Judgments

Decision Information

Decision Content

A-31-04

2005 FCA 1

Yi Mei Li (Appellant)

v.

The Minister of Citizenship and Immigration (Respondent)

Indexed as: Li v. Canada (Minister of Citizenship and Immigration) (F.C.A.)

Federal Court of Appeal, Rothstein, Noël and Malone JJ.A.--Toronto, November 30, 2004; Ottawa, January 5, 2005.

Citizenship and Immigration -- Status in Canada -- Persons in need of protection -- Appeal from Federal Court decision dismissing judicial review of Refugee Protection Division decision denying appellant's claim to be person in need of protection under Immigration and Refugee Protection Act, s. 97(1) -- Three certified questions -- (1) Standard of proof for purposes of s. 97 balance of probabilities -- (2) Degree of risk envisaged by "substantial grounds for believing" in Act, s. 97(1)(a), whose words mirror Convention Against Torture and Other Cruel, Inhuman or Degrading Treatment or Punishment, Art. 3 balance of probabilities or more likely than not based on judicial interpretation of Art. 3 -- (3) That test also applying to Act, s. 97(1)(b) -- Appeal dismissed.

Construction of Statutes -- Appellant claiming to be person in need of protection under Immigration and Refugee Protection Act, s. 97(1) -- S. 97(1)(a) defining person in need of protection as person whose removal would subject them to a danger, believed on substantial grounds to exist, of torture -- Question as to requisite degree of risk of torture envisaged by s. 97(1)(a) -- Words of s. 97(1)(a) mirroring those of Convention Against Torture and Other Cruel, Inhuman or Degrading Treatment or Punishment, Art. 3 -- F.C.A. interpreting words of Art. 3 in Suresh v. Canada, adopting "balance of probabilities" ("more likely than not") test -- Convention, Art. 3 and Act, s. 97(1)(a) almost identical, dealing with same subject-matter, should be interpreted same way -- When statutory provision (e.g. s. 97(1)(a)) apparently modelled on existing legislation, whether from same jurisdiction or not and including United Nations Conventions (e.g. Art. 3), interpretation of model legislation presumed to have been known and taken into account by legislator -- Here, House of Commons Debates, proceedings of Standing Senate Committee on Human Rights, indicating Parliament aware of Suresh -- Reasonable to infer Parliamentary intention that degree of torture test in Act, s. 97(1)(a) same as in Suresh, i.e. balance of probabilities or more likely than not.

This was an appeal from a decision of the Federal Court dismissing the appellant's judicial review application. That application stemmed from a decision of the Refugee Protection Division (RPD) of the Immigration and Refugee Board, which denied the appellant's refugee claim and claim for protection.

The appellant claimed to be a person in need of protection under subsection 97(1) of the Immigration and Refugee Protection Act. The standard of proof and degree of danger of torture applied to determine that the appellant was not in need of protection gave rise to three questions certified by the Federal Court Judge. The Judge found that under subsection 97(1), the evidence must establish, on a balance of probabilities, that it is more likely than not that the appellant would be tortured or subjected to cruel and other degrading treatment upon his return to China. The appellant argued that the test to apply is whether there is a reasonable chance (which was said to be less than probability) of a danger of torture or a reasonable chance of a risk to life or of cruel and unusual treatment or punishment.

Held, the appeal should be dismissed.

The first certified question concerned the standard of proof for purposes of section 97. The parties and the Court agreed that the standard is proof on a balance of probabilities. However that standard must be distinguished from the legal test to be met in order to be entitled to protection under paragraph 97(1)(a).

