the split ratio in assessing the projected sales of BaWang Shampoo Products.

565. The second flaw lies with the use of comparables.  Mr Dick insists to use actual market data to assess the forecast revenue of BaWang Shampoo Products.  In doing so, Mr Dick uses the data of comparable companies and products in working out the projected sales.  However, I agree with the Plaintiffs that the 2 Comparables chosen by Mr Dick are not appropriate because they do not reflect the Plaintiffs’ position as the leader in the herbal shampoo market.  Further, Mr Dick uses the post-publication data of the 2 Comparables in the calculation.  The whole market condition might have changed after the publication of the Article, and the publication might have affected the sales of the 2 Comparables as well.  As the court is trying to project the sales of BaWang Shampoo Products but for the publication, it would not be appropriate to use these post-publication figures in the calculation.  I therefore also reject the Defendant’s methodology.

566. As there is serious doubt about the use of the 2 Comparables, Mr Dick, in the course of his testimony, also puts forward an alternative calculation known as the Scenario B’s calculation.  In Scenario B, Mr Dick assumes that BaWang could have maintained the share of the Mainland shampoo market that it achieved in 2009.  The growth rate for the general hair care products market in the Mainland was 9% in 2010 and 2011.  Mr Dicks assume that BaWang’s revenue would have grown at the same rate as the general hair care products market as a whole in 2010 and 2011, namely 9% per annum on a calendar year basis.  Based on that assumption, he works out the forecast revenue to be RMB¥1,619,187,000.[244]

567. Again I reject the Scenario B’s calculation.  Mr Dick still uses the split ratio between the 1^st and 2^nd halves of the year in assessing the projected sales in the Relevant Period.  As mentioned above, the split ratio approach should not be adopted, and the absurdity of the Scenario B’s calculation is shown by the fact that despite the general growing trend in the sales in the general market, the final figure worked out by Mr Dick (i.e. RMB¥1,619,187,000) is still less than the actual sales figure from 1 July 2009 to 31 June 2010 (RMB¥1,837,399,823 x 11.5/12 = RMB¥1,760,841,496).  Hence, I also reject the Scenario B calculation made by Mr Dick.

568. How should the court then assess the projected sales figures in such circumstances?  As in many difficult cases concerning assessment of damages, the court has to try its best to assess the quantum based on the available evidence and data.  In my judgment, the best approach is to try to work out the notional annual growth rate of BaWang Shampoo Products at the Relevant Period.  After obtaining the notional growth rate, the court can then apply the growth rate to the actual sales figures of the immediate 12 months of BaWang Shampoo Products prior to the publication of the Article to work out the projected sales figures.

569. There is no serious dispute that the annual growth rate of the hair care products in the Mainland market in the Relevant Period was about 9%.  I decide to use the growth rate of the general hair care products market instead of the specific herbal shampoo products market, because one of the brands of BaWang Shampoo Products, Royal Wind, had to compete in the more competitive general market and not the specific market of herbal shampoo. Further, the market data shows that the growth rate of the original BaWang brand (which competed in herbal shampoo market) had slowed down in the years before the publication of the Article.  Hence, I adopt 9% as the notional annual growth rate.

570. The Relevant Period consisted of 11½ months and not the full 12 months.  As the parties cannot establish a specific pattern for the sales in a particular period of the year, I just reduce the projected sales figures on a pro gratia basis.  Hence, I assess the projected sales revenue of BaWang Shampoo Products in the Relevant Period as follows:

RMB¥1,837,399,823 x 11.5/12 x 109% = RMB¥1,919,317,230

571. In his evidence, Mr Dick tells me that this methodology does not work.  9%, which relates to a calendar year growth, can only be applied to calendar year profit or revenue.  It cannot be applied to profit or revenue which straddles 2 calendar years because that would be applying apples to oranges.  If 9% were to be applied to the figure of RMB¥1,837,000,000 as mentioned above, annual growth would come to 18% which plainly exceeds the growth that was in fact 9%.

572. I disagree.  I fully understand that 9% was the annual calendar year growth, and so it may not match the calculation of the annual growth in the Relevant Period which was from 15 July 2010 to 30 June 2011.  Yet this would be the best figure for the assessment of the annual growth in the period.  The actual sales of BaWang Shampoo Products from 1 July 2009 to 30 June 2010 was RMB¥1,837,399,823.  If there was a general growth in the region of 9% in the Relevant Period, then the figure of RMB¥1,919,317,230 makes much more sense to me than the figure of RMB¥1,619,186,000 worked out by Mr Dick under Scenario B, which marks a decline in the sales despite a general growing trend of 9% at the Relevant Period.

573. Then I come to the forecast variable costs of sales, which would be calculated by applying a specific percentage to the projected sales figures.  Mr Skalak uses a percentage of 32.6% based on the cost structures and relevant cost to revenue ratio in the 1^st half of 2010.  On the other hand, Mr Dick adopts a percentage of 42.2% by looking at the costs actually experienced by BaWang during the 2^nd half of 2010 which was after the publication of the Article.

574. Mr Dick uses post-publication figures rather than pre-publication figures because of indications that BaWang’s manufacturing and packaging costs had increased during the Relevant Period, and the increasingly competitive environment might have made it difficult to pass those cost increases onto consumers.  Mr Dick also backs up his observations by some market data, for example cost indexes of the various raw materials required for the production of BaWang Shampoo Products and the rising labour costs during the Relevant Period.  Apart from these factors, decrease in sales due to the change in sales channel mix and increase in bundling sales might also have an impact on the gross margin generated by BaWang during the Relevant Period.

575. On the other hand, the Plaintiffs argue that the cost ratio in the 2^nd half of 2010 was tainted by the effect of the Article.  After the publication of the Article, the Plaintiffs might have to reduce the price of their products or to engage in other kinds of promotion activities thereby lowering their profit margin.  Further, their competitors might seize the opportunity created by the 1,4-dioxane incident to capture the market share of BaWang.

576. There are some justifications in these arguments. It may not be appropriate to rely solely on the post-publication cost structure because the publication of the Article might have certain effect on the cost structure of BaWang Shampoo Products.  Further I accept Mr Skalak’s evidence that, had the publication not been made, the Plaintiffs would have been in a better position to absorb the costs increase by other measures.  In fact, some of the fixed term agreements made by the Plaintiffs for the supply of packaging material prior to the publication of the Article might be able to absorb part of the increasing costs faced by BaWang.[245]

577. Despite that, the market data supplied by Mr Dick supports the Defendant’s case that the variable costs of sales of BaWang Shampoo Products were likely to have increased during the Relevant Period.  Furthermore, the cost ratio relied on by the Plaintiffs was actually the lowest one in recent years.  In fact, the cost ratio was as high as 36.2% in the 1^st half of 2008.  In view of the rising labour, raw materials and packaging costs in the Mainland, I have serious doubt as to whether the Plaintiffs were able to maintain the same cost ratio even without the publication of the Article.

578. Balancing all these factors, I adopt the median cost ratio of 37.4% ((32.6% + 42.2%) ÷ 2).  Hence, the forecast variable costs of sales can be calculated as follows:

RMB¥1,919,317,230 x 37.4% = RMB¥717,824,644

579. There is no dispute between the experts that the court should adopt the percentage of 39.44% of the revenue in calculating the forecast variable other operating expenses.  As the forecast revenue is estimated to be RMB¥1,919,317,230, the amount of forecast variable other operating expenses is:

RMB¥1,919,317,230 x 39.44% = RMB¥756,978,716  

580. For actual variable costs of sales, the difference between the parties is small.  The Plaintiffs and the Defendant adopt the respective percentages of 41.3% and 42.5%.  As the parties have not argued on the difference, I adopt the median percentage of 41.9%.

581. Again, there is no dispute between the parties in using the same 39.44% in calculating the actual variable other operating costs.

582. So far as the calculation of the actual profits is concerned, another dispute between the parties is to how to handle the effect of two other incidents after the publication of the Article which might have a negative impact on the sales of the Plaintiffs’ products.  First, there was a report on 16 July 2010 in the National Business Daily about missing and expired SFDA registrations in respect of some of the products of the Plaintiffs.  Second, there was another report on 10 September 2010 in the Ming Pao Daily about mislabelling in respect of 2 Men’s Series products of the Plaintiffs.  The Plaintiffs also make some adjustment in respect of the excessive sales return as a result of a rebranding campaign.

583. In a nutshell, Mr Skalak increases the actual revenue by adding back some sales returns that had been deducted, both in respect of the Men’s Series products due to the mislabelling incident and also in respect of what the Plaintiffs allege are “excessive” sales returns during the rebranding campaign.  On the other hand, Mr Dick does not make any corresponding adjustment in respect of sales returns of the Men’s Series products because: (i) it is not possible to separately project the forecast revenue attributable to that particular line of products; and (ii) the impact of the mislabelling incident was not restricted to the Men’s Series products and would also have impacted the overall consumers’ confidence in the BaWang’s products.

584. In respect of the alleged “excessive” sales returns, Mr Skalak has based his view that such returns were excessive because that was what had been advised by BaWang’s management.  However, he admits at the trial that he is not in a position to offer an opinion on this matter. Significantly, there is a complete absence of contemporaneous evidence confirming that the alleged “excessive” sales returns, which took place one year after the publication of the Article, were attributable to the said publication.  In such circumstances, I am not prepared to make the adjustment in respect of the “excessive” sales returns.

585. I also agree with Mr Dick that the mislabelling incident would have an impact on the overall consumers’ confidence in BaWang’s Shampoo Products and so it is not appropriate to make the simple adjustment in the manner as alleged by Mr Skalak. 

