Date: 20201210

Docket: A-206-19

Citation: 2020 FCA 213

Heard by online video conference hosted by the registry on October 22, 2020.

Judgment delivered at Ottawa, Ontario, on December 10, 2020.

Date: 20201210

Docket: A-206-19

Citation: 2020 FCA 213

REASONS FOR JUDGMENT

LOCKE J.A.

I.  Overview

[1]  This decision concerns an appeal from a decision of the Tax Court of Canada (2019 TCC 121, per Pizzitelli J.) which dismissed the appellant’s motion to compel the respondent, as part of discovery, to provide certain documents and answers to certain questions. The appellant’s motion was in the context of an appeal before the Tax Court in which the parties agree that the only issue in dispute concerns the interpretation of a statutory provision. Specifically, the question is whether trading activities of the kind carried on by the appellant constitute carrying on a business so as to require the payment of tax despite the appellant being a Tax-Free Savings Account (TFSA) trustee.

[2]  The appellant alleges many errors by the Tax Court, but not all of these have to be addressed in order to dispose of this appeal.

[3]  The parties agree, and I concur, that the standard of review in this appeal is as contemplated in Housen v. Nikolaisen, 2002 SCC 33, [2002] 2 S.C.R. 235 (Housen). The standard of correctness applies to questions of law (see Housen at para. 8), but findings of fact or of mixed fact and law, unless there is an extricable question of law, are reviewable only where the court below has made a palpable and overriding error (see Housen at paras. 10 and 36).

[4]  The appellant adds that this Court should not defer to a motions judge on findings of fact or mixed fact and law where that judge has failed to demonstrate impartiality. The appellant argues that the Tax Court gave rise to a reasonable apprehension of bias by (i) basing its decision on a number of issues that had not been raised or addressed by the parties, and (ii) awarding costs in any event of the cause.

II.  Bias

[5]  I will deal with the bias argument first. I start by noting that bias allegations do not lend themselves to a standard of review analysis at all: Canadian Pacific Railway Company v. Canada (Attorney General), 2018 FCA 69 at para. 55. If an appellant is successful in showing that a lower court’s decision gave rise to a reasonable apprehension of bias, then the appeal will be allowed based on a breach of natural justice arising from that bias. On the other hand, if an appellant is not successful in such an argument, then bias cannot be the basis of a successful appeal. Either way, deference is not a factor. Parties are simply entitled to an impartial decision.

[6]  In any case, the appellant’s arguments on bias come nowhere near what is required. The appellant recognizes its legal burden, which is to show that an informed person, viewing the matter realistically and practically – and having thought the matter through – would think that it is more likely than not that the decision-maker, whether consciously or unconsciously, would not decide fairly (Wewaykum Indian Band v. Canada, 2003 SCC 45, [2003] 2 S.C.R. 259 at para. 60; Yukon Francophone School Board, Education Area #23 v. Yukon (Attorney General), 2015 SCC 25, [2015] 2 S.C.R. 282 at paras. 20-21, 26; Badawy v. 1038482 Alberta Ltd., 2019 FCA 150 at para. 21).

[7]  The appellant’s assertions concerning issues not raised or addressed by the parties and costs awarded in any event of the cause are clearly insufficient. If the Tax Court had erred on either of these points, such an error might be a basis to allow the appeal, but it would not, without more, suggest bias. Moreover, the appellant’s list of seven findings and rulings by the Tax Court that were not subject to argument (see paragraph 38 of the appellant’s memorandum of fact and law) is unconvincing. Most of those points were not determinative and were unnecessary to the Tax Court’s decision. For example, as discussed below, the appellant mischaracterizes the Tax Court’s comments concerning the limits of its power to review a decision to redact portions of documents produced under the Access to Information Act, R.S.C. 1985, c. A-1 (AIA). Moreover, those comments were not the basis of the Tax Court’s decision to refuse to order the production of unredacted versions of those documents.

[8]  I agree with the respondent, and the appellant acknowledges, that bias allegations should not be undertaken lightly because they call into question not simply the personal integrity of the judge, but the integrity of the administration of justice (R. v. S.(R.D.), [1997] 3 S.C.R. 484, 151 D.L.R. (4th) 193 at para. 113). I also agree with the respondent that bias allegations should not have been made in this appeal.