The second certified question dealt with the requisite degree of risk of torture envisaged by the expression "substantial grounds for believing that" which appears in paragraph 97(1)(a). That paragraph was adopted to give effect to Canada's obligations as a signatory to the Convention Against Torture and Other Cruel, Inhuman or Degrading Treatment or Punishment (CAT), and its words mirror those used in Article 3 of the Convention. In Suresh v. Canada (Minister of Citizenship and Immigration) (C.A.), Robertson J.A. interpreted the words of Article 3 and rejected the lower "mere possibility" of torture test and the higher "highly probable" test in favour of what he called the "balance of probabilities test" (which can also be expressed as "more probable than not" or "more likely than not"). General Comments of the United Nations Committee Against Torture, decisions of that Committee and a decision of the United States Court of Appeals for the Ninth Circuit are consistent with the approach of Robertson J.A.

Suresh is dispositive of the test for danger of torture under paragraph 97(1)(a). Because the words in paragraph 97(1)(a) and Article 3 are almost identical and deal with the same subject-matter, they should be interpreted the same way. Paragraph 97(1)(a) was enacted after Suresh. When a statutory provision (in this case paragraph 97(1)(a)) appears to be modelled on existing legislation, whether from the same or another jurisdiction and including United Nations Conventions (in this case, Article 3 of the CAT), interpretation of the model legislation is presumed to have been known and taken into account in drafting the new legislation. Here, the Debates of the House of Commons indicate that Parliament was aware of the Suresh case, and in the proceedings of the Standing Senate Committee on Human Rights, there is express reference to the F.C.A. decision in Suresh. Parliament had an opportunity to enact a lower-level test in paragraph 97(1)(a) but did not do so. In these circumstances, it is reasonable to infer that the Parliamentary intention was that the test for the degree of danger of torture in paragraph 97(1)(a) is a balance of probabilities or more likely than not.

Although the wording of this test is equivalent to the wording of the legal test to be met in order to be entitled to protection under paragraph 97(1)(a), the two are distinct steps. Proof on a balance of probabilities is the standard the RPD will apply in assessing the evidence adduced before it for purposes of making its factual findings. The test for determining the danger of torture is whether, on the facts found by the RPD, the RPD is satisfied that it is more likely than not that the individual would personally be subjected to a danger of torture.

In response to the appellant's argument that it makes "no rational sense" to adopt a higher degree of danger test for protection against torture than for determining Convention refugee status (section 96), there are a number of differences between section 96 and paragraph 97(1)(a) because of which, it cannot be said that the provisions are so closely related that it would be irrational if the tests under both sections were not identical.

As for the final certified question, which was whether the test under paragraph 97(1)(b) is the same as the one under paragraph 97(1)(a), in the absence of some compelling reason suggesting a particularly low or a particularly high-level test for paragraph 97(1)(b), it is.

statutes and regulations judicially

considered

Convention against Torture and Other Cruel, Inhuman or Degrading Treatment or Punishment, December 10, 1984, [1987] Can. T.S. No. 36, Art. 3.

Immigration and Refugee Protection Act, S.C. 2001, c. 27, ss. 74(d), 95(1)(b), 96, 97(1).

cases judicially considered

applied:

Suresh v. Canada (Minister of Citizenship and Immigration), [2000] 2 F.C. 592; (2000), 18 Admin. L.R. (3d) 159; 5 Imm. L.R. (3d) 1; 252 N.R. 1 (C.A.); revd on other grounds [2002] 1 S.C.R. 3; (2002), 208 D.L.R. (4th) 1; 37 Admin. L.R. (3d) 152; 90 C.R.R. (2d) 1; 18 Imm. L.R. (3d) 1; 281 N.R. 1; 2002 SCC 1.

considered:

O.K.K. (Re), [2002] R.P.D.D. No. 483 (QL); Adjei v. Canada (Minister of Employment and Immigration), [1989] 2 F.C. 680; (1989), 57 D.L.R. (4th) 153 (C.A.); A.R. v. Netherlands, Communication No. 203/2002: Netherlands 21/11/2003 (CAT/C/31/D/203/2002); Selvaratnam v. Ashcroft, 81 Fed. Appx. 907 (9th Cir. 2003).

referred to:

Jose Pereira E Hijos, S.A. v. Canada (Attorney General) (2002), 299 N.R. 154; 2002 FCA 470; Chalk River Technicians and Technologists v. Atomic Energy of Canada Ltd., [2003] 3 F.C. 313; (2002), 298 N.R. 285; 2002 FCA 489; Miller v. Minister of Pensions, [1947] 2 All E.R. 372 (K.B.D.); Lyons et al. v. The Queen, [1984] 2 S.C.R. 633; (1984), 58 A.R. 2; 14 D.L.R. (4th) 482; [1985] 2 W.W.R. 1; 15 C.C.C. (3d) 417; 43 C.R. (3d) 97; 56 N.R. 6; Bell ExpressVu Limited Partnership v. Rex, [2002] 2 S.C.R. 559; (2002), 212 D.L.R. (4th) 1; [2002] 5 W.W.R. 1; 166 B.C.A.C. 1; 100 B.C.L.R. (3d) 1; 18 C.R.R. (4th) 289; 93 C.R.R. (2d) 189; 2002 SCC 42; Chan v. Canada (Minister of Employment and Immigration), [1995] 3 S.C.R. 593; (1995), 128 D.L.R. (4th) 213; 187 N.R. 321.

authors cited

Canada. Parliament. Standing Senate Committee on Human Rights. Proceedings, 1st Sess., 37th Parl., No. 3 (24 September 2001).

House of Commons Debates, 080 (18 September 2001), 5525.

Sullivan, Ruth. Sullivan and Driedger on the Construction of Statutes, 4th ed. Toronto: Butterworths, 2002.

United Nations. Committee Against Torture. General Comment No. 01: Implementation of Article 3 of the Convention in the Context of Article 22. U.N. Doc. A/53/44, Annex 1X, CAT General Comment No. 01 (21 November 1997).

APPEAL from a decision of the Federal Court ([2004] 3 F.C.R. 501; 2003 FC 1514), which found that to be a person in need of protection under subsection 97(1) of the Immigration and Refugee Protection Act, the evidence must establish, on a balance of probabilities, that it is more likely than not that the appellant would be tortured or subjected to cruel and other degrading treatment upon his return to China. Appeal dismissed.

appearances:

Michael E. Korman for appellant.

Ian Hicks for respondent.

solicitors of record:

Otis & Korman, Toronto, for appellant.

Deputy Attorney General of Canada for respondent.

The following are the reasons for judgment rendered in English by

[1]Rothstein J.A.: This is an appeal of a decision of Gauthier J. of the Federal Court [[2004] 3 F.C.R. 501] on three questions certified by her under paragraph 74(d) of the Immigration and Refugee Protection Act, S.C. 2001, c. 27 [the Act].

FACTS

[2]Yi Mei Li, a citizen of China, arrived in Canada in April 2001 on a boat operated by "human" smugglers known as snakeheads. He made a refugee claim on the basis of fear of persecution by reason of his alleged religious beliefs and membership in a particular social group. He also claimed to be a person in need of protection because of a risk to his life, because he would be subjected to a risk of cruel and unusual treatment or punishment or because he would be in danger of being tortured if returned to China.

[3]A panel of the Refugee Protection Division of the Immigration and Refugee Board denied Mr. Li's refugee claim and claim for protection because of a lack of credibility [O.K.K. (Re), [2002] R.P.D.D. No. 483 (QL)]. The panel concluded that he did not establish a well-founded fear of persecution based on any ground contained in the Convention refugee definition [section 96 of the Act]. It also concluded that there was no persuasive evidence that Mr. Li was at risk of losing his life or of being subjected to cruel and unusual treatment or punishment or of being in danger of torture if he was returned to China.

[4]On judicial review, Gauthier J. found no reviewable error by the panel in respect of its credibility findings, its rejection of alleged threats to Mr. Li from the snakeheads and its rejection of Mr. Li's allegation of denial of state protection by the Chinese government against the snakeheads.

ISSUES

[5]The standard of proof and degree of danger of torture applied by the panel in determining that Mr. Li was not a person in need of protection under subsection 97(1) of the Act gave rise to the questions certified for appeal by Gauthier J.

[6]Mr. Li argues that under subsection 97(1), the test to apply in respect of an individual in need of protection is whether there is a reasonable chance of a danger of torture or a reasonable chance of a risk to life or of cruel and unusual treatment or punishment. Reasonable chance is said to be less than probability. Mr. Li says the panel wrongly assessed his claim to protection on the basis of probability rather than reasonable chance.

[7]Gauthier J. found that under subsection 97(1), the evidence must establish, on a balance of probabilities, that Mr. Li faces a substantial danger of being tortured upon his return, that is, that it is more likely than not that he would be tortured or be subjected to cruel and other degrading treatment. She found that the panel properly assessed Mr. Li's claim under subsection 97(1). As a result, she denied the application for judicial review.

QUESTION No. 1

[8]Question No. 1 is:

Does section 97 of the Act require that a person establish, on a balance of probabilities, that he or she will face the danger or risks described in paragraphs 97(1)(a) and (b)?

[9]The parties are agreed that the answer to this question is in the affirmative. I think they are correct. Unless the words of a statute or the context requires otherwise, the standard of proof in civil cases is always proof on a balance of probabilities.

[10]However, the standard of proof must not be confused with the legal test to be met. The distinction was recognized in Adjei v. Canada (Minister of Employment and Immigration), [1989] 2 F.C. 680 (C.A.), in the context of a claim for Convention refugee status. The relevant provision is now section 96 of the Immigration Refugee and Protection Act, which provides:

96. A Convention refugee is a person who, by reason of a well-founded fear of persecution for reasons of race, religion, nationality, membership in a particular social group or political opinion,

(a) is outside each of their countries of nationality and is unable or, by reason of that fear, unwilling to avail themself of the protection of each of those countries; or

(b) not having a country of nationality, is outside the country of their former habitual residence and is unable or, by reason of that fear, unwilling to return to that country.

[11]At page 682 of Adjei, McGuigan J.A. stated:

It was common ground that the objective test is not so stringent as to require a probability of persecution. In other words, although an applicant has to establish his case on a balance of probabilities, he does not nevertheless have to prove that persecution would be more likely than not. [Emphasis added.]

[12]McGuigan J.A. adopted the "reasonable chance [of] persecution" test as the legal test to meet to obtain Convention refugee status, i.e. not necessarily more than a 50 percent chance but more than a minimal possibility of persecution.

[13]The certified question deals with subsection 97(1). The relevant portions of subsection 97(1) provide:

97. (1) A person in need of protection is a person in Canada whose removal to their country or countries of nationality or, if they do not have a country of nationality, their country of former habitual residence, would subject them personally

(a) to a danger, believed on substantial grounds to exist, of torture within the meaning of Article 1 of the Convention Against Torture; or

(b) to a risk to their life or to a risk of cruel and unusual treatment or punishment . . . .

[14]As was found by McGuigan J.A. to be the case with respect to section 96, nothing in subsection 97(1) suggests that the standard of proof to be applied in assessing the danger or risk described in paragraphs 97(1)(a) and (b) is anything other than the usual balance of probabilities standard of proof. The answer to the first certified question is therefore:

The standard of proof for purposes of section 97 is proof on a balance of probabilities.

QUESTION No. 2

[15]The second certified question is:

What is the requisite degree of risk of torture envisaged by the expression "substantial grounds for believing that"?

[16]The words "believed on substantial grounds to exist" appear in paragraph 97(1)(a). I interpret the certified question to ask what degree of danger of torture is envisaged by these words.

The Test in Paragraph 97(1)(a)

[17]As observed by Gauthier J., paragraph 97(1)(a) was adopted to give effect to Canada's obligations as a signatory to the Convention Against Torture and Other Cruel, Inhuman or Degrading Treatment or Punishment, December 10, 1984, [1987] Can. T.S. No. 36 [Convention Against Torture]. Article 3 of the Convention Against Torture provides:

Article 3

1. No State Party shall expel, return ("refouler") or extradite a person to another State where there are substantial grounds for believing that he would be in danger of being subjected to torture.

[18]It is apparent that the words in paragraph 97(1)(a):

97. (1) . . . would subject them personally

(a) to a danger, believed on substantial grounds to exist, of torture. . . . [Emphasis added.]

mirror closely the words in Article 3 of the Convention Against Torture:

1. . . .where there are substantial grounds for believing that he would be in danger of being subjected to torture. [Emphasis added.]

Because the words used in Article 3 and paragraph 97(1)(a) are almost identical and because paragraph 97(1)(a) was adopted by Parliament to give effect to Article 3, the jurisprudence that interprets Article 3 is of assistance in interpreting paragraph 97(1)(a).

[19]The relevant words in Article 3 were interpreted by this Court in Suresh v. Canada (Minister of Citizenship and Immigration), [2000] 2 F.C. 592 (C.A.) reversed on other grounds, [2002] 1 S.C.R. 3. At paragraph 152, Robertson J.A. stated:

If we reject the two extreme threshold tests, "mere possibility" and "highly probable", we are left with the intermediate standard framed in terms of a "balance of probabilities". That threshold can be conveniently recast by asking whether refoulement will expose a person to a "serious" risk of torture.

Robertson J.A. rejected the lower "mere possibility" of torture test and the higher "highly probable" test for what he called the "balance of probabilities" test.

[20]General Comments of the United Nations Committee Against Torture and decisions of that Committee are consistent with the approach of Robertson J.A. In General Comment No. 01: Implementation of Article 3 of the Convention in the Context of Article 22 (U.N. Doc A/53/44, Annex IX), the Committee rejected both a low-level test--theory or suspicion, and a high-level test--highly probable. The Committee stated at paragraphs 6 and 7:

6. Bearing in mind that the State party and the Committee are obliged to assess whether there are substantial grounds for believing that the author would be in danger of being subjected to torture were he/she to be expelled, returned or extradited, the risk of torture must be assessed on grounds that go beyond mere theory or suspicion. However, the risk does not have to meet the test of being highly probable.

7. The author must establish that he/she would be in danger of being tortured and that the grounds for so believing are substantial in the way described, and that such danger is personal and present. All pertinent information may be introduced by either party to bear on this matter.

[21]In the Committee's decision in A.R. v. Netherlands, Communication No. 203/2002: Netherlands 21/11/2003 (CAT/C/31/D/203/2002), the Committee elaborated that the risk must be "foreseeable, real and personal." At paragraph 7.3, the Committee stated:

7.3 The Committee recalls its General Comment on article 3, which states that the Committee is to assess whether there are substantial grounds for believing that the author would be in danger of torture if returned, and that the risk of torture must be assessed on grounds that go beyond mere theory or suspicion. The risk need not be highly probable, but it must be personal and present. In this regard, in previous decisions, the Committee has determined that the risk of torture must be foreseeable, real and personal.

The Committee found that it is required to assess whether there are substantial grounds for believing that the individual would be in danger of torture, and that the risk must be personal, present, foreseeable and real, but not highly probable.

[22]This Court has found that use of the word "would" requires a showing of probability. See Jose Pereira E Hijos, S.A. v. Canada (Attorney General) (2002), 299 N.R. 154 (F.C.A.), at paragraph 14, per Stone J.A. and Chalk River Technicians and Technologists v. Atomic Energy of Canada Ltd., [2003] 3 F.C. 313 (C.A.), at paragraph 52. Had the Convention used the words "could", "might", or "may", I think a lower-level test might be implied. But the word "would" in the Convention, together with the other words used by the Committee, imply that the Committee adopted a probability test.

[23]In Selvaratnam v. Ashcroft, 81 Fed. Appx. 907 (9th Cir. 2003), the United States Court of Appeals for the Ninth Circuit adopted the "more likely than not" test in interpreting Article 3 of the Convention Against Torture. The words "balance of probabilities" may be expressed as "more probable than not" (see Miller v. Minister of Pensions, [1947] 2 All E.R. 372 (K.B.D.), at page 374, per Denning J. (as he then was)) or the equivalent--more likely than not.

[24]Although the word formulas used by the Committee and the Ninth Circuit vary, and although their decisions are not binding on this Court, their approaches are consistent with the approach of Robertson J.A. in Suresh, namely, that the test is balance of probabilities.

[25]The approach of Robertson J.A. in Suresh implies that on a theoretical spectrum of tests of the meaning of "substantial grounds for believing" in Article 3 of the Convention Against Torture, there are three possible tests: i.e. mere possibility, balance of probabilities and highly probable. Mr. Li submits that the Adjei test relating to section 96 "reasonable chance [of] persecution", i.e. more than a minimal possibility but less than a probability, should be added to the spectrum. Thus, on the theoretical spectrum, there could be four tests: mere possibility, reasonable chance, balance of probabilities and highly probable.

[26]Turning to paragraph 97(1)(a), there is no authority to suggest that the highly probable or the mere possibility tests might apply and they can be summarily rejected. That leaves the reasonable chance (Adjei) and balance of probabilities (Suresh) tests as potential meanings of the words

97. (1) . . . would subject them personally

(a) to a danger, believed on substantial grounds to exist, of torture

in paragraph 97(1)(a) of the Act.

[27] In my opinion, Suresh is dispositive of the test for danger of torture under paragraph 97(1)(a). Because the words in paragraph 97(1)(a) and Article 3 are almost identical and deal with the same subject-matter, they should be interpreted the same way. In Suresh, this Court interpreted the words in Article 3 of the Convention Against Torture as meaning balance of probabilities. Parliament enacted paragraph 97(1)(a) after this Court's decision in Suresh. When a statutory provision appears to be modelled on existing legislation, whether from the same or another jurisdiction, interpretation of the model legislation is presumed to have known and taken into account in drafting the new legislation (see R. Sullivan, Sullivan and Driedger on the Construction of Statutes, 4th ed. (Toronto: Butterworths, 2002), at page 509, and Lyons et al. v. The Queen, [1984] 2 S.C.R. 633, at pages 687-689).

[28]I see no reason why this principle would not apply where the model is a Convention of the United Nations and the statutory provision in question was adopted to give effect to Canada's obligation as a signatory to the Convention. The Debates of the House of Commons indicate that Parliament was aware of the Suresh case: see House of Commons Debates, 080 (18 September 2001), 5525. In the proceedings of the Standing Senate Committee on Human Rights, there is express reference to the Federal Court of Appeal decision in Suresh (see Standing Senate Committee on Human Rights, Proceedings: Issue No. 3, 37th Parl. (24 September 2001)). Parliament had an opportunity to enact a lower-level test in paragraph 97(1)(a), notwithstanding this Court's decision in Suresh. It did not do so. In these circumstances, it is reasonable to infer that the Parliamentary intention is consistent with this Court's decision in Suresh and that the test for the degree of danger of torture in paragraph 97(1)(a) is a balance of probabilities or more likely than not.

Distinguishing the Standard of Proof and the Test under Paragraph 97(1)(a)

[29]It is immediately apparent that the words used to describe the standard of proof--balance of probabilities --are equivalent to the words used to describe the legal test to be met in order to be entitled to protection under paragraph 97(1)(a)--more likely than not. Although the words are equivalent, there are two distinct steps involved. Proof on a balance of probabilities is the standard of proof the panel will apply in assessing the evidence adduced before it for purposes of making its factual findings. The test for determining the danger of torture is whether, on the facts found by the panel, the panel is satisfied that it is more likely than not that the individual would personally be subjected to a danger of torture.