586. For the above reasons, the 1^st Plaintiff’s loss of profits can be assessed as follows:

Forecast revenue	¥1,919,317,230
Forecast variable cost of sales	(¥717,824,644)
Forecast variable other operating expenses	(¥756,978,716)
Forecast profits (A)	¥444,513,870
Actual revenue	¥588,688,000
Actual variable costs of sales	(¥246,660,272)[246]
Actual variable other operating expenses	(¥232,178,547)[247]
Actual profits (B)	¥109,849,181
1st Plaintiff’s Lost Profits (A-B)	¥334,664,689

587. It is common ground that such loss of profits is calculated on a consolidated basis, i.e. at the level of the 1^st Plaintiff as the holding company of the BaWang Group.  After ascertaining the loss at that level, the parties then go on to ascertain the loss of profits suffered specifically by the 2^nd Plaintiff.

588. The entities within the BaWang Group that were engaged in the sales of BaWang Shampoo Products at the time of the publication of the Article were:

(i)  the 2^nd Plaintiff;

(ii)  BaWang International Group Holding (HK) Limited (“HK1”); and

(iii)  Hong Kong BaWang International Trading Limited (“HK2”).

589. The 2^nd Plaintiff was responsible for the sales of the BaWang Shampoo Products in the Mainland market, whilst HK1 and HK2 were responsible for the sales in Hong Kong and other overseas markets. The 2^nd Plaintiff was the main operating subsidiary of the BaWang Group.

590. In trying to work out the loss of profits suffered specifically by the 2^nd Plaintiff, the experts agree to calculate the loss of profits suffered by HK1 and HK2 and then subtract that figure from the overall loss of profits as worked out earlier.

591. Different calculations of the loss of profits of HK1 and HK2 are set out below:[248]

	Plaintiffs’ Expert RMB’000	Defendant’s Expert RMB’000
Forecast revenue	104,709	73,270
Forecast variable cost of sales	(37,139)	(28,928)
Forecast variable other operating expenses	(26,439)	(18,500)
Forecast profits (A)	41,131	25,842
Actual revenue	21,313	21,313
Actual variable costs of sales	(8,731)	(8.731)
Actual variable other operating expenses	(5,381)	(5,381)
Actual profits (B)	7,201	7,201
HK1 & HK2’s lost profits before adjustments (A-B)	33,930	18,641
Lost profits attributable to men’s series products	(737)	-
HK1 & HK2’slost profits	33,193	18,641

592. Both experts agree that it is appropriate to apply the ratio of the actual revenue of HK1 and HK2 to actual revenue of the BaWang Group for the affected products in the 1^st half of 2010 to arrive at the forecast revenue for HK1 and HK2 during the Relevant Period, but disagree on the methodology in calculating such ratio.[249]  Mr Skalak has also set out his calculation in some detail in appendix 4 of his 1^st report dated 6 March 2012.

593. As mentioned above, I do not accept the methodology of both experts in working out the loss of profits of the whole BaWang Group, in particular as to projection of the forecast revenue of BaWang Shampoo Products in the Relevant Period.  As a result, the court estimates the projected sales revenue by adopting a notional growth rate of 9% which was the growth rate of the Mainland general hair care products market at the Relevant Period.

594. Because of such change, the calculation of the loss of profits of HK1 and HK2 has to be adjusted as well.  Both experts would adopt a percentage of the forecast sales of the whole BaWang Group to assess the projected revenues of HK1 and HK2.  The forecast revenue of HK1 and HK2 assessed by Mr Skalak (i.e. ¥104,709,000) is about 5.04% of the forecast revenue of the whole BaWang group at the Relevant Period (i.e. ¥2,077,913,000). On the other hand, the forecast revenue worked out by Mr Dick (i.e. ¥73,270,000) is about 4.92% of the forecast revenue of the whole BaWang Group at the Relevant Period (i.e. ¥1,489,206,000).  Both percentages are in the region of 5% and so I use such percentage to assess the forecast revenue of HK1 and HK2 at the Relevant Period: ¥1,919,317,230 x 5% = ¥95,965,861.

595. Then I come to forecast variable cost of sales. Mr Skalak estimates the forecast variable cost of sales based on the historical cost of sales as percentages of the corresponding revenues in the 1^st half of 2010 (i.e. 35.47%).  Mr Dick uses the actual ratio of variable cost of sales for HK1 and HK2 to actual revenues for HK1 and HK2 during the Relevant Period (i.e. 39.48%).  I adopt the median percentage of 37.475%.  Hence, the forecast variable cost of sales is: ¥95,965,861 x 37.475% = ¥35,963,206.

596. The remaining parts of the calculation are not in dispute, and the parties also agree to use the percentage of 25.2504% to calculate the forecast variable other operating expenses.  Hence the loss of profits suffered by HK1 and HK2 can be assessed as follows:

Forecast revenue	¥95,965,861
Forecast variable cost of sales	(¥35,963,206)
Forecast variable other operating expenses	(¥24,231,764)	[250]
Forecast profits (A)	¥35,770,891
Actual revenue	¥21,313,000
Actual variable costs of sales	(¥8,731,000)
Actual variable other operating expenses	(¥5,381,000)
Actual profits (B)	¥7,201,000
HK1 & HK2’s lost profits (A-B)	¥28,569,891

597. The loss of profits suffered by the 2^nd Plaintiff relates to the reduction of sales of BaWang Shampoo Products in the Mainland market.  Such loss can therefore be assessed as follows:

Loss of profits of the 1st Plaintiff at the group level	¥334,664,689
Loss of profits of HK1 and HK2	(¥28,569,891)
2nd Plaintiff’s lostprofits	¥306,094,798

598. In assessing the quantum of the Plaintiffs’ claim, it may be relevant to look at the loss of profits resulting from the reduction of sales of BaWang Shampoo Products only in Hong Kong.  There is no dispute that HK1 and HK2 conducted sales other than in Hong Kong, and so the loss of profits suffered by HK1 and HK2 is not exactly the same as the loss of profits for the sales in Hong Kong.

599. Both experts agree to apply the same ratio of 71.1253% to the forecast revenue of HK1 and HK2 to arrive at the forecast revenue sustained in Hong Kong.[251] The experts apply different percentages based on different costs structure to assess the forecast variable cost of sales attributable to the Hong Kong revenue, about 28.4% in the case of the Plaintiffs and 39.87% in the case of the Defendant.[252]  For the same reasons given earlier, I adopt the median percentage of 34.135%.  Both experts agree to use the same percentage of 35.3% in assessing the forecast variable other operating expenses.  For the remaining items, there is no serious dispute between the parties and the only slight differences are due to rounding differences.

600. Hence, the loss of profits suffered in Hong Kong can be assessed as follows:

Forecast revenue	¥68,256,006	[253]
Forecast variable cost of sales	(¥23,299,188)	[254]
Forecast variable other operating expenses	(¥24,094,370)	[255]
Forecast profits (A)	¥20,862,448
Actual revenue	¥ 9,355,000
Actual variable costs of sales	(¥4,439,000)
Actual variable other operating expenses	(¥3,303,000)
Actual profits (B)	¥1,613,000
Lost profits in Hong Kong (A-B)	¥19,249,448

601. In conclusion, I assess the loss of profits suffered by the BaWang Group during the Relevant Period at different levels as follows:

Loss suffered at the group level	¥334,664,689
Loss suffered by the 2nd Plaintiff	¥306,094,798
for the reduction in sales in the Mainland
Loss suffered by HK1 and HK2 for the reduction in sales in Hong Kong	¥19,249,448

602. Despite this elaborate exercise, the Plaintiffs, for the reasons given in the later part of this Judgment, are not able to recover any loss of profits allegedly suffered by them during the Relevant Period.  Nevertheless, I feel obliged to conduct such exercise in the event that this case is going elsewhere and a contrary view is taken against my other rulings on quantum as mentioned below.

(ii)  Juridical basis in claiming for loss resulting from republications

603. As mentioned above, the bulk of the Plaintiff’s claim relates to the loss of sales caused by the republications by third parties in the Mainland.  One thing unique about the tort of defamation is that it is also a tort to repeat the defamation by some other person.  In other words, replication of the original defamatory article is a separate tort and so there is an issue as to whether there is juridical basis for the claimant to recover from the original publisher damages caused by the replication of the original defamatory article by others.

604. Mr Yu invites the court to hold that it is not permissible for the Plaintiffs to rely on the republications in the Mainland not as causes of action but only for the purpose of recovery of damages.

605. As agreed by Mr Yu, there have been a line of authorities to support that a plaintiff can, as an alternative to suing on a republication as a separate cause of action, treat the further publication as increasing the damage flowing from the first.  In Gatley, the learned authors said:[256]

“Where a defendant’s defamatory statement is voluntarily republished by the person to whom he published it or by some other person, the question arises whether the defendant is liable for the damage caused by that further publication. In such case the claimant may have a choice: he may (1) sue the defendant both for the original publication and for the republication as two separate causes of action, or (2) sue the defendant in respect of the original publication the damage which he has suffered by reason of its repetition, so long as such damage is not too remote.”

606. In Slipper v BBC[257], the defendant showed television film defamatory of the plaintiff and the plaintiff was allowed to rely on the effect of the newspaper reviews of the film as damage flowing from the broadcast of the film.  The film reviews, being further publications though not republications, were said to have been foreseeable, natural, provable and perhaps even intentional results of the publication sued upon, and were held to be a legitimate claim.  In McManus v Beckham[258], the plaintiff was able to rely upon republications not as separate causes of action, but as part of the damage which flowed from the original publication as the republications had been foreseesable.  In Tang Chiu Yuk v Hung Ka Cheun[259], in awarding damages, Chung J took into account the fact that there had been extensive republication of the libel in quite a number of local newspapers as the republications were the natural and probable results of the original publication.

607. However, Mr Yu submits that the said authorities did not address the juridical basis of the proposition that a plaintiff can recover damages flowing from a publication in respect of which he cannot establish primary liability on the part of the defendant.  Such argument originated from a submission made by the counsel in Baturina v Times Newspapers Ltd which was endorsed Eady J in the first instance judgment[260].