III.  Analysis of Substantive Issues on Appeal

[9]  My discussion of the substantive issues is not organized in the same way as the appellant has organized its arguments, but it addresses all of the discovery questions in issue.

[10]  I preface this analysis by noting an overarching argument by the appellant that the errors allegedly made by the Tax Court were extricable errors of law, and hence subject to review on a standard of correctness. The appellant argues that, though the Tax Court correctly stated the legal test to be applied, it failed to apply that test. I disagree that this case concerns any extricable error of law. Rather, the appellant’s arguments are effectively that the Tax Court erred in its application of the law to the facts of this case. I am not convinced that the Tax Court failed to apply the legal test described in its decision, or applied a different legal test. It follows that the standard of palpable and overriding error applies.

A.  Public Documents

[11]  The Tax Court found that certain of the documents sought by the appellant are in the public domain, and refused to order production for that reason, since such an order would amount to requiring the respondent to do the appellant’s research.

[12]  The appellant argues that the documents said to be available to the public are not, in fact, publicly available since the appellant has been unable to obtain the documents despite diligent efforts. The appellant does not argue that such documents are confidential, or impossible to obtain. For example, question 11 asserts that the documents sought therein are “very difficult to obtain.” Also, question 14 argues simply that “[i]t may be easier for the respondent to locate” the documents sought therein.

[13]  In my view, such documents are publicly available even if they are allegedly difficult to obtain. The Tax Court did not err in characterizing the appellant’s request as an effort to have the respondent do its research, and in refusing to order production of such documents.

[14]  This section addresses the appellant’s questions 11-14.

B.  Non-Public Documents

[15]  The Tax Court also refused to order production of, and answers to questions about, internal Department of Finance documents on the basis, among other things, that they are of very little or no relevance to the statutory interpretation issue in the underlying appeal before the Tax Court.

[16]  The discussion in this section addresses the appellant’s questions 24-30, 36, 38 and 39. Questions 24-30 all concern a redacted table that the appellant obtained pursuant to a request under the AIA. The table is entitled “Comparison of tax characteristics of proposed LSP and DSP and existing RRSP and RESP” and is dated January 19, 2007. The appellant notes that LSP later came to be referred to as TFSA when they were introduced in the Budget Implementation Act, 2008, S.C. 2008, c. 28, which was tabled on February 26, 2008, and received Royal Assent on June 18, 2008. The appellant focuses on the third page of the table, and a row entitled “Carrying on a business.” On that row, entries under the columns headed “LSP” and “RRSP” read, respectively, “Exemption applies only to income derived from investing of funds, thus unrelated business income would be taxable” and “Taxable on unrelated business income.” Questions 24-30 are reproduced here:

24.  Does this table accurately reflect the Department of Finance’s policy in January 2007 on the kind of business income that is taxable or exempt to a TFSA and an RRSP?

25.  If so, what is meant by “unrelated business income”?

26.  The table seems to imply that “related business income” is exempt. If so, what is meant by related taxable income?

27.  Does this table accurately reflect the Department of Finance’s (and Parliament’s) legislative intent and policy in relation to the kind of business income that is taxable or exempt to a TFSA and an RRSP when section 246.2(6) [sic, 146.2(6)] was enacted?

28.  If the table does not accurately reflect such policy, or the policy was subsequently changed as a matter of fact, when was that policy changed?

29.  As a matter of fact, what policy replaced it?

30.  If the policy was changed please produce any document produced by or in the possession of the Department of Finance which supports such a change (including, but not limited to memoranda or other guidance provided by the Policy Division of the Department of Finance).

[17]  By these questions, the appellant asks about the meaning of certain references in the table, whether the table reflects the “legislative intent and policy” of the Department of Finance and Parliament, any changes to such policy, and documents related to any such changed policy.

[18]  The remaining questions in issue in this section, questions 36, 38 and 39, seek (i) an unredacted version of this and other documents obtained pursuant to a request under the AIA, and (ii) admissions that said documents were made in the usual and ordinary course of business.