Mr. Li's Arguments

[30]Mr. Li argues that it makes no "rational sense" to adopt a higher degree of danger test for protection against torture than for determining Convention refugee status (as determined in Adjei), that different tests constitute discrimination because the law will not provide equal protection for everyone and that a higher standard offends the Convention Against Torture.

[31]I do not lightly dismiss the "no rational sense" argument. Indeed, paragraph 95(1)(b) of the Act confers "refugee protection" on a person determined to be a Convention refugee or a person in need of protection. Paragraph 95(1)(b) states:

95. (1) Refugee protection is conferred on a person when

. . .

(b) the Board determines the person to be a Convention refugee or a person in need of protection;

There is therefore an argument to be made that the context suggests that the same test should apply to section 96 and paragraph 97(1)(a). On the other hand, the Act provides two distinct processes to a common result and it is not obvious that the tests for risk of persecution and danger of torture must be identical.

[32]If Parliament intended the same test be applied, the words in section 96 "well-founded fear" could have been used in paragraph 97(1)(a). But Parliament chose not to do so, instead adopting words in paragraph 97(1)(a) that mirror the words in Article 3 of the Convention Against Torture. Prima facie the adoption of different words in paragraph 97(1)(a) than in section 96 suggests a different meaning. Of course, the "words-in-context" approach remains determinative and might lead to a different conclusion. See, for example, Bell ExpressVu Limited Partnership v. Rex, [2002] 2 S.C.R. 559, at paragraph 26.

[33]It is true that at a refugee hearing a panel may be asked to consider both whether an individual is a Convention refugee and whether that individual is in need of protection. Some of the evidence may apply to both determinations. However, there are differences between section 96 and paragraph 97(1)(a). For example, a claim for protection under paragraph 97(1)(a) is not predicated on the individual demonstrating that he or she is in danger of torture for any of the enumerated grounds of section 96. Further, there are both subjective and objective components necessary to satisfy the requirements of section 96: see Chan v. Canada (Minister of Employment and Immigration), [1995] 3 S.C.R. 593, at paragraph 120, per Major J., while a claim under paragraph 97(1)(a) has no subjective component. Because of such differences, it cannot be said that the provisions are so closely related that it would be irrational if the test under paragraph 97(1)(a) was not identical to the test under section 96.

[34]Mr. Li's discrimination argument was not well developed. There was no identification of an enumerated or analogous ground of discrimination. There is no basis for this Court to conclude that different tests under different statutory provisions constitute some form of discrimination.

[35]Finally, it is not easy to understand why a test for degree of danger being "more likely than not" offends the Convention Against Torture. The United Nations Committee Against Torture itself has adopted a test for purposes of Article 3 of the Convention Against Torture that, for the reasons I have given is, for all intents and purposes, the "more likely than not" test.

[36]In the result, the answer to the second question is:

The requisite degree of danger of torture envisaged by the expression "believed on substantial grounds to exist" is that the danger of torture is more likely than not.

QUESTION No. 3

[37]The final certified question is:

Is the same degree of risk required under paragraph 97(1)(b)?

[38]Mr. Li says that the reasonable-chance test should apply to paragraph 97(1)(b). However, there are no words that qualify the term "risk" in paragraph 97(1)(b) or that suggest the test in section 96 should apply to paragraph 97(1)(b). In the absence of some compelling reason suggesting a particularly low or a particularly high-level test, I do not see why the degree of risk for purposes of paragraph 97(1)(b) should not be that it is more likely than not that the individual would be subjected, personally, to a risk to his life or to a risk of cruel and unusual treatment or punishment if the person was returned to his country of nationality.

[39]The answer to the third question is:

The degree of risk under paragraph 97(1)(b) is that the risk is more likely than not.

CONCLUSION

[40]Having agreed with the analysis and conclusion of Gauthier J., I am of the opinion that the appeal should be dismissed.

Noël J.A.: I agree.

Malone J.A.: I agree.

 You are being directed to the most recent version of the statute which may not be the version considered at the time of the judgment.