608. The submission was summarised in §52 of the judgment of Eady J in Baturina:

“[The defendant’s counsel] queries the juridical basis of the proposition that a claimant can recover damages flowing from a publication in respect of which he could not establish primary liability on the part of the defendant. It is difficult to reconcile these two propositions as a matter of causation. Attention is drawn by the learned editors of Gatley … … to observations made in the Queensland case of Timms v Clift [1998] 2 Qd R 100 Qd CA, at 108:

“If, as was held, the plaintiff had no right to recover in respect of the cause of action for the second publication, that in the magazine, it is difficult to understand why cause of action in respect of first publication should have included a right to damages in respect of the second publication.”

609. In my judgment, the lack of juridical basis argument only supports the proposition that if the claimant has no right to bring a cause of action for the second publication, either because there is a defence available for the second publication or some other reasons, the claimant cannot sue the defendant in respect of the original publication for the damages resulting from the second publication.  Such observation is actually supported by the following passage in Gatley:[261]

“ … … this might seem to suggest damages can be recovered for the republication regardless of whether the later publication would be actionable by the claimant. Thus, even if the publisher of the later publication would have had a defence, or had been released as a joint tortfeasor, in respect of that publication, damages may still be recovered by the claimant for the republication. As Eady J. pointed out in Baturia v Times Newspapers, the juridical basis of the proposition that a claimant can recover damages flowing from a publication in respect of which he could not establish primary liability on the part of the defendant is difficult to ascertain. It is submitted therefore that the correct view is that where no claim would lie against the defendant in respect of the later publication, the claimant should not be as a matter of principle be allowed to recover damages in respect of that publication. If the later publication is not actionable then, even if it was caused by the original publication, it would be unjust to make the defendant liable for any harm caused by that publication. If that is right, then regardless of whether a claimant relies on a republication as a cause of action or in aggravation of damages, a defendant would be entitled to meet the claim in respect of that publication with any relevant defence. That is not to say however that the original statement must be repeated word for word in the republication. Provided a media report of the initial publication convey the sting of the original, in whole or in part, it may be relied on to increase the damages flowing from the initial publication even if it cannot be said to ‘repeat’ what was then said.”

610. This is quite different from the proposition that, in case the claimant has not brought a separate cause of action in respect of the republication or further republication, he should not be allowed to claim against the original publisher for the damages resulting from the later publication if such damages are foreseeable.

611. Apart from the requirement to establish actionability in respect of the subsequent republication, there are also other issues involved if the claimant seeks to recover from the original publisher damages caused by republications.  The case of Baturina v Times Newspapers Ltd[262] highlights some of these issues.

612. In Baturina, the claimant, a Russian businesswoman and wife of the former Mayor of Moscow, brought a libel claim against the defendant, Times Newspaper, for its publication of an article which stated that she had purchased a large mansion in Highgate for £50 million through an offshore company based in BVI and that she was planning to spend a further £50 million renovating and extending the property.  The claimant denied purchasing any such property and claimed that the articles defamed her in the eyes of readers who were aware of her declaration of assets made a few months previously in accordance with Russian anti-corruption legislation.  Her declaration had not included this property.  The claimant therefore relied as part of her claim upon an innuendo meaning that the claimant failed to declare her ownership of and interest pursuant to Russian law.  The readers who had such knowledge would be in Russia.  The story was widely repeated and embellished in the Russian media.

613. There were 5 categories of publications which might cause damage to the reputation of the claimant:

(i)  the hard copies of the newspaper circulated in England;

(ii)  the hard copies of the newspaper circulated in Russia;

(iii)  publications by third parties in Russia which took the matter a stage further by making direct allegation of illegality against the claimant;

(iv)  publications by third parties in Russia which replicated the defendant’s words (or some of them) without putting an interpretation or gloss upon them; and

(v)  possible internet publications of the article in the defendant’s Times Online website.

614. In the application to strike out the claimant’s claim, Eady J struck out the claim in respect of the first category of publications as, in the absence of evidence as to specific readers with the relevant knowledge, there was no realistic prospect of establishing the relevant innuendo for such class of publications.  For the second type of publications, the judge allowed the claim to proceed because it would be a matter for the trial to determine whether the readers in Russia would have had knowledge of the facts alleged in support of the pleaded innuendo meaning.  Nonetheless, it would be necessary for the claimant to establish that such publications were actionable by Russian law, in particular with reference to the Russian law on innuendo.  For the fifth category, Eady J struck out the claim because the court was not prepared to assume that there would have been any internet readers, whether in England or in Russia, who would have had access to or downloaded the relevant material and who possessed knowledge of the extraneous facts relied on in support of the innuendo meaning.

615. It is the third and fourth categories of publications which are relevant here.  Eady J struck out the third category’s claim.  The judge held that the defendant should not be held liable in respect of allegations in publications produced by third party publishers which went beyond the allegations made in the Sunday Times article and which overtly defamed the claimant.  Each such publication would fall within the definition of novus actus interveniens.  The article in Sunday Times might well have been a cause, or indeed the primary cause, of such subsequent publications, but that was not sufficient to establish liability against the defendant for those publications.  There was no evidence that the defendant would have reason to know of such risk at the time of the publication of the original article.  Regarding the fourth category, i.e. third party publications in Russia that merely repeated the defendant’s allegation that the claimant had purchased or acquired an interest in the property, the claim should be allowed provided that it was properly pleaded.

616. Furthermore, since the defendant was not directly responsible for such third party publications, to establish liability against the defendant, the claimant would also need to establish that any such republication was a reasonably foreseeable consequence of the defendant’s initial publication.  This was so irrespective of where the alleged republication took place, whether it was in the local or some other foreign jurisdictions. The only additional requirement for the republication in the foreign jurisdiction would be that of establishing double actionability.

617. The case went on appeal.  The English Court of Appeal allowed the claimant’s appeal and permitted the claimant to proceed with the claims relating to the first and fifth categories of publications on the basis that such claims were not wholly unarguable.  It also allowed the defendant’s appeal in part with the effect that the claimant had to identify specific readers in each class of readers who, on reading the article, had inferred the innuendo meaning alleged.  The appellate judgment is therefore not relevant for our purposes.

618. Hence, if a claimant seeks to hold the original publisher liable for damages resulting from further republications by third parties, the following issues become relevant:

(i)  the damage flowing from the republications must be foreseeable and not too remote;

(ii)  whether there is any material difference in the contents between the original publication and the republication so that the latter may become a novus actus interveniens (which is very much a causation issue);

(iii)  whether there is a defence available to the third party in respect of the republication; and

(iv)  in the case that the republications took place in foreign jurisdiction, whether the claimant has established the requirement of double actionability.

(iii)  Causation and remoteness of damages

619. I first deal with the questions of causation and remoteness of damages.  The problem here is that the republications were made by third parties and the contents of the republications were very different from those contained in the original publication, and so it poses serious questions as to whether the further damages caused by the republications by third parties were caused by the original publication and whether the further damages were within the reasonable foreseeability of the original publisher.

620. In Gatley[263], the learned authors said the following on the issues of causation and remoteness of damages:

“It is clear that if the republication was by a person for whom the defendant is vicariously liable under the principles of master and servant or agency then the defendant is liable for that. Beyond this it may be that the original publisher should only be liable as a publisher of the republished statement where he authorised or intended it, but some of the cases speak in broader terms and assume that the same principles relating to the responsibility of the defendant apply to both situations. In any event,

‘the law would part company with realities of life if it held that the damages caused by publication of a libel began and ended with publication to the original publishee. Defamatory statements are objectionable not least because of their propensity to percolate through underground channels and contaminate hidden springs.’

The question is essentially the same as that in any other tort case where it is sought to make the defendant liable for harm which is directly attributable to the voluntary act of a third person.  That is a question of causation but it is not a pure question of fact, nor is the inquiry a value-free one: the ‘reality is that the court has to decide whether, on the facts before it, it is just to hold [the defendant] responsible for the loss in question.’  No doubt it is still true that the starting point is that the defendant is prima facie not liable for the further damage, because it is incumbent on the claimant to show that there is an adequate causative link between the tort and the damage, but subject to that, the defendant will be liable if he is actually aware that what he says or does is likely to be reported or if a reasonable person in his position should have appreciated that there was a significant risk that what he said would be repeated in whole or in part and that that would increase the damage caused by what he said.  For reasons of clarity it is better not to direct the jury in terms of reasonable foreseeability, though as a matter of principle that is the underlying concept.”

621. It is the Plaintiffs’ case that there is a strong causative link between the original publication and the republications in the Mainland.  After the original publication of the Article, the Words, or their gist, were widely repeated and republished by the Hong Kong local newspapers and the internet publishers in the Mainland.  Unlike the Magazine, a number of these Hong Kong newspapers were listed as permissible publications which could be purchased by the public in the Mainland.  According to the Chief Financial Controller of the BaWang Group, Mr Wong Sin Yung (“Mr Wong”), there were at least 83 articles by the Hong Kong newspapers repeating or republishing the Words, or their gist, which were made available to the public in the Mainland through the circulation of those newspapers and the internet.  Furthermore, the Words or their gist were repeated or republished in at least over 140 Mainland newspapers and internet sites, which according to Mr Wong, a substantial number of which had national circulation.  At least 43 of those Mainland republications made references to the reports from the Hong Kong media or the Magazine as the source of the report.

622. However, the Plaintiffs’ claim is not free from difficulty.  As mentioned above, the bulk of the Plaintiffs’ claim relates to the loss of sales of BaWang Shampoo Products in the Mainland.  The court has no information about the circulation of the “permissible Hong Kong publications” in the Mainland, which should have been rather limited.  In support of the claim, the Plaintiffs therefore mainly rely on the effect caused by the republications in the Mainland which consist of 147 articles included in various places from items 165 - 421 of Bundle C of the trial bundle (“the Republications”).  At the trial, the parties have submitted a comprehensive table listing out the contents of the Republications.  In trying to keep this Judgment to a reasonable length, I do not intend to include this lengthy table in this Judgment.