[19]  On the overarching issue of relevance, the Tax Court correctly noted various general principles applicable to discovery, including (i) that it should be broadly and liberally construed, (ii) that the threshold is lower in discovery than at trial, (iii) that earlier drafts of a final position paper do not have to be disclosed, and (iv) that even where relevance is established, the Court has a residual discretion to refuse document production.

[20]  The appellant argues that the redacted internal Department of Finance documents are relevant to statutory interpretation, and should be produced in unredacted form. The appellant argues that documents prepared by government employees participating in the legislative process are admissible as permissible extrinsic aids. The appellant argues that relevance is not limited to documents that are published or otherwise available to the public.

[21]  The Tax Court based its finding that the internal documents in question are of marginal relevance on Superior Plus Corp. v. Canada, 2016 TCC 217 at para. 34, which provides such documents are not relevant to ascertaining the Minister’s mental process in auditing and assessing a taxpayer, unless they have been communicated to the Minister. The respondent argues that the Tax Court was correct to apply the same reasoning to statutory interpretation: internal finance documents that have not been communicated to the Minister are not relevant to ascertaining Parliamentary intent.

[22]  It is tempting to follow this reasoning and to agree with the respondent’s position that documents must be publicly available in order to be relevant to statutory interpretation. Otherwise, it would be possible for members of the public to be left without access to certain information that is necessary to fully understand a particular law with which they are required to comply. Such a situation would be problematic for the reasons mentioned in Pepper (Inspector of Taxes) v. Hart, [1992] 3 W.L.R. 1032 at 1042 (U.K.H.L.):

A statute is, after all, the formal and complete intimation to the citizen of a particular rule of the law which he is enjoined, sometimes under penalty, to obey and by which he is both expected and entitled to regulate his conduct. We must, therefore, I believe, be very cautious in opening the door to the reception of material not readily or ordinarily accessible to the citizen whose rights and duties are to be affected by the words in which the legislature has elected to express its will.

[23]  Notwithstanding this concern, the appellant argues that the scope of documents that could be relevant to statutory interpretation is viewed more broadly. For example, the appellant cites Delisle v. Canada (Attorney General), [1999] 2 S.C.R. 989 (Delisle), which concerned an argument that a provision of a federal statute violated the Canadian Charter of Rights and Freedoms, Part I of The Constitution Act, 1982, being Schedule B to the Canada Act 1982 (U.K.), 1982, c 11. As part of his analysis on behalf of the majority of the Court in Delisle, Bastarache J. considered the purpose of the statutory provision in question in the course of interpreting it. At para. 17, he stated as follows:

[…]Although extrinsic sources may be used to interpret legislation and to determine its true meaning, when the meaning of the challenged provision is clear, they are of little assistance in determining the purpose of a statute in order to evaluate whether it is consistent with the Charter. Generally, the Court must not strike down an enactment which does not infringe the Charter in its meaning, form or effects, which would force Parliament to re-enact the same text, but with an extrinsic demonstration of a valid purpose. That would be an absurd scenario because it would ascribe a direct statutory effect to simple statements, internal reports and other external sources which, while they are useful when a judge must determine the meaning of an obscure provision, are not sufficient to strike down a statutory enactment which is otherwise consistent with the Charter. Legislative intent must have an institutional quality, as it is impossible to know what each member of Parliament was thinking. It must reflect what was known to the members at the time of the vote. It must also have regard to the fact that the members were called upon to vote on a specific wording, for which an institutional explanation was provided. The wording and justification thereof are important precisely because members have a duty to understand the meaning of the statute on which they are voting. This is more important than speculation on the subjective intention of those who proposed the enactment. (emphasis added)

[24]  This passage recognizes the potential relevance of internal documents to the interpretation of “obscure” statutory provisions. It is not clear what constitutes “obscure”, and I do not reach a conclusion on this point. I note that this passage does not state clearly whether internal, non-public documents can be relevant to statutory interpretation. In fact, the focus on “what was known to the members [of Parliament] at the time of the vote,” suggests that non-public documents are not relevant.