623. After reading the original texts of the Republications, I can make the following observations.

624. Firstly, most of the Republications were in the form of reportage and did not repeat the gist of any of the imputations against the Plaintiffs, rather most of the Republications only referred to the fact that BaWang Shampoo Products were reported to contain 1,4-dioxane which was a carcinogenic substance (致癌物).  A lot of the articles discussed the social impact of such incident, the regulation of the presence of such substance in consumer products in the Mainland, the effect of the incident on the confidence of the consumers in BaWang Shampoo Products, the consumers’ queries as to why BaWang Shampoo Products (advertised as natural products) did contain carcinogenic chemical, and the effect of the incident on the share price of the BaWang Group.

625. Secondly, a lot of the articles contained similar information about the chemical known as 1,4-dioxane.[264]  Most of the articles referred the substance as “微毒” (mildly toxic), a term which had not been used in the original Article.  It was clear that such information was not copied from the original Article, and the authors of these Republications did their own researches and included the information about 1,4-dioxane from some other sources.  In a few articles, the authors actually conducted their own interviews with other experts in the Mainland about the issue.

626. Thirdly, some of the articles referred to poll results.[265] The public was asked as to whether they believed BaWang Shampoo Products did contain carcinogenic substance, or whether they would continue to buy the Products.  They also contained reports that consumers might not buy the Products as a matter of caution, irrespective of whether they believed that the Products did contain carcinogenic substance or the Products did pose a health risk to humans.

627. Fourthly, most articles included BaWang’s explanation about the presence of 1,4-dioxane in their shampoos.  Some reports referred to the statements and the stance of the Mainland SFDA confirming that BaWang Shampoo Products were safe.

628. Fifthly, there were discussions in a few articles which may be regarded as imputations to the Plaintiffs.  However, these articles did not focus on the presence of 1,4-dioxane in BaWang Shampoo Products.  Instead, the criticisms mainly related to: (i) the alleged improper advertising by BaWang describing BaWang Shampoo Products as wholly natural[266]; (b) the mishandling of the incident by BaWang[267]; or (c) other issues such as report of the crisis facing privately-owned enterprises in the Mainland[268]. Some of the articles compared the difference between the present incident with the previous incidents about the discovery of 1,4-dioxane in Johnson & Johnson’s products and toxic chemicals in the dairy products in the Mainland.  The articles in Items 205, 372, 373 and 398 were actually in favour of BaWang, which claimed that BaWang was the victim of another scandal created by someone with ulterior motives.

629. Sixthly, there were a few references to BaWang Shampoo Products being unsafe.  The article in Item 170b stated that the level of 1,4-dixcane in BaWang Shampoo Products was near the rim of danger.  The article in Item 178 mentioned that long term use of BaWang Shampoo Products might cause serious damage to human health.  The articles in Items 185, 349 and 391 referred to BaWang Shampoo Products as “毒薬” (poison) or “有毒” (containing poison).  The articles in Items 198 and 207 referred to the test results as contained in the original Article. 

630. Eady J in Baturina v Times Newspapers Ltd[269] quoted a passage in Gatley[270] and said that a plaintiff should only be allowed to recover as a consequence of the original publication the damage which they have suffered by reason of its repetition, if the republication conveys the sting of the original publication, either in whole or in part, otherwise it would be unjust to make the defendant liable for harm caused by the republication.

631. I agree with the Defendant’s case that the stings of the defamation as imputation to the Plaintiffs as corporate entities, for example the Plaintiffs were corporations lacking commercial morality and the Plaintiffs used inferior raw materials in the production of BaWang Shampoo Products with a mercenary motive to save costs, did not appear in the Republications. Most of the Republications only stated that carcinogenic substance was found in BaWang Shampoo Products.  This fact, by itself, was true because 1,4-dioxane had been classified as a possible human carcinogen.  In fact, the authors of most of the Republications had conducted their own researches about the possible harmful effect of 1,4-dioxane.

632. Further, the court is not familiar with the market condition and consumers’ behaviour in the Mainland.  It was not certain why the Mainland consumers suddenly stopped buying BaWang Shampoo Products. They did not have access to the original Article and so they learnt about the 1,4-dioxane incident from other sources including perhaps the Republications or some other not so formal sources.  As mentioned in some of the reports in the Republications, some of the consumers did not buy BaWang Shampoo Products as a matter of caution, probably due to the presence of a possible carcinogenic substance known as 1,4-dioxane in the Products irrespective of whether the level of such substance did pose a health risk to humans.  In other words, the loss of sales might not have been caused by the replication of the defamatory stings in the original Article, but rather simply by their concern about the presence of 1,4-dioxane in the shampoos.

633. As mentioned by Eady J in Baturina, as the Republications did not repeat the defamatory stings in the original Article, it would be unjust to make the Defendant liable for the additional damages caused by the Republications.  This is like the 3^rd category of publications in Baturina.  Though the original Article might well have been a cause, or indeed the primary cause, of the Republications, in the absence of the repetition of the defamatory stings, it would not be sufficient for the Plaintiffs to establish liability on the part of the Defendant for the damages caused by the Republications.

634. Mr Pow has also relied on market reports[271] to show that the market share of BaWang Shampoo Products had dropped significantly after the publication of the Article.  At the same time, the Plaintiffs’ other competitors, including Rejoice which was another shampoo named in the Article as containing 1,4-dioxane, were able to maintain relatively stable market shares from 2006 to 2011.  According to Mr Pow, if Mainland consumers stopped buying BaWang Shampoo Products only because of their concern about the presence of 1,4-dioxane in shampoo, they would have stopped buying Rejoice shampoos as well.  Hence, the drop in sales was caused by the defamatory stings contained in the original Article as replicated in the Republications.

635. I do not accept that such argument can advance the Plaintiffs’ case any further.  As mentioned above, the court has very little information about the market condition and consumers’ behaviour in the Mainland.  There may be a lot of reasons to explain the consumers’ behaviour in the Mainland.  It may be the case that consumers may have more confidence in brands with a longer track record, and as a result, the news that shampoos containing 1,4-dioxane might have a greater impact on the Plaintiffs’ products than shampoos of other brands.  The most important point here is that the defamatory stings as contained in the original Article had been replicated in at least most of the Republications, and so the Plaintiffs have failed to discharge the burden of proving that all or any of the loss of sales in the Mainland was caused by the public reading the defamatory stings allegedly replicated in the Republications, which I do not accept it to be the case.

636. It is arguable that some of the defamatory stings were replicated in a few specific Republications.  Some of the Republications did refer to the test results contained in the original Article, and a few Republications did mention that long term use of BaWang Shampoo Products might cause serious harm to human health.  However, it is uncertain how many consumers had actually read these few specific Republications.

637. In fact, this brings out another problem which had briefly been mentioned in Baturina.  In order to determine the loss caused by a particular libel, the court must first know about the circulation of the publication concerned.  In Baturina, the fifth category of publication was the article appearing in the Defendant’s website, and Eady J said: “[the] courts in recent years been approaching internet publication generally on the basis that there is no rebuttable presumption that material placed on a generally accessible website has been published to a substantial number of person (whether within the jurisdiction or elsewhere.”[272]  Eady J struck out the claim based on the internet publication on the ground that there was no evidence about its circulation.  The English Court of Appeal held that since there were a large number of hits on the defendant’s website from Russia, the claim should be allowed to proceed.  But the ultimate question remains that there must be some evidence about the circulation of a particular republication.

638. In the present case, most of the Republications were internet publications.  The court has no idea about the number of readers who have read these specific Republications, or indeed all the other Republications relied on by the Plaintiffs.  Hence, it would be difficult for the Plaintiffs to prove that all the loss of sales in the Mainland was caused by the readers reading the defamatory stings contained in the original Article which allegedly were replicated in the Republications, which I do not find it to be the case.

639. There is another issue about causation.  As mentioned above[273], there were other negative reports about the BaWang Group after the publication of the Article, i.e. the mislabelling incident about the men’s series products and missing and expired SFDA registration in respect of some Plaintiffs’ products. Such negative reports would also have adverse effect on the sales of BaWang Shampoo Products.  Although I agree that the effect of these incidents might have been smaller as compared with the 1,4-dioxane incident, the court would not be in a position to say how much discount to quantum should be made for the other unrelated incidents, in particular the court has no idea about the market condition in the Mainland.

640. The Plaintiffs also have the burden to establish that the damages caused by the Republications are not too remote to be recovered.  In this regard, the court has to determine whether the Defendant should have foreseen that the defamatory stings contained in the Article would be republished to the public in the Mainland.  

641. It is common ground that it was not possible to import the Magazine into the Mainland for sale.  There were only 49 subscribers of the Magazine with a Mainland phone number or email address as at July 2010, and only a total of 24 of those subscribers viewed the Article and constituted “hits”. In such circumstances, it would not be foreseeable that the general public in the Mainland would be able to view the whole Article.  It might be foreseeable the news that BaWang Shampoo Products did contain a carcinogenic substance known as 1,4-dioxane might be spread to the public in the Mainland, but it might not be foreseeable that the defamatory stings would be replicated in other publications in the Mainland.  Hence, I also find that the damages allegedly caused by the Republications are too remote to be recovered.

642. The Plaintiffs are claiming substantial damages for loss of sales in the Mainland.  For the reasons above, I am not satisfied that the Plaintiffs have discharged the burden of proving that all such loss was caused by the replication of the defamatory stings in the Republications.  The loss is also too remote to be recovered.  More importantly, as I will demonstrate in the following part of this Judgment, the Plaintiffs are not entitled to claim for the loss of sales caused by the Republications in the Mainland because of conflict of law principle.