[25]  Ruth Sullivan, in Sullivan on the Construction of Statutes, 6th ed., (Toronto: LexisNexis, 2014) at §23.11 casts a broad net for the types of documents that can be relevant to statutory interpretation:

Like evidence of external context, opinions about the purpose and meaning of legislation can be found anywhere: before enactment, in the materials generated by government employees participating in the legislative process (instructing officers, drafters, legal opinion givers) and, after enactment, in interpretive guidelines issued by administrative agencies, in judicial or administrative case law and in the daily decisions of government employees charged with administering the legislation. Until recently, the primary source of opinion about the meaning of legislation was judicial case law. Courts were unwilling to look at the practice of bureaucrats or the opinions of administrative tribunals and, except for standard textbooks, scholarly opinion was largely ignored. The current tendency, however, is to look at any material that meets the threshold test of relevance and reliability.

[26]  Again, this passage does not state clearly that non-public documents can be relevant to statutory interpretation. However, it does appear that the legislative process (during which relevant documents could be created) begins early. In Mikisew Cree First Nation v. Canada, 2018 SCC 40, [2018] 2 S.C.R. 765 (Mikisew) at para. 120, Brown J. stated that “the legislative process begins with a bill’s formative stages, even where the bill is developed by ministers of the Crown.” Brown J went on in paragraph 121 to state

Public servants making policy recommendations prior to the formulation and introduction of a bill are not “executing” existing legislative policy or direction. Their actions, rather, are directed to informing potential changes to legislative policy and are squarely legislative in nature.

[27]  Care must be taken not to read Mikisew too broadly. That case concerned whether the law-making process (described at paragraph 116 thereof as the steps from initial policy development to royal assent) was subject to the Crown’s duty to consult indigenous peoples about steps that could adversely affect their rights. Mikisew was not concerned with statutory interpretation.

[28]  Sullivan, relying on the Newfoundland Court of Appeal decision in Reference re Upper Churchill Water Rights Reversion Act (1982), 134 D.L.R. (3d) 288, 36 Nfld. & P.E.I.R. 273, rev’d [1984] 1 S.C.R. 297 (Upper Churchill), goes on at §23.13 to state:

When the purpose of a provision is discussed or its meaning explained during the enactment process, and the legislation is then passed on that understanding, the explanation or discussion offers persuasive (if not conclusive) evidence of the legislature's intent.

[29]  However, the Supreme Court of Canada in Upper Churchill offered a more nuanced approach to the relevance of extrinsic evidence. After discussing the relaxation of the former general exclusionary rule against admissibility of extrinsic evidence, the Court stated at p. 318:

It will therefore be open to the Court in a proper case to receive and consider extrinsic evidence on the operation and effect of the legislation. In view of the positions of the parties, particularly the appellants’ contention that the Reversion Act has extra-provincial effect, this is, in my opinion, such a case.

I agree with the Court of Appeal in the present case that extrinsic evidence is admissible to show the background against which the legislation was enacted. I also agree that such evidence is not receivable as an aid to construction of the statute. However, I am also of the view that in constitutional cases, particularly where there are allegations of colourability, extrinsic evidence may be considered to ascertain not only the operation and effect of the impugned legislation but its true object and purpose as well. This was also the view of Dickson J. in the Reference re Residential Tenancies Act, 1979, [[1981] 1 S.C.R. 714], at p. 721, where he said:

In my view a court may, in a proper case, require to be informed as to what the effect of the legislation will be. The object or purpose of the Act in question may also call for consideration though, generally speaking, speeches made in the Legislature at the time of enactment of the measure are inadmissible as having little evidential weight.

This view is subject, of course, to the limitation suggested by Dickson J., at p. 723 of the same case, that only evidence which is not inherently unreliable or offending against public policy should be admissible…

[30]  Not only does the Supreme Court leave room for cases where extrinsic evidence will not be relevant, but it also limits the issues to which such evidence might be relevant. Moreover, it should be noted that Upper Churchill was in a constitutional law context, in which the Supreme Court has traditionally been more open to extrinsic evidence (see p. 317).

[31]  In the end, though there are good reasons to be reluctant to consider non-public documents in the exercise of statutory interpretation, it is difficult to state unequivocally that such documents could never be relevant. The better question is whether the documents in question in the present appeal have an institutional quality such that they could represent the government’s position concerning the legislation at issue. If not, such documents are not relevant.

[32]  It is important to bear in mind that this appeal concerns whether the Tax Court made a palpable and overriding error in applying the law to the facts. Accordingly, this Court owes deference to the Tax Court on issues of mixed fact and law.

[33]  It is also important to bear in mind that jurisprudence must always be considered in its proper factual context. It is helpful to observe that questions 24-30 in this case concern a document that predates the legislative provision of interest. That document is hence similar to the “earlier drafts of a final position paper” which do not have to be disclosed (see paragraph 19 above).

[34]  The Tax Court was equivocal about the relevance of the internal documents in question. At paragraph 18 of its decision, in dealing with discovery question 39, it concluded that they “have very little or no relevance.” The Tax Court did not state here that the documents in question had no relevance whatsoever. Therefore, I understand the Tax Court’s refusal to order production of the unredacted documents sought in question 39 to be an exercise of its residual discretion to refuse document production even when the documents in question may have marginal relevance.

[35]  Later, the Tax Court reached the same conclusion at paragraph 34 concerning questions 24-30. Here, the Tax Court did not equivocate as it did at paragraph 18; it simply stated that the internal documents in question were “irrelevant.” However, the Tax Court stated that it was adopting the same reasoning as in paragraph 18, and I take that to include the equivocation. Finally, at paragraphs 37 and 38, the Tax Court reached the same conclusion about questions 36 and 38, agreeing with the respondent’s position that the documents in question were “irrelevant.” Again, I take the Tax Court’s conclusion to include the equivocation of paragraph 18.

[36]  In my view, the Tax Court made no palpable and overriding error in refusing to order production of unredacted copies of the internal documents in question. Nor did the Tax Court make any palpable and overriding error in refusing to order that the respondent answer questions related to such documents.

[37]  Before concluding this section, I wish to address two aspects of the Tax Court’s reasons. First, the appellant argues that paragraph 12 suggests that the Tax Court does not have the power to overrule the Information Commissioner’s refusal to order production of an unredacted version of a document that has been obtained, with redactions, by means of an AIA request. That is not what the Tax Court stated, and it is incorrect. The criteria applicable to production of documents during discovery are quite distinct from those applicable to disclosure of documents under the AIA. It is clear from the Tax Court’s reasons that it was of the view that the internal documents in question were of marginal relevance. As indicated above, this was a proper basis to refuse production. The reference to challenging the redaction of documents with the Information Commissioner seems to have been simply a suggested alternative approach that the appellant could try.

[38]  The second aspect of the Tax Court’s reasons that I wish to address concerns paragraph 38. There, the Tax Court referred to “earlier versions” of certain documents that the appellant seeks. The appellant asserts that it never requested earlier versions of any documents. Again, the conclusion concerning the marginal relevance of internal documents remains, and that was a proper basis for refusing questions thereon. If the Tax Court was confused about the specific internal documents in issue, it makes no difference here. Even if such an error were palpable, it would not be overriding. It is clear that the Tax Court would have refused to order the questions answered regardless of whether or not they were earlier versions. In any case, I believe that the Tax Court was not confused. As indicated in paragraph 33 above, I believe that it was referring to old documents discussing possible policies that might be considered “earlier drafts of a final position paper,” and which need not be produced.

C.  Legal Position vs. Legal Argument

[39]  The Tax Court refused to order that the respondent provide information concerning factual assumptions surrounding the object, spirit and purpose and the policy behind the statutory provisions in issue. It based itself, in part, on the view that the appellant was seeking more than the respondent’s legal position, but rather information concerning the respondent’s legal argument (to which it is not entitled in discovery). The Tax Court was satisfied that the respondent’s legal position was already clear.

[40]  The appellant objects arguing that the information it seeks concerns the respondent’s legal position and is relevant.