(iv)  Conflict of law issue and the question of double actionability

643. As the Plaintiffs seek to recover damages relating to the loss of sales in the Mainland caused by the Republications in the same foreign jurisdiction, there is a conflict of law issue as to whether the Plaintiffs are entitled to recover such loss.

644. The Defendant argues under the present circumstances, Mainland law applies, and under Mainland law, the Defendant would only be liable to the Plaintiffs for its own publication of the Words, and not for the Republications in the Mainland.  Hence, the Plaintiffs should not be allowed to recover the loss resulting from the Republications in the Mainland.

645. According to Mr Yu, Mainland law is the law with the most significant relationship with the occurrence and with the parties in relation to the issue of damage allegedly sustained by the Plaintiffs in the Mainland.  Relying on authorities such as Boys v Chaplin[274] and Red Sea Insurance Co Ltd v Bouygues SA[275], Mr Yu submits that, as the Republications relied upon by the Plaintiffs took place in the Mainland, and the bulk of any loss or damage which the Plaintiffs might have suffered in consequence would have occurred in the Mainland, Mainland law is the lex loci delicti in the present case.  Further, a tort is committed where a defamatory statement is published, namely where the material is read, heard or seen.  Since the Words were alleged to have been republished in Mainland, Mainland is the place where the tort would have been committed.  The Plaintiffs cannot avoid this result simply by stating that they are not suing on those Republications as separate causes of action.

646. Mr Yu also argues that any defence which is valid under the lex loci deliciti (Mainland law according to the Defendant) is normally available to a defendant, irrespective of the lex fori (Hong Kong law in the present case), unless it is a procedural defence.  The law relating to damages is partly procedural and partly substantive.  Generally, remoteness and permissible heads of damages are questions of substance, while measure or quantification of damages is question of procedure.  Mr Yu submits that the extent of the Defendant’s liability for loss or damage caused by the Mainland Republications is a matter of substance and not procedure and is therefore governed by Mainland law as being the lex loci delicti or as having the most significant relationship with the occurrence and the parties.

647. On the other hand, Mr Pow argues that the place for the commission of the defamation was Hong Kong.  As mentioned above[276], the Plaintiffs have 2 options either to sue the Defendant both for the original publication and for the republication as two separate causes of action, or to sue the Defendant in respect of the original publication the damage which they have suffered by reason of its repetition so long as such damage is not too remote.  The Plaintiffs have chosen the second option.  Under such circumstances, the Plaintiffs are only suing on one tort, i.e. the Defendant’s publication of the Article in Hong Kong.  The fact that damages flowing from this tort beyond the territorial confines of Hong Kong is therefore not a conflict of law issue.  The place in which the tort was committed was the place of the publication.  Although the Words had been republished in the Mainland, the Republications were not sued upon as separate torts.  Libel is actionable per se, and damage is presumed in law and does not constitute an essential ingredient.  The place of occurrence of loss and damage is therefore irrelevant to the occurrence and completion of the tort of libel.  In such case, the place in which the tort was committed was the place of the publication, i.e. Hong Kong.

648. Further, relying on some Australian authorities such as Nygh’s Conflict of Laws in Australia[277] and David Syme v Grey[278], Mr Pow submits there can only be one place for the commission of the tort, and the court should insist on one single locus deliciti in the choice of law context.  It is not permissible to “change” the applicable law in a single tort action by reference to where the damages actually occurred in a case where the tort is actionable per se, as there can only be one place of the tort, and hence one single lex loci delicto.

649. As the locus deliciti was in Hong Kong, Mr Pow argues that there is no room for the application of foreign law. He relies on the following passage in Johnson: The Conflict of Law in Hong Kong (2 ed)[279]:

“It is not uncommon for there to be some scope for argument as to the place in which a tort has been committed. The point has less importance than it used to have, because the Red Sea exception now gives the court the ability to avoid the harshest consequences of the mechanical double actionability rule. Its remaining significance is two-fold:

1. If the place in question is Hong Kong, then, at least according to the law as currently understood through the English authorities, solely Hong Kong law will apply. On that basis, the court will have no discretion to apply another law even if the matter clearly has its most significant relationship with a legal system other than Hong Kong.

2.  If the place in question is outside Hong Kong then the general rule of double actionability still carries prima facie weight, notwithstanding the Red Sea exception.”

650. The position of the Plaintiffs is therefore double actionability rule does not apply and expert evidence on Mainland law is irrelevant.

651. Despite the able submission of Mr Pow, the locus deliciti of the tort of defamation is not a straightforward issue.  It is stated in Dicey, Morris & Collins:[280]

“Very difficult questions could arise in trying to determine the place of commission of the torts of libel or slander, slander of title, slander of goods, or other malicious falsehood. However, in the context of defamation the rule appears to be have been applied is that the relevant place is the country in which the defamatory statement was published. This appears to have been accepted without argument in one case where the English court was faced with a defamation claim arising out of a publication in a foreign country. This rule creates no practical difficulty where the statement is published in only one country. Where publication occurs in several countries, each publication may, potentially, give rise to a fresh cause of action and the law of each country in which publication takes place may become relevant, it being open, presumably, to the claimant to limit the claim to one such place if he or she so wishes.”

652. Baturina v Times Newspapers Ltd[281] is also the authority to support that foreign law is relevant in determining the Plaintiffs’ claim for the loss of sales caused by the Republications in the Mainland.  That is actually the scenario covered by the fourth category of publications in that case, i.e. third parties’ republications in Russia which replicated the defendant’s libel.  In allowing the plaintiff to proceed with the claim for damages arising from such kind of republications, Eady J held that the rule of double actionability applied.  If foreign law is irrelevant, there is no room for the application of such rule.  This is also consistent with the dicta of Eady J about the juridical basis to claim for damages resulting from republication as mentioned above.[282]  If there is a defence (presumably under foreign law) available to the original publisher or the person who republished the libel in a foreign jurisdiction, the plaintiff should not be allowed to claim either against the original publisher or the republisher for the loss resulting from such republications.  Hence, foreign law and the rule of double actionability are relevant.

653. I then turn to the evidence by the Mainland law experts.  The Plaintiffs’ expert is Mr Gao Jun (“Mr Gao”) while the Defendant’s expert is Mr Xu Jingbin (“Mr Xu”).  There is no dispute that both of them are qualified experts in Mainland law.

654. By the order of Poon J (as he then was) dated 20 April 2013, the court gave leave to the parties to adduce expert evidence on Mainland law on the following question:

“If a magazine publishes a defamatory statement and that statement is republished in the [Mainland] by other media, newspapers or internet sites and such republication causes loss and damage to the plaintiff in the [Mainland], would the original publisher be liable for any loss and damage caused by the republication and if so what is the extent to which the original publisher would be liable for such loss and damages?”

655. The aforesaid question has been framed based on the issue pleaded in the pleading.  The Defendant does not run the defence that, under Mainland law, the persons who republished the Words are not liable for defamation, rather the Defendant claims that, under Mainland law, the original publisher would not be liable for the damages caused by third parties’ republications.  This should be regarded as a matter of substance relating to a particular type of damages that can be claimed by a claimant and not about quantum.

656. Both experts agree that, under Mainland law, there is no specific legal provision concerning the liability of the original publisher for damages caused as a result of republication.  Nor is there any specific legal provision to show that an original publisher may be liable for a republication if it is the natural and probable consequence or reasonably foreseesable that it will be republished.  Both experts therefore try to answer the question posed by the court by analysing other relevant legal provisions and the judicial practice as demonstrated by the various cases decided by the courts in the Mainland.

657. Mr Gao is of the view that Mainland law is very similar to the common law concept concerning the original publisher being responsible for the foreseeable loss arising out of the publication.

658. Under Mainland law, there are 4 conditions under the general tort law to establish the claimant’s entitlement to sue in defamation (“the 4 Conditions”)[283]: (i) illegal act, i.e. an act of infringement to the legal person’s right to reputation; (ii) subjective fault on the part of the wrongdoer, i.e. the act is conducted intentionally or out of negligence; (iii) damage to the reputation of the legal person; and (iv) causation between the illegal act and the damage suffered by the legal person.  The court needs to consider the question of “subjective fault” in assessing both the issues of liability and quantum.  In addressing this question, the court has to determine whether the act is conducted intentionally or out of negligence, both contain the notion of foreseeability. If the tortfeasor does the infringing act intentionally, he clearly foresees the consequences of his intentional act.  As for doing an act negligently, it means that the tortfeasor foresees the consequences of his act but because he is over-confident, he fails to avoid such foreseeable consequences.

659. According to Mr Gao, his opinion is supported by: (i) Article 150 of the Opinions of the Supreme People’s Court Concerning the Implementation of the General Principles of the Civil Law issued on 26 January 1988 and revised on 18 December 2008[284]; and (ii) Answers of the Supreme People’s Court on Certain Issues Concerning the Trial of Cases Involving the Rights to Reputation (“the Answers”). The courts in the Mainland will therefore consider the concept of foreseeability in determining both the issues of liability and quantum.  In the context of republication, in assessing the “subjective fault” of the tortfeasor, the Mainland courts would assess whether it is foreseeable that the defamatory statements would be republished, and if so, whether the defendant intended such foreseeable republication or was negligent in failing to avoid such foreseeable republication.  The courts will also take that into account in assessing the compensation to be awarded to the claimant.  In other words, the liability of the original publisher for republication shall be analysed according to the 4 Conditions.  Once the 4 Conditions for deciding liability for infringement of the right to reputation have been established, the defendant is liable even for the republications, and the claimant does not need to consider whether anyone else is also liable unless and until the republishers are joined in the proceedings and are found to be liable for infringing the claimant’s right to reputation through their republications.

660. On the other hand, Mr Xu is of the opinion that, under Mainland law, the concept of forseeability is only relevant to the question of fault which is one of the 4 Conditions in determining the liability for defamation.  There is no reference in either the Tort Liability Law or the Explanation of Article 6 of the Tort Liability Law to foreseeability in respect of causation of a tortfeasor or the amount of damages.  According to the said Article 6, the question of causation is just whether “the conduct and the result links naturally without being broken by the external events”.  There is therefore no concept of foreseeability in the sense of remoteness of damage in Mainland law.  Insofar as Mr Gao is relying on said Article 150 and the Answers, Mr Xu is of the opinion that they actually deal with the question of the extent of liability of an original publisher focusing on the fault of the party being sued and not with the issue of damages.

661. According to Mr Xu, Article 13 of the Mainland Contract Law does refer to the concept of foreseeability in the context of determining the amount of compensation to be awarded.  Hence, the fact that foreseeability is not referred to in respect of compensation in the Tort Liability Law means that there is no such concept in assessing damages in tort cases.  Being a civil law system with codified laws, the difference in the two statutes is significant and the courts cannot go beyond the provisions in the statutes. 

662. In the opinion of Mr Xu, the republications by persons or parties other than the original publisher constitute separate and independent causes of action for which the republishers are legally liable to the claimant. If the republishers do not fulfil their duty to examine and verify the truthfulness of the news report when extracting or republishing an article, they are liable for defamation if such untrue report aggravates the damage to the reputation of another person by their republication which enlarges the dissemination range of the untrue report.  Such damages arising from the republications are separate from and different to any damages arising from the first publication of the libel by the original publisher.

663. Further, even if the original publication of the defamatory material and the republications by the republishers together cause the same damage, the original publisher is neither jointly, nor jointly and severally, liable with the republishers to the claimant in respect of the republications.  The original publisher and the republishers shall bear proportional share liability under Article 12 of the Tort Liability Law.[285]  This is because the acts of the original publisher and the republishers respectively are insufficient to cause the entire damage: there would be no republication without first publication, but if there had only been the original publication then there would not have been losses resulting from the republication.

664. In determining whether the original publisher would be liable for the damages caused by the republications, Mr Xu says that the applicable test, as supported by the various authorities in the Mainland, is whether the original publisher has control over the republications.  If he has no such control, he would not be liable.

665. In reply to Mr Xu’s opinion, Mr Gao takes the view that once the 4 Conditions for deciding liability for infringement of the right to reputation have been established, the original publisher is liable even for the republications, and the claimant does not need to consider whether anyone else is also liable unless and until the republishers are joined in the proceedings and are found to be liable for infringing the claimant’s right to reputation through their republications.  Mr. Gao considers that as the test for liability lies with the 4 Conditions for deciding liability for infringement of the right to reputation, Articles 8, 11 and 12 of the Tort Liability Law[286] are irrelevant.  However, if the court joins the republishers and finds them liable, then Article 11 of the Tort Liability Law shall apply, i.e. that the original publisher’s act is sufficient to cause the entire damage.

666. As Mainland law is based on the civil law system, I doubt whether we can use the same common law approach in analysing the statutes and the relevant legal documents in the Mainland. Both counsel have spent a lot of time in analysing the 4 Conditions, whether the concept of foreseesability exists in assessing damages in libel cases and the applicability of various provisions in the Tort Liability Law and the other various legal documents.  In determining which opinion reflects the law in the Mainland, I am of the view that the best starting point is to look at the cases decided by the Mainland courts.  Although the legal system in the Mainland is based on the civil law system with no concept of case law, I am of the view that the cases would show how the Mainland judges understood and applied the law, and this provide the best guidance in understanding the law in the Mainland.

667. In this regard, I agree with Mr Yu that the decided cases in the Mainland tend to support the opinion of Mr Xu that an original publisher should not be held liable for damage arising from the republication.  In the Mainland, the law places a very heavy duty of examination on the part of the republisher to verify the truth of the contents of the republication.[287] It is not disputed that a republisher may in principle have his own legal liability for defamation resulting from republication.[288]  In order to ensure that the republisher does fulfil such duty, the cases show that the republisher would have to be held responsible for loss flowing from the republication and the original publisher would not be liable for such loss unless he is in a position to control the republication or he makes the republication himself.  To hold the contrary may dilute the heavy duty imposed on the republisher in verifying the truth of the contents of the republication.

668. In Wang Hai v Blue Moon[289], the libel appeared in the defendant’s micro-blog, and the claimant claimed for RMB¥ 3 million on the basis that the impact of the libel was increased substantially as a result of republications.  However, both at first instance and on appeal, the court only awarded RMB¥100,000.  The reasoning of the first instance court was that the defendant was only responsible for the libel in the micro-blog which had limited number of followers.

669. In Beijing Shusheng Electronic Technology Co., Ltd. v Shanghai Shanda Network Development Co Ltd[290],the claimant tried to ask the original publisher to apologise for its actions not only on its own website, but also on all the republication media.  However, the court held that as the republishers’ actions were beyond the control of the original publisher, the original publisher need only to apologise on its own website. This case demonstrates that the original publisher should be responsible only for his own action, and not for any adverse impact caused by republications by others.

670. The case of Sun Mou v. 21 Well-known Newspapers and Websites[291] illustrates that there is a distinction between liability for the original publication and liability arising as a result of republication.  Although the claimant asked the court to find that the original publisher should be held responsible for the republications, the court stated that republishers were taking their own actions (各自進行), which constituted a tort and which the original publisher had no control over.  As the original publisher and republishers were not carrying out a joint act, they did not have joint liability and were not liable as joint tortfeasors (連代責任).

671. In Wenzhou Tietong Electrical Alloy Industry Co., Ltd. v. Guangdong 21^st Century Global Economy Newspaper House and Others[292], the 1^st defendant was the original publisher whilst the other defendants were republishers, and the liability of the respective defendants was held to be several liability and not joint liability.  The orders made against the 1^st defendant, namely the original publisher, and the various other republishers, were different.

672. In Jiaxing Phillips Automobiles Company Limited v. Business Time & Ors[293], one of the republishers, Sina, had fulfilled its duty of verification, and was therefore not liable, but another defendant, Luyuan Diandongche Gongsi, who had republished the offending article on its own website, was held liable for failing to take care of the truthfulness of what it republished.  More importantly, the court held that the original publisher, Caijing Shi Bao, and Luyuan should each bear responsibility, in other words there should be share proportional liability instead of joint liability.

673. Even Mr Gao accepts that in both Beijing Shusheng and Sun Mou, the respective claimants sought to persuade the courts to impose a remedy on the original publisher that took into account the acts of republication.  Yet both attempts failed.  In my judgment, these cases support Mr Xu’s view that the original publisher should not be held liable for damages arising from the republications.

674. On the other hand, Mr Gao relies on some judicial cases to support his opinion, but I do not accept that these cases can take the Plaintiffs’ case any further.

675. Jia Yinghua v. Wang Qingxiang[294]seems to support that the original publisher was liable for the loss of the republications, but the facts in that case show that the republications, though in the internet, were published in the name of the original publisher.  It was not the case that the republications were made in the name of a third party.  Further, the original libel was made in a book written by the defendant which had been widely published through the whole country.

676. Hu Nan Shun Tian Construction Group Co. Ltd. v. Gong Houqin[295] is also not a case on the point.  The claimant only claimed RMB¥1 as economic loss and notarial fee expenses in the sum of RMB¥3,900, and so there was no finding by the court that the original publisher was liable for compensation to the claimant for the economic loss suffered as a result of the republication.

677. Mr Gao also relies on Tian Shi Guo v Shandong Shang Bao[296], but this was another case in which both the original publication and the so-called republications were published by the defendant itself.  In fact, the libel in that case was contained in the hard copy format of a newspaper and also the electronic form of the newspaper, both published by the defendant and not by third parties.  Naturally, the defendant would have to pay compensation resulting from both publications.

678. A careful analysis of the cases shows that the courts in the Mainland have used the concept of control to explain the principle why an original publisher should not be responsible for the consequences of the republications.[297]  As the original publisher has no control over the republications, he should not be liable for the additional damages resulting from the republications in the absence of the concept of remoteness of damages.  Even Mr Gao has to accept that the courts in these cases used the notion of control to determine whether there was causation.

679. Further, the decided cases do not support Mr Gao’s opinion that foreseeability is the test for whether an original publisher should be held liable for damages resulting from republication. Neither of the experts has been able to find a single case in which a plaintiff sought to hold an original publisher liable for all the damage to his reputation on the basis of foreseeability, including damages resulting from republications.  Mr Gao tries to cite a number of cases to support his view[298], but it is clear from these cases that the concept of foreseeability was mentioned in the context of determining whether fault existed (i.e. liability) and not about the quantum of damages.  If foreseesability is such an important concept in determining quantum, I would expect the Mainland judges to have mentioned such concept in assessing damages and in deciding what remedy to be given.  Yet no such reference was made in the cases.

680. Apart from the decided cases, the following provisions of the Tort Liability Law also support Mr Xu’s opinion that the original publisher is not liable for the additional loss resulting from the republications:

“ Art 8  Where two or more persons jointly commit a tortious act which causes damage to another person, they shall be jointly and severally liable.

… … …

Art 11   Where two or more persons commit tortious acts respectively which cause the same damage, and each tortious act is sufficient to cause the entire damage, the tortfeasors shall be jointly and severally liable.

Art 12  Where two or more persons commit tortious acts respectively which cause the same damage, and if the extent to which each person is liable can be determined, the tortfeasors shall respectively bear their corresponding liabilities; or if it is difficult to determine the extent to which each person is liable, the tortfeasors shall be liable for compensation in equal shares.”

681. As explained by the Law Working Committee of the Standing Committee of the People’s Congress (法治工作委員會) in the Explanation to Article 12 of the PRC Tort Liability Law, the difference between Articles 11 and 12 is that Article 11 requires that “each tortious act is sufficient to cause the entire damage”, whereas Article 12 caters for the situation where each tortious act is not sufficient to cause the entire damage.  This is also my understanding by looking at the plain wordings of the provisions.

682. In the present case, there is no suggestion that the Defendant and the republishers acted jointly and so Article 8 does not apply.  Further, the acts of the Defendant and the republishers respectively were insufficient to cause the entire damage.  It is the Plaintiffs’ case that the loss of sales and profits in the Mainland was caused mainly by the Republications in the Mainland as the Magazine was not freely available or circulated in the Mainland.  In such case, it is clear that Article 12 of the Tort Liability Law applies and the Defendant and the republishers shall bear proportional share liability respectively.

683. For the above reasons, I find in favour of the Defendant on the Mainland law issue.  If a magazine publishes a defamatory statement and that statement is republished in the Mainland by other media, newspapers or internet sites, and such republications cause additional loss and damage to the plaintiff in the Mainland, the original publisher, as a matter of substantive law in the Mainland, would not be liable for any loss or damage caused by the republications.

684. By reason of the fact that the republications were separate torts committed in the Mainland, I accept that Mainland law is the law with the most significant relationship with the occurrence and with the parties in relation to the issue of damage allegedly sustained by the Plaintiffs in the Mainland.  Further, any defence which is valid under Mainland law being the lex loci deliciti, which is substantive in nature, is available to the Defendant.  Hence, the Plaintiffs are not entitled to claim for any additional damages against the Defendant resulting from the republications in the Mainland.

685. Based on the aforesaid analysis, it is not necessary for me to consider the second part of the question posed by the court as to the extent to which the Defendant, being the original publisher, would be liable for such loss and damages.

(v)  Right to claim for loss of the sales and profits in Hong Kong

686. The issue of reflective loss is relatively a simple one.  The Defendant argues that the non-Mainland loss of profits was actually suffered by HK1 and HK2 which are the 100% Hong Kong subsidiaries indirectly held by the 1^st Plaintiff.  By reason of the no-reflective-loss principle, the 1^st Plaintiff (being the ultimate holding company) is not entitled to sue for the loss of profits suffered by its subsidiaries, i.e. HK1 and HK2.

687. The Court of Final Appeal in Waddington v Chan Chun Ho[299]has confirmed the no-reflective-loss principle which ensures that a defendant can only be sued once for the same loss and, in doing so, prioritizes the company’s claim as a matter of principle.  Thus, a shareholder cannot sue for the loss suffered by the company that he owns on the ground that the shareholder suffers a loss in the value of the shareholding, as such loss is merely “reflective” of the company’s loss.  Instead, it should be the company who can sue for the loss.  As Lord Millet NPJ said:[300]

“If the shareholder is allowed to recover in respect of such loss, then either there will be double recovery at the expenses of the defendant or the shareholder will recover at the expense of the company and its creditors and other shareholders. Neither course can be permitted. This is a matter of principle; there is no discretion involved.”

688. Nevertheless, Mr Pow argues that a prerequisite of the application of the no-reflective-loss principle is that the company is able to sue for the loss.  If the company is unable to do so, there will be no double recovery and the shareholder’s claim will not be at the expense of the company.  In George Fischer (Great Britain) v Multi Construction Ltd[301], the parent company contracted with the defendant to install equipment at its subsidiary. However, the equipment was defectively installed causing loss to the subsidiary. Since the contract was made by the parent company, the subsidiary did not have a cause of action against the defendant for breach of contract.  The English Court of Appeal held that the parent company was entitled to sue for the loss suffered at the subsidiary level.

689. Mr Pow submits that HK1 and HK2 may not have a good claim against the Defendant because the loss suffered was caused by the loss of reputation of the 1^st or 2^nd Plaintiff.  As the companies (HK1 and HK2) actually suffered the loss do not have a valid cause of action of libel or malicious falsehood against the Defendant, their shareholder (i.e. the 1^st Plaintiff) should be allowed to pursue the loss suffered by HK1 and HK2.

690. But this argument runs contrary to the Plaintiffs’ own case.  In justifying the inclusion of the 2^nd Plaintiff as a claimant, Mr Pow has argued earlier that, though the readers might not know the specific names of each subsidiary, the Words actually referred to all the subsidiaries of the BaWang Group responsible for the manufacturing and marketing of the BaWang Shampoo Products in Hong Kong.[302] It is therefore the Plaintiffs’ own case that the reputation of HK1 and HK2, being such companies, had also been defamed by the publication of the Article, and so there is no reason why they cannot maintain a cause of action, if properly established, against the Defendant for libel.

691. Even stronger argument applies in the case of the claim for malicious falsehood.  According to the Plaintiffs, the Words had wrongly suggested that BaWang Shampoo Products were hazardous to human health. The nature of the cause of action does not relate to the reputation of the 1^st Plaintiff or the BaWang Group, and the loss of sales in Hong Kong and other non-Mainland countries was actually suffered by HK1 and HK2.  In such circumstances, it is clear that HK1 and HK2 can maintain a cause of action, if properly established, against the Defendant.

692. It is difficult to understand why the Plaintiffs have not joined HK1 and HK2 as additional claimants.  The Defendant had raised the no-reflective-loss defence at the early stage of the proceedings. Apparently, the 2^nd Plaintiff was joined as an additional claimant after considering such defence, and yet the Plaintiffs have not joined the other relevant subsidiaries as additional plaintiffs.  As pointed out by Lord Millet in Waddington, the court has no discretion in such matter and the Plaintiffs cannot therefore claim for the loss of profits in respect of BaWang Shampoo Products outside the Mainland including Hong Kong.

(vi)  The effect of the said rulings on the assessment of damages

693. By reason of the aforesaid rulings, the Plaintiffs would not be able to claim for any actual loss of profits caused by the reduction in sales of BaWang Shampoo Products, whether in the Mainland or in Hong Kong.

694. The Plaintiffs cannot claim for the loss of profits relating to the sales in the Mainland because:

(i)  the Plaintiffs have failed to discharge the burden of proving that all the loss in sales in the Mainland was caused by the replication of defamatory stings in the Republications;

(ii)  the damages resulting from the Republications are too remote to be recovered by the Plaintiffs; and

(ii)  more importantly, under Mainland law, the original publisher would not be liable for the additional damage caused by republications by third parties of which he has no control, and so the Defendant is entitled to rely on such defence to defend the Plaintiffs’ claim for the loss resulting from the Republications which were published in a foreign jurisdiction (i.e. the Mainland).

695. The Plaintiffs also cannot claim for the loss of profits relating to the sales in Hong Kong because such loss was actually suffered by the two subsidiaries of the 1^st Plaintiff, namely HK1 and HK2.  As they have not been joined as the claimants in these proceedings, the existing Plaintiffs are not able to claim the loss on their behalf or to sue for the loss of the value of their shareholdings because of the no-reflective-loss principle.

696. I also disallow the claim for all the special damages by the 2^nd Plaintiff.  The 2^nd Plaintiff was the 1^st Plaintiff’s subsidiary responsible for conducting the sales of BaWang Shampoo Products in the Mainland.  Hence, I have reason to believe that all the special damages allegedly suffered by the 2^nd Plaintiff relate to the additional damage caused by the Republications in the Mainland.  The expenses were incurred with a view to promote the image of the BaWang Group in the Mainland.  For the same conflict of law principle in disallowing the Plaintiffs to claim for the loss of profits for the sales in the Mainland, the 2^nd Plaintiff likewise cannot claim for such special damages.

697. Further, the expenses for purchasing the television advertisement slots did not arise from the publication of the Article.  The television advertising contract was made before the publication of the Article, and what resulted from the publication was that the 2^nd Plaintiff decided to use the slots it had already purchased for other advertisements.  In other words, even without the publication, those expenses had to be incurred in any event.

698. I allow the 1^st Plaintiff’s claim in the sum of HK$4,652.50 relating to the cost of issuing the public announcement.  There was a sudden drop in the share price of the 1^st Plaintiff after the publication of the Article, and so the 1^st Plaintiff had to issue an announcement as required by the listing rules in Hong Kong.  Such loss was directly related to and caused by the publication of the Article.  Further, the 1^st Plaintiff itself had been defamed and so there is no room for the application of the no-reflective-loss principle as contended for by Mr Yu.

699. Given that I disallow the Plaintiffs’ claim for loss of profits, it is quite unnecessary for me to address the other arguments of the Defendant, such as whether the Plaintiffs should have mitigated the loss of profits by telling the public that they would reformulate the shampoos by reducing or eliminating the amount of 1,4-dioxane rather than simply repeating that their shampoos were safe, or that the quantum of the claim should be adjusted downwards to take into account the tax liability of the Plaintiffs in respect of the potential profits made.

700. That leaves only the claim for general damages and exemplary damages.

(vii)  General damages

701. The Plaintiffs have all along focused their claim on the loss of profits, and so the parties have not devoted a lot of time at the trial to deal with the question of general damages.  Given my rulings above, the assessment of general damages may now become more important, and so I have invited the parties to make further submissions on this issue.

702. It is established law that in an action for defamation, the claimant does not only recover pecuniary loss, but is entitled to general damages for injury to his reputation and vindication of his reputation.[303]

703. The purpose of general damages is to compensate the claimant for the effects of the defamatory statement.[304]  General compensatory damages serve 3 functions:[305]

(i)  to console the claimant for hurt and distress suffered;

(ii)  to repair the harm that has been done to his reputation (including his business reputation); and

(iii)  to vindicate his reputation.

704. Since the Plaintiffs are companies, there will be no compensation for injury to feelings.

705. The need for vindication is well explained by Lord Hailsham in Cassell & Co Ltd v Broome:[306]

“Not merely can [the claimant] recover the estimated sum of his past and future losses, but, in case the libel, driven underground, emerges from its lurking place at some future date, he must be able to point to a sum awarded by [the court] sufficient to convince of the baselessness of the charge.”

706. The effect of the libel can be long-lasting.  In Hill v Church of Scientology, the Supreme Court of Canada made the following observations:[307]

“a defamatory statement can seep into the crevasses of the subconscious and lurk there ever ready to spring forth its cancerous evil. The unfortunate impression left by a libel may last a lifetime. Seldom does the defamed person have the opportunity of replying and correct the record in a manner that will truly remedy the situation.”

707. It should be noted that “the court should be careful not to drive down damages in libel cases to a level which publishers might with equanimity be tempted to risk having to pay”.[308]

708. In Gatley[309], the learned authors summarised the matters affecting the level of award: the conduct of the claimant, his credibility, his position and standing, and the subjective impact that the libel has had on him, the nature of the libel, its gravity and the mode and extent of its publication, the absence or refusal of any retraction or apology, and the conduct of the defendant from the time when the libel was published down to the verdict.

709. The court should also be guided by the overarching principle that the sum to be awarded must be proportionate to the damage suffered and that it is reasonably required to compensate the claimant and re-establish his reputation.[310]

710. In assessing the amount of damages, the court will normally arrive at a “global figure”, rather than seeking to allocate a particular sum to each of the heads of damages.[311]

711. By reason of the same conflict of law principle as mentioned above, the court, in assessing the general damages, should not take into account the additional damage to the reputations of the Plaintiffs caused by the Republications in the Mainland.

712. As I see it, the main issue here is whether the court should use the figure for the loss of profits in respect of the sales in Hong Kong as a yardstick to assess the general damages.

713. Mr Pow has referred me to the authorities such as Lewis v Daily Telegraph Ltd[312], Ratcliffe v Evans[313], Cambridge v Makin[314], Metropolitan International Schools Ltd v Designtechnica Corporation[315] and Chu Siu Kuk Yuen v Apple Daily Ltd[316], and argues that the court should take into account the general loss of business in making an award of general damages.

714. However, the claimants in all these cases were the same persons or entities who suffered the general loss of business.  These claimants had difficulty in proving the actual loss or special damage caused by the libels, and so the courts took into account the claimants’ general loss of business in assessing the quantum of general damages.

715. The present case is different.  It is not that the Plaintiffs are not able to prove the actual loss of business, but that such loss in Hong Kong was actually suffered by some other parties, i.e. HK1 and HK2.  In such circumstances, I do not accept that the court should take into account the actual loss of business suffered by other parties in assessing the general damages to the Plaintiffs.  The approach by the court should therefore be, assuming that the court is to allow HK1 and HK2 to claim for the actual loss of profits in Hong Kong for the Relevant Period, what is the proper amount that should be awarded to the Plaintiffs separately as general damages for the harm caused to their reputations.

716. In assessing the quantum of general damages, the court should therefore focus on the damage caused to the reputations of the Plaintiffs by the publication of the Article, and to vindicate the reputations of the Plaintiffs in this regard.  The Plaintiffs have been portrayed as irresponsible and evasive companies lacking commercial morality.  Knowing the danger of 1,4-dioxane to human health, they nevertheless proceeded to supply the BaWang Shampoo Products to the public.  Further, they used inferior raw materials to produce the Products with a mercenary motive to save costs.

717. At the time of the publication of the Article, the Plaintiffs could be regarded as successful companies.  The BaWang Shampoo Products, which were advertised as shampoos made with Chinese herbals, had been quite successful in the market, though the growth rate had slowed down before the publication of the Article.

718. Mr Yu submits that the court should not make any award on general damages because there is no evidence before the court which shows loss to the Plaintiffs’ reputation.  Further, the loss of goodwill to a company can only be assessed with proper expert evidence from accountant because there is special accounting treatment for valuation of goodwill.

719. I do not accept that the court should be debarred from making any award on general damages because of the lack of expert evidence on the valuation of goodwill of the Plaintiffs damaged by the publication of the Article.  After all, the Plaintiffs are only claiming for general damages with a view to repair the harm that has been done to their reputation and to vindicate their reputation.  In libel cases, damages are often “at large” in the sense that they cannot be assessed by reference to any mechanical, arithmetical or objective formula and the court is entitled to take account a wide range of matters in reaching an award for damages[317], including but not limited to the scale of business operated by the Plaintiffs.  Hence, the court has to try its best to arrive at a figure which would, hopefully, serve the purpose of repairing the harm and vindicating the reputation of the Plaintiffs.

720. I find that the Words in the Article had caused serious damage to the reputations of the Plaintiffs, making them more difficult to market their existing products or to launch new products in the future. Such loss would go beyond the Relevant Period, and it may emerge “from its lurking place at some future date”.  The award should therefore not be too low, and it must be sufficient “to convince a bystander of the baselessness of the charge”. Although the court should not use the actual loss of profits as a yardstick to assess the quantum of general damages, the court can take into account the Plaintiffs’ scale of business in Hong Kong in the assessment exercise.  Further, it must not be at “a level which publishers might with equanimity be tempted to risk having to pay”.

721. On the other hand, the court has to ensure that the award would not be too high to the extent it would constitute an impediment to freedom of expression.  In Tolstoy v Miloslavsky v United Kingdom[318], the European Court struck down an award of £1,500,000 as an infringement of Article 10 of the European Convention as the size of the award was disproportionate and could not be justified as being “necessary in a democratic society”.  The Hong Kong Court of Appeal applied Tolstoy and struck down the award in Cheung Ng Sheong Steven v Eastweek Publisher Ltd[319] on the ground that if an excessive award is allowed to stand, there is a risk that it will constitute an impediment to freedom of expression and of opinion laid down in Article 16 of HKBOR Ordinance and cannot be regarded as necessary to protect the reputation of the plaintiff.

722. Both parties cannot find any comparable cases on quantum of general damages with facts similar to those in the present case.  So far as I know, HK$3 million awarded as general damages in Chu Siu Kuk Yuen v Apple Daily Ltd & Ors[320] was one of the highest awards made in defamation cases.  The facts of that case were quite different.  It involved a claimant who was a person and not a corporation.  As a result of a defamatory article published in the Defendant’s sister newspaper, the claimant, who was a solicitor, suffered depression which materially contributed to her child being born almost 13 weeks premature and remained in a life-threatening condition for some time thereafter.  It is not surprising that such a high award was made in that particular case.

723. Without the assistance of comparable cases, the court can only try its best to assess a figure taking into account all the factors mentioned above.  Obviously, the damages suffered by the 1^st Plaintiff would be higher because it would be regarded by the public as the ego of the whole BaWang Group.  The 2^nd Plaintiff was the actual manufacturer of BaWang Shampoo Products in Hong Kong, and to me, the damages suffered by it should be less than that of the 1^st Plaintiff.  Having carefully considered all the factors in the present case, I award the sums of HK$2 million and HK$1 million as general damages to the 1^st and 2^nd Plaintiffs respectively.

(viii)  Exemplary damages

724. To make a defendant liable for exemplary damages, the claimant has to show that the defendant would have had to know that or be reckless whether he was committing a tort, and have had decided to publish, because the material advantages from publication outweighed the prospects of material loss.[321] However, the fact that a defendant is engaged in an activity aimed at profit, for example publishing a magazine, does not of itself justify an award of exemplary damages.  There must be something more specific than that.[322]

725. Further, for there to be exemplary damages, the court must be satisfied that the defendant had no genuine belief in the truth of what he published.  The defendant must have suspected that the words were untrue and have deliberately refrained from taking obvious steps which, if taken, would have turned suspicion into certainty.[323]

726. Mr Pow complains that the Defendant had deliberately published the Article in a sensational manner.  The Defendant did not seek further clarification about the NICNAS standard from the two experts and ignored any materials in support of the Plaintiffs’ side of the story, and the only reasonable explanation was that the Defendant wanted to avoid the issue. The presentation of the contrary view would lessen the impact of the Article targeting the Plaintiffs, with the result it would reduce the attraction of the Article to the readers.  Hence, Mr Pow says that the Defendant was reckless as to the truth of the contents in the report.

727. I have already dealt with this argument in the earlier part of this Judgment on responsible journalism[324].  In short, the Defendant had adopted a naïve and unprofessional approach in reporting the story.  The Defendant might have been grossly negligent, but I do not find any sinister motive behind the making of the report.  The mere fact that the Magazine was published for profit is not by itself enough to justify exemplary damages.

728. Neither would I attach a lot of weight to the fact that the Defendant refused to retract the statements in the Article as demanded by the Plaintiffs.  After all, this case involves matters of some importance and scientific materials of some controversy, and so the court should not hold against the Defendant for adopting such stance.  Hence, I disallow the claim for exemplary damages.

VIII  CONCLUSION

729. As the Article was a one-off publication, I do not see any necessity of granting injunctive relief in the present case.

730. For the above reasons, I grant judgment in favour of the 1^st Plaintiff in sum of HK$2,004,652.5 and 2^nd Plaintiff in the sum of HK$1,000,000.  Since the Plaintiffs do not succeed in a number of issues, in particular those relating to quantum, I make a costs order nisi that the Defendant do pay to the Plaintiffs 80% of the costs of this action with certificate for 2 counsel, which shall be made absolute 21 days after the date of the handing down of this Judgment.

731. I have not dealt with the question of interest. If the Plaintiffs seek to make any claim for interest, such claim has to be made within 21 days from the date of the handing down of this Judgment.

732. That concludes one of the most difficult cases that I have dealt with in my judicial career.  With the diligence of counsel, I think that they have covered every possible argument in favour of their clients.  Although it makes my work more difficult, I must express my gratitude to them for all the assistance they have provided to me, without which I do not think I could navigate through the massive evidence presented to the court.