[41]  In my view, the Tax Court did not err either in observing that factual assumptions are matters for pleading, not discovery, or in finding that the respondent has already communicated its legal position. The appellant has not convinced me that the Tax Court made a palpable and overriding error or erred on an extricable question of law in concluding that the respondent had already made its position clear and that ordering answers to the questions in issue would invite legal argument.

[42]  Paragraph 27 of the Tax Court decision (as well as paragraph 18 thereof) cites an obiter dicta statement in Madison Pacific Properties Inc. v. Canada, 2019 FCA 19 at para. 28 (MP Properties) in support of the conclusion that the questions in issue lacked relevance because the statutory interpretation issue in dispute in the underlying appeal before the Tax Court is a question of law, not a question of fact:

[27] […]The statutory interpretation of these sections is a question of law, and not a matter of fact and is for the Court to ultimately determine at trial as referenced in MP Properties by Gleason J. at paragraph 28 above referred to, the Appellant is entitled to know the Respondent’s position on the law, but not its [sic] evidence it relies on nor its legal argument.[…]

[43]  In my view, the statements in MP Properties and by the Tax Court should not be taken as an indication that a party is not entitled during discovery to obtain, and ask questions about, relevant documents simply because they concern a question of law. Rather, I understand these statements to mean that a party may not use discovery to question an opposing party as to which of the documents on the record will be relied on for which legal arguments. Such questions essentially seek the opponent’s legal argument rather than its legal position.

[44]  The discussion in this section addresses the appellant’s questions 1, 2, 10 and 19.

D.  Draft Statement of Agreed Facts

[45]  The Tax Court refused to order that the respondent indicate any facts described in the draft Statement of Agreed Facts that it disputes. In doing so, the Tax Court noted the appellant’s own position that the facts at issue are already admitted by the respondent. The Tax Court also noted that the draft Statement of Agreed Facts is privileged and confidential, and further that it was not proper to ask whether the respondent refutes any question amongst many.

[46]  In my view, the appellant’s assertion that all of the facts included in the draft Statement of Agreed Facts have already been admitted in the pleadings is sufficient to dismiss this aspect of the appeal (see paragraphs 83 and 125 of its memorandum). In light of this acknowledgement, it follows that the appellant does not point to any facts in the draft Statement of Agreed Facts that are in dispute, which would be necessary to justify requiring the respondent to answer. Reviewing the pleadings carefully, it appears that the respondent does not actually admit everything in the draft Statement of Agreed Facts. Though most of the statements in the document are admitted, it seems that a few are not. Therefore, the appellant’s assertion that all of the facts in the document are admitted appears to be untrue. In any case, this does not change the fact that it was appropriate to refuse to order an answer to the appellant’s question since the appellant does not assert any unadmitted facts which could justify such an order.

[47]  Even though it is not necessary to address this point, I note that the appellant argues that the Tax Court erred at paragraph 22 in stating:

A party should not be put in the position on discovery of having to recall by memory all of the facts that may have been admitted in the pleadings.

[48]  The appellant suggests that the reference to “recall by memory” indicates that the Tax Court failed to understand that the discovery in issue was in writing (not in person) and that the Minister had abundant time to prepare its answers. I disagree. I do not believe that the Tax Court misunderstood the circumstances of the discovery. Paragraphs 1 and 12 of the reasons clearly recognize that the discovery was in writing. Reading paragraph 22 as a whole, I believe that the Tax Court was concerned with the burden on the respondent of having to answer such a broad question with myriad potential implications.

[49]  The discussion in this section addresses the appellant’s question 31.

IV.  Conclusion

[50]  For the foregoing reasons, I would dismiss the present appeal with costs.

[51]  The parties agreed on the amounts of their respective costs. Accordingly, I would set the respondent’s costs of this appeal at an all-inclusive amount of $2,750.

"George R. Locke"

“I agree.

Johanne Gauthier J.A.”

“I agree.

Yves de Montigny J.A.”

FEDERAL COURT OF APPEAL

NAMES OF COUNSEL AND SOLICITORS OF RECORD

APPEARANCES:

SOLICITORS OF RECORD: