Citation: 2017 FCA 79
CANADIAN NATIONAL RAILWAY COMPANY
EMERSON MILLING INC. and CANADIAN TRANSPORATION AGENCY
Heard at Vancouver, British Columbia, on September 19, 2016.
Judgment delivered at Ottawa, Ontario, on April 18, 2017.
REASONS FOR JUDGMENT BY:
CONCURRED IN BY:
Citation: 2017 FCA 79
CANADIAN NATIONAL RAILWAY COMPANY
EMERSON MILLING INC. and CANADIAN TRANSPORATION AGENCY
REASONS FOR JUDGMENT
Crops are harvested and delivered to the facilities of companies like Emerson Milling Inc. To handle the arriving crops, Emerson orders railcars from the Canadian National Railway Company. Where Emerson has
“traffic offered for carriage,” CN must supply railcars and then
“without delay, and with due care and diligence, receive, carry and deliver the traffic”: subsection 113(1) of the Canada Transportation Act, S.C. 1996, c. 10. Normally there is no controversy: the railcars are ordered, the railcars arrive, the crops are loaded onto the railcars and CN transports them away.
 But there was controversy in 2013-2014. As usual, growers delivered crops to Emerson’s facility. As in previous years, Emerson periodically ordered railcars from CN to transport the crops. But CN delivered only some of the cars, not all. 2013-2014 was a bumper crop year. Also the winter of 2014 was extremely cold, restricting some of CN’s operations.
As it is entitled to do under the Act, Emerson complained to the Canadian Transportation Agency. It alleged that CN had failed to receive, carry and deliver
“traffic offered for carriage” and, thus, violated subsection 113(1) of the Act. CN alleged that Emerson’s orders for cars were unreasonable and that in the circumstances CN acted as reasonably as it could under challenging circumstances.
 In a decision dated July 10, 2015 (case no. 14-06408), the Agency sided with Emerson. It ordered CN to provide to Emerson the railway cars that Emerson asked for and still required in order to satisfy its customers.
 CN appeals, with leave of this Court, from the Agency’s decision. For the reasons set out below, I would dismiss the appeal with costs.
A. Preliminary issue: the requirement in subsection 41(1) of the Act that there be a “question of law” or a “question of jurisdiction”
 Emerson submits that subsection 41(1) bars CN’s appeal in whole or in part because CN’s appeal does not raise a question of law or a question of jurisdiction.
Under subsection 41(1) of the Act,
“[a]n appeal lies from the Agency to the Federal Court of Appeal on a question of law or a question of jurisdiction on leave to appeal being obtained from that Court.” Among other things, this means that this Court must be satisfied that an appellant has raised a
“question of law” or a
“question of jurisdiction” before it can entertain the appeal.
 We usually deal with this sort of submission on a preliminary basis before delving into the merits of an appeal: see, e.g., Canadian National Railway Company v. Dreyfus, 2016 FCA 232 at para. 18; Canadian National Railway Company v. BNSF Railway Company, 2016 FCA 284. Often we follow this practice in other contexts where our subject-matter jurisdiction is in issue, especially where to decide the merits might invade the right to decide of another body that might have jurisdiction: National Indian Brotherhood v. Juneau (No. 2),  F.C. 73 (C.A.); see also, e.g., Pfizer Canada Inc. v. Teva Canada Limited, 2016 FCA 218, 141 C.P.R. (4th) 165. Often considerations of legality and practicality favour proceeding in this way—a pronouncement on the merits of the matter without jurisdiction is a nullity: P.E.I. (Provincial Secretary) v. Egan,  S.C.R. 396, 3 D.L.R. 305.
 This practice is prudent: putting aside narrow areas of inherent or plenary jurisdiction and the responsibility to develop and apply the common law, it has been accepted for at least a quarter of a millennium that courts can act only within the limits of the law set by the legislator: see, e.g., Green v. Rutherford (1750), [1558-1774] All E.R. Rep. 153, 1 Ves. Sen. 462 at page 471; Penn v. Lord Baltimore (1750), [1558-1774] All E.R. Rep. 99, 1 Ves. Sen. 444 at page 446; A.G. v. Lord Hotham (1827), [1814-23] All E.R. Rep. 448, 3 Russ. 415; Thompson v. Sheil (1840), 3 Ir. Eq. R. 135. And of even longer standing is the principle of legislative supremacy, one corollary of which is that laws bind courts, just like everyone else: Re Resolution to Amend the Constitution,  1 S.C.R. 753 at pp. 805-806; Reference re Secession of Quebec,  2 S.C.R. 217, 161 D.L.R. (4th) 385 at paras. 71-72; Reference re Remuneration of Judges of the Provincial Court of Prince Edward Island,  3 S.C.R. 3, 150 D.L.R. (4th) 577 at para. 10.
The only exceptions are where a legislative limit is unconstitutional or the rule of law justifies court intervention: Crevier v. A.G. (Québec) et al.,  2 S.C.R. 220, 127 D.L.R. (3d) 1; Immeubles Port Louis Ltée. v. Lafontaine (Village),  1 S.C.R. 326 at p. 360. The latter, once described as
“a fundamental postulate of our constitutional structure” that
“lie[s] at the root of our system of government,” is now expressly set out as an operative principle in our Constitution: Roncarelli v. Duplessis,  S.C.R. 121 at p. 142; British Columbia (Attorney General) v. Christie, 2007 SCC 21,  1 S.C.R. 873 at para. 19; preamble to the Constitution Act, 1982, Schedule B to the Canada Act 1982 (U.K.), 1982, c. 11. It is the constitutional authorization for judicial review even in the face of legislative provisions restricting or forbidding it (e.g., so-called privative clauses): Dunsmuir v. New Brunswick, 2008 SCC 9,  1 S.C.R. 190 at paras. 27-28. Among other things, the rule of law provides that
“the law is supreme over officials” and
“thereby preclusive of the influence of arbitrary power”: Re Manitoba Language Rights,  1 S.C.R. 721 at p. 748; British Columbia v. Imperial Tobacco Canada Ltd., 2005 SCC 49,  2 S.C.R. 473 at paras. 57-58. Those who wield public power cannot be a law unto themselves, immunized from truly independent review and shielded from meaningful scrutiny: Thorson v. Attorney General of Canada,  1 S.C.R. 138 at p. 145; Canada (Attorney General) v. Downtown Eastside Sex Workers United Against Violence Society, 2012 SCC 45,  2 S.C.R. 524 at paras. 31-33, citing Canadian Council of Churches v. Canada (Minister of Employment and Immigration),  1 S.C.R. 236 at pp. 250-251; Paradis Honey Ltd. v. Canada, 2015 FCA 89, 382 D.L.R. (4th) 720 at para. 108 and Canada (Attorney General) v. Bri-Chem Supply Ltd., 2016 FCA 257 at para. 49, both citing Slansky v. Canada (Attorney General), 2013 FCA 199,  1 F.C.R. 81 at paras. 313-314.
For the purposes of subsection 41(1) of the Canada Transportation Act and sections worded like it, what is a
“question of law” and what is a
“question of jurisdiction”? To interpret these terms, we need to consider their plain meaning, their context within the Act and the purpose of subsection 41(1) and the Act itself: Re Rizzo & Rizzo Shoes Ltd.,  1 S.C.R. 27, 154 D.L.R. (4th) 193 and Bell ExpressVu Limited Partnership v. Rex, 2002 SCC 42,  2 S.C.R. 559.
 Under the Canada Transportation Act, the Agency is continued and empowered as a specialized regulator in the transportation sector. Its decisions are informed by understandings of how the sector operates and other specialized appreciations and policy considerations, such as the National Transportation Policy set out in section 5 of the Act. Indeed, under sections 24 and 43 of the Act, the Governor in Council can issue policy directions concerning any matter that comes within the jurisdiction of the Agency and the Agency must follow them. Appeals are not available for pure questions of fact (see section 31 of the Act). But appeals to the Governor in Council are available under section 40 of the Act; this provides a way to appeal, among other things, factually suffused and policy-imbued decisions of the Agency: Canadian National Railway Co. v. Canada (Attorney General), 2014 SCC 40,  2 S.C.R. 135 (“CN 2014”).
 From these provisions, one can see Parliament’s intention behind subsection 41(1): factually suffused and policy-imbued decisions are not to be appealed to this Court. Parliamentary debates also support this: CN 2014 at para. 46. Such questions can be appealed elsewhere. Instead, only matters turning on questions of law or questions of jurisdiction may be appealed to this Court with leave granted on the basis that there is an arguable issue: CKLN Radio Incorporated v. Canada (Attorney General), 2011 FCA 135, 418 N.R. 198; Rogers Cable Communications Inc. v. New Brunswick (Transportation), 2007 FCA 168, 367 N.R. 78. Given the terms of subsection 41(1), given the fact that a denial of leave is merits based, and given the availability of review under other sections of the Act for other questions, it would be hard to view subsection 41(1) as immunizing Agency decision-making in a problematic way.
What does a
“question of jurisdiction” mean? We begin with a bit of a conundrum. Today in administrative law we are often encouraged not to speak of jurisdiction. This trend has been underway since 1979 when Dickson J. (as he then was) warned against describing issues as jurisdictional when they are
“doubtfully so”: C.U.P.E. v. N.B. Liquor Corporation,  2 S.C.R. 227 at p. 233.
The reasoning goes like this. To say that an administrative decision-maker has jurisdiction to do something is to say that it has powers that have been granted to it expressly, impliedly or necessarily by legislation in certain circumstances or over certain subject-matters: Chrysler Canada Ltd. v. Canada (Competition Tribunal),  2 S.C.R. 394, 92 D.L.R. (4th) 609; Tranchemontagne v. Ontario (Director, Disability Support Program),  1 S.C.R. 513, 266 D.L.R. (4th) 287 at para. 16. For example, whether an agency can exercise a power to compel a witness to give testimony turns on what its statute says and how we interpret it—in reality a question of law. Thus, a
“question of jurisdiction” for the purposes of judicial review is really just a question of statutory interpretation, in other words a question of law: Halifax (Regional Municipality) v. Nova Scotia (Human Rights Commission), 2012 SCC 10,  1 S.C.R. 364; and see the detailed discussion in Canada (Border Services Agency) v. C.B. Powell Limited, 2010 FCA 61,  2 F.C.R. 332 at paras. 39-46.
 On this reasoning, many so-called questions of jurisdiction that are appealed under subsection 41(1) could easily be said today to be questions of law. Subsection 41(1) speaks of questions of law and questions of jurisdiction as if they are two different things. But it would seem that the latter is often just a subset of the former.
But, on closer examination, the phrase
“question of jurisdiction” in subsection 41(1) still adds something above and beyond the phrase
“question of law.” A bit of legislative history shines a light on this.
The Canada Transportation Act is a successor to various Acts stretching back to the National Transportation Act, S.C. 1966-67, c. 69, which was enacted in 1967. The phrase
“question of jurisdiction” in subsection 41(1) of the current Canada Transportation Act first appeared in 1971 as a requirement for appeals to this Court in subsection 64(2) of the National Transportation Act after it was amended by R.S.C. 1970 (2nd Supp.), c. 10. At that time, Parliament understood
“jurisdiction” to include failures of procedural fairness and other fundamental legal flaws: see, e.g., In re Ontario Labour Relations Board,  2 S.C.R. 18,  3 D.L.R. 561 (sometimes known as the Toronto Newspaper Guild case). Ever after, Parliament has decided to maintain
“question of jurisdiction” in the subsection even though, as mentioned, today
“question of jurisdiction” essentially means
“question of law” and
“question of law” is already in the subsection. This must mean something, as Parliament is not in the business of legislating redundancies: Nanaimo (City) v. Rascal Trucking Ltd., 2000 SCC 13,  1 S.C.R. 342 at para. 23.
Based on this legislative history, I conclude that
“question of jurisdiction” in subsection 41(1) includes at least issues of procedural fairness, even if those issues are factually suffused. Thus, under subsection 41(1) of the Act, a party may appeal on the basis that a decision of the Agency is procedurally unfair.
Now to the meaning of a
“question of law” under subsection 41(1) of the Act. Sometimes the Agency will state a pure question of law or a legal standard in its decision and then will resolve it. There is no doubt that such a question of law or legal standard can be the proper subject-matter of appeal under subsection 41(1) of the Act.
But sometimes the question of law or legal standard is mixed up with questions of fact. For example, the Agency might have a legal view of how a particular statutory provision works and rather than stating that view explicitly instead might proceed directly to its bottom-line conclusion. In reality, the Agency’s conclusion is an amalgam of its view of the law/legal standards and its view of the evidence, and how the former applies to the latter. In that context, where the law and the facts are mussed together, is there a
“question of law” for the purposes of subsection 41(1) of the Act?
 These questions of mixed fact and law are best seen on a spectrum. At one end are questions where the legal content is low and the result is driven by findings of facts or the adjudicator’s interpretation of the evidence as a whole. At the other end are questions where the legal content is high and the result is driven mainly by law/legal standards.
 The Supreme Court discussed this spectrum, albeit in a different context, in Housen v. Nikolaisen, 2002 SCC 33,  2 S.C.R. 235 at paras. 28 and 36. Housen concerns appellate review of decisions outside of the administrative law context and is not relevant to the administrative law context here. However, Housen is useful here because the Supreme Court was trying to solve the same question facing us here: given that questions of law/legal standards should be treated differently from questions of fact, how do we deal with questions of mixed fact and law that lie on a spectrum?
 In Housen, for the purposes of the appellate standard of review, the Supreme Court decided that questions of law/legal standards were subject to correctness review and questions of fact were subject to review for palpable and overriding error. But it asked itself how questions of mixed fact and law should be handled given that they sit on a spectrum ranging from very low legal content to very high legal content.
The Supreme Court’s solution was that where a question of law or an issue of legal principle is
“extricable” from the question of mixed fact and law, there is indeed a
“question of law”:
To summarize, a finding of negligence by a trial judge involves applying a legal standard to a set of facts, and thus is a question of mixed fact and law. Matters of mixed fact and law lie along a spectrum. Where, for instance, an error with respect to a finding of negligence can be attributed to the application of an incorrect standard, a failure to consider a required element of a legal test, or similar error in principle, such an error can be characterized as an error of law, subject to a standard of correctness. Appellate courts must be cautious, however, in finding that a trial judge erred in law in his or her determination of negligence, as it is often difficult to extricate the legal questions from the factual. It is for this reason that these matters are referred to as questions of “mixed law and fact”. Where the legal principle is not readily extricable, then the matter is one of “mixed law and fact” and is subject to a more stringent standard. The general rule, as stated in [Jaegli Enterprises Ltd. v. Taylor,  2 S.C.R. 2], is that, where the issue on appeal involves the trial judge’s interpretation of the evidence as a whole, it should not be overturned absent palpable and overriding error.
(Housen at para. 36.)
 This same approach should be adopted here. Extricable questions of law/legal standards are best regarded as questions of law of the sort intended by Parliament to be reviewed by this Court under subsection 41(1). In a number of cases, this Court determined appeals where extricable questions of law/legal standards (in addition to other legal and jurisdictional questions) were present:
Canadian National Railway Company v. Canadian Transportation Agency, 2010 FCA 65,  3 F.C.R. 264 (“CN 2010”) and Canadian National Railway Co. v. Canada (Canadian Transportation Agency), 2008 FCA 363, 383 N.R. 349 (“CN 2008”). What matters fall into certain defined terms in the Act, triggering the revenue cap in the Act? The extricable legal question was the definition of the defined terms in the Act.
Dreyfus, above at para. 18. Two issues were raised that involve extricable questions of law, namely statutory interpretation. Does the
“evaluation approach,”a methodology adopted by the Agency for deciding questions under sections 113-116, deviate from the proper interpretation of the sections? Did the Agency fail to consider matters that the statute requires it to consider?
Canadian National Railway Company v. Richardson International Limited, 2015 FCA 180, 476 N.R. 83. Do the facts of the case constitute a
“line of railway”and a
“connection”for the purposes of triggering the carrier’s interswitching obligations? The extricable question of law was the meaning of these terms.
Canadian National Railway Company v. Viterra Inc., 2017 FCA 6. On the facts, were the obligations of the carrier under section 113 triggered? Was the carrier’s rationing methodology a confidential contract under subsection 113(4) of the Act?
On occasion, this Court has defined the phrase
“questions of law” in subsection 41(1) as including questions of mixed fact and law as long as there is
“enough of a legal component” to the issue raised: Northwest Airlines Inc. v. Canadian Transportation Agency, 2004 FCA 238 at para. 28, 325 N.R. 147; Canadian National Railway Company v. Canada (Transportation Agency), 2016 FCA 266 at para. 22 (“CN 2016”). The phrase
“enough of a legal component” suffers from some ambiguity and lack of clarity: for example, how much is
“enough” and is the assessment of sufficiency a qualitative one, a quantitative one or both? The
“extricable questions of law or legal principle” standard is more concrete and clear, especially since appellate courts considering the appellate standard of review under Housen regularly have to grapple with the phrase and define it. In both Northwest Airlines and CN 2016 there were extricable questions of law or legal principle supporting the determination of the appeals under subsection 41(1) of the Act.
Therefore, in future, this Court should adopt the
“extricable questions of law or legal principle” standard for determining whether a question of mixed fact and law should be regarded as a
“question of law” under subsection 41(1) of the Act.
Turning to the facts of this case, does subsection 41(1) apply to restrict or eliminate CN’s appeal? To answer that, we first must identify the subject-matter of the appeal. We do this by construing the originating document, here the notice of appeal, to gain
“a realistic appreciation” of the appeal’s
“essential character.” The say-so of a party that a
“legal test” or
“the Act” is involved is not enough:
“skilful pleaders” who are
“armed with sophisticated wordsmithing tools and cunning minds” can express grounds in such a way as to make them sound like legal questions
“when they are nothing of the sort.” We must look at the substance of what is being raised, not the form. See generally Canada (National Revenue) v. JP Morgan Asset Management (Canada) Inc., 2013 FCA 250,  2 F.C.R. 557 at paras. 49-50.
 Sometimes an appellant’s memorandum of fact and law articulates the grounds set out in a notice of appeal in a different way. The memorandum can be useful in gaining a realistic appreciation of the appeal’s essential character, as presented in the notice of appeal. As we shall see, CN’s memorandum does assist us in this case.
 CN’s notice of appeal alleges that the Agency erred in two ways:
1. The Agency applied the wrong evidentiary threshold and onus such that it effectively and automatically conflated a car order request placed by a shipper with “traffic offered for carriage” under [subsection 113(1) of] the Act; and
2. The Agency applied the wrong legal test in determining whether CN had breached its level of service obligations by treating unfulfilled car order requests in a given week as constituting cumulative “traffic offered for carriage” under [subsection 113(1) of] the Act in subsequent weeks, months and years.
CN submits that these errors are questions of law relating to the proper interpretation of the phrase
“traffic offered for carriage” in subsection 113(1) of the Act.
 The first ground in the notice of appeal is phrased as a question of mixed fact and law—how the law should be applied to the facts of this case—not as a pure question of law. The second error alleges that the Agency applied a wrong legal test, but it could also be construed as an objection to how the Agency characterized the unfulfilled orders based on the facts of this case. In both cases, Emerson suggests that these are factually suffused points that are, so-to-speak, insufficiently legal and, thus, cannot be appealed under subsection 41(1) of the Act.
 I disagree. In my view, the essential character of this ground of appeal is that the Agency erred in law by taking a legally incorrect view of subsection 113(1).
For the purposes of this point, subsection 113(1) has two parts: an event that triggers the carrier’s legal obligation to transport, namely that the shipper has
“traffic offered for carriage,” and the shipper’s legal obligation
“without delay, and with due care and diligence, [to] receive, carry and deliver the traffic.” Both grounds concentrate on the triggering event: what is
“traffic offered for carriage”? What, as a matter of law, must a shipper like Emerson do in order to trigger the carrier’s legal obligation to receive, handle and transport? In other words, in law, what is the triggering event for the carrier’s legal obligation and in what circumstances does it happen?
The first ground in the notice of appeal suggests that the Agency
“car order request” by a shipper, Emerson, with
“traffic offered for carriage.” In other words, according to CN, the Agency erred in holding that as soon as Emerson says it has traffic for carriage, the triggering event has happened and CN’s onerous obligations under subsection 113(1) kick in. Put another way, under the Agency’s legally wrong view of the matter, CN’s legal obligation arises when Emerson says simply that it has a shipment that needs to be transported, without any other demonstration that there is indeed a shipment that needs to be transported. By misinterpreting subsection 113(1), the Agency allows the significant, sometimes onerous legal obligation on CN to receive, carry and deliver traffic to arise too easily, almost automatically. CN says that subsection 113(1), properly interpreted, requires much more from Emerson.
In short, the question raised by the first ground of appeal is a matter of statutory interpretation: what is the meaning of
“traffic offered for carriage,” or, put another way, what evidence must a shipper adduce to establish that there is
“traffic offered for carriage”? In the first ground of appeal, we have an issue of statutory interpretation: an issue of law. The resolution of the first ground of appeal is driven by the law/legal standards; this is not an issue where the parties agree upon the law/legal standards and the outcome is driven by the facts.
 Several statements from CN’s memorandum confirm that the first ground of appeal raises a question of law:
“The [Act] requires actual goods subject to a genuine request for transportation by rail”and
“a railway has no obligation [under the Act] to carry putative, hypothetical or speculative traffic.”To hold otherwise
“is an error of law”: CN’s memorandum of fact and law at para. 42;
“[Under subsection 113(1)] an applicant must prove that it has ‘traffic’ for carriage…and a [mere] car order cannot be presumed to be ‘traffic’”: CN’s memorandum of fact and law at para. 44;
Under the Act,
“[a] railway is not presumed to be in breach of its level of service obligations.”Instead, an applicant must first
“prove that it offered traffic for carriage—a railway’s service level obligations [under subsection 113(1)] are not triggered otherwise”: CN’s memorandum of fact and law at paras. 49 and 52. The
“traffic offered for carriage”under subsection 113(1)
“must be actual goods offered for transport”and not just
“putative or potential traffic”: CN’s memorandum of fact and law at para. 52.
Any other conclusion
“fails to accord with the words, context, object and intent of the [Canada Transportation Act]”: CN’s memorandum of fact and law at para. 53.
Interpreting subsection 113(1) in this way violates the Supreme Court’s holding in A.L. Patchett & Sons Ltd. v. Pacific Great Eastern Railway Co.,  S.C.R. 271, 17 D.L.R. (2d) 449
“that a railway’s common carrier obligations are ‘permeated with reasonableness”: CN’s memorandum of fact and law at para. 72.
A sub-point raised by CN and really the flipside of the foregoing is that the Agency must have improperly taken
“judicial notice of the fact that car orders invariably reflect concomitant demand for the carriage of actual traffic”: CN’s memorandum of fact and law at para. 65. Whether the Agency can do this is a question of law.
CN’s memorandum of fact and law also runs this statutory interpretation point in a different way. CN takes particular issue with a methodology or test established by the Agency and followed by in this case. It is known as the
 The first step of the Agency’s evaluation approach is to assess whether the request for service is reasonable. This entails looking at factors such as whether the request for cars was properly communicated, whether the car request provided adequate notice and whether the shipper has the capacity to receive, load and release the cars requested: CN’s memorandum of fact and law at para. 54.
CN complains that
“none of these criteria relate in any way to the issue of whether the shipper has actual traffic for carriage,” which it says subsection 113(1) requires; instead they
“relate solely to matters of the form and timing of communication, and capacity to receive and release a car spot”: CN’s memorandum of fact and law at para. 56. It also complains that the Agency found that the
“sole requirement for a shipper to meet the first step of the evaluation approach is to place an order,” again contrary to subsection 113(1): CN’s memorandum at para. 64. Lest there be any doubt that the argument CN makes is based on the proper interpretation of subsection 113(1), CN adds that the Agency’s evaluation approach
“obviates [the] statutory requirement [of showing that there is “traffic offered for carriage”]—a shipper need only make a car request in order to be conclusively deemed to have ‘traffic’”: CN’s memorandum of fact and law at para. 44.
 All of these submissions in CN’s memorandum relate to the gist of the first ground in the notice of appeal. They confirm that we are dealing with an issue of statutory interpretation, which is a question of law that this Court can entertain under subsection 41(1) of the Act.
 The issue of statutory interpretation raised by the first ground of appeal is very much like the question at issue before this Court in CN 2010, above and CN 2008, above, namely what matters fall into certain defined terms in the Act, triggering the revenue cap in the Act. In considering the evaluation approach and in raising the question whether the approach is consistent with the Act, this case is also very much like Dreyfus. All of these cases passed muster under subsection 41(1) of the Act.
 Overall, I find that the first ground of appeal raises a question of law that can be appealed to this Court under subsection 41(1) of the Act.
The second ground of appeal complains that the Agency
“applied the wrong legal test” in determining whether CN had breached its level of service obligations by
“treating unfulfilled car order requests in a given week as constituting cumulative ‘traffic offered for carriage’ under [subsection 113(1) of] the Act in subsequent weeks, months and years.” Although this ground refers to
“legal test” and
“the Act,” it is still incumbent on us to construe the originating document, here the notice of appeal, to gain
“a realistic appreciation” of the appeal’s
“essential character”: see paragraph 29 above.
 This ground of appeal is best understood by viewing it in light of what CN argues in its memorandum of fact and law and in light of what the Agency decided. When this is done, we see that the second ground of appeal raises an issue of statutory interpretation. CN and the Agency have contrasting views about how to go about analyzing cases under subsection 113(1), views based on a different interpretation of subsection 113(1):
CN’s view. Subsection 113(1) requires the Agency to assess on a week-by-week basis whether CN is justified for that week in failing to provide enough cars. If non-delivery during a particular week is justified by the circumstances, then any non-delivery of cars in that week is excused and cannot be the basis for a later finding that CN breached its obligations as a carrier under subsection 113(1) of the Act. In the words of the second ground in the notice of appeal,
“treating unfulfilled car order requests in a given week”that are justified cannot be regarded
“as constituting cumulative ‘traffic offered for carriage’ under [subsection 113(1) of] the Act in subsequent weeks, months and years.”See generally paras. 77-95 of CN’s memorandum of fact and law.
Agency’s view. Subsection 113(1) allows the Agency to look at the matter more globally, as it did here, and assess from the available data whether over the entire complaint period CN met its subsection 113(1) obligations. It need not conduct a week-by-week analysis. See generally paras. 65-68 of the Agency’s decision.
Which view of subsection 113(1) should prevail: the week-to-week approach or the global approach? Fundamentally, this is a question of statutory interpretation, a question of law.
 We can glean CN’s view of subsection 113(1)—which varies from the Agency’s view—from its memorandum. CN begins by arguing that the Agency recognized that in certain weeks CN was justified in not delivering all of the cars that Emerson had ordered: CN’s memorandum of fact and law at para. 77. According to CN, if it was justified in failing to deliver a certain portion of car requests at the time the requests were made, by definition it complied with its obligations under the Act: CN’s memorandum of fact and law at para. 77.
 CN puts this same point a different way. It says that under the Act a railway’s obligation to move traffic is triggered when presented with traffic for carriage: CN’s memorandum of fact and law at para. 80. Then, once traffic for carriage is presented, the railway either moves the traffic or does not. If the railway fails to deliver the traffic, the Agency’s task is to determine whether the railway’s failure was justified as of the time of the request: CN’s memorandum of fact and law at para. 82. If it was justified, then under subsection 113(1) the railway has offered the requisite level of service: CN’s memorandum of fact and law at para. 86.
Given that there were numerous crop weeks where CN was absolved of its obligation to deliver cars,
“it was not open for the Agency to perfunctorily declare that CN was in breach of its obligation to deliver cars”: CN’s memorandum of fact and law at para. 94. The Agency reached this conclusion
“improperly” and committed an
“error of law” by
“treating unfulfilled orders as though they continued to represent traffic offered for carriage within the meaning of ss. 113(1)(a)”: CN’s memorandum of fact and law at para. 95.
As mentioned above, our task is to gain
“a realistic appreciation” of the appeal’s
.” When the second ground in the notice of appeal is read together with CN’s memorandum of fact and law, it becomes evident that CN is taking issue with how the Agency read and applied the statute.
 This is not a case where the Agency and CN have a common view of how the statutory provision, subsection 113(1), should be read and CN merely takes issue with the way the Agency has applied it to the facts of the case. That would be a question of mixed fact and law where the facts drive the answer.
 Rather, this is a case where the Agency and CN have a different view on how subsection 113(1) is to be read. In my view, the second ground set out in the notice of appeal raises an extricable question of law sufficient for an appeal under subsection 41(1) of the Act.
 Before leaving this issue, I wish to offer some further guidance for future cases concerning subsection 41(1) of the Act, guidance that may be useful for similarly worded sections.
 This is a relatively close case under subsection 41(1). As mentioned, in determining whether we have jurisdiction in the face of a jurisdiction-limiting provision like subsection 41(1) we must examine the essential character of a notice of appeal with the assistance of the appellant’s memorandum and, like all pleadings, construe it generously with due allowance for infelicities in wording. But there are limits to the Court’s examination and its generosity. In this case, those limits were almost reached.
 Drafters of notices of appeal are now on notice. When the grounds of appeal are drafted in the form of questions of mixed fact and law, the Court may well conclude that there is no question of law or jurisdiction in the appeal and dismiss it. This is especially so where the questions of mixed fact and law are presented in a heavily fact-laden way. Instead, drafters should identify with clarity and precision the questions of law or of jurisdiction, including any extricable questions of law or legal principle, and explain how these emerge from the decision below. While the Court may look to the appellant’s memorandum of fact and law to construe the notice of appeal, the notice of appeal, as the originating document, remains the primary focus of the Court and must be carefully drawn.
 These observations have ramifications for motions for leave to appeal under provisions like subsection 41(1). When the Court grants leave to appeal, it has not decided the subsection 41(1) issue; for one thing, in granting leave the Court may have considered the issue of jurisdiction to be uncertain but fairly arguable: Canadian Pacific Railway Co. v. Canada (Transportation Agency), 2003 FCA 271,  4 F.C.R. 558 at para. 17. Despite the granting of leave, the subsection 41(1) issue remains live during the appeal and the Court must dismiss any appeal over which it does not have jurisdiction.
 This being said, those of the view that this Court has no jurisdiction under subsection 41(1) should forcefully argue the point at the leave stage and, where possible, this Court should determine it. Increasingly, courts must conserve scarce judicial resources and adopt a new, more efficient litigation culture: Hryniak v. Mauldin, 2014 SCC 7,  1 S.C.R. 87. Appeals or issues in appeals that cannot be entertained by this Court should not be allowed to meander through to a merits hearing. Instead, at the earliest opportunity, they should be stopped in their tracks. The guidance given by these reasons to drafters of notices of appeal applies equally to drafters of notices of motion for leave to appeal.
B. Analysis of the merits of the appeal
As mentioned above, subsection 113(1) of the Canada Transportation Act imposes certain obligations upon a railway company once there is
“traffic offered [to the railway company] for carriage” within the meaning of the subsection. Subsection 113(1) provides as follows:
 The parties agree that this Court should review the Agency’s interpretation of subsection 113(1) of the Act on the basis of the standard of reasonableness. The parties’ agreement does not bind us: Monsanto Canada Inc. v. Ontario (Superintendent of Financial Services), 2004 SCC 54,  3 S.C.R. 152. But on the current state of the authorities I agree that the standard of review is reasonableness.
Reasonableness is presumed to be the standard of review of administrative decision-makers’ interpretations of provisions in
“[their] own [legislation] or [legislation] closely connected to [their] function, with which [they have] particular familiarity,” i.e., their home statute: Alberta (Information and Privacy Commissioner) v. Alberta Teachers’ Association, 2011 SCC 61,  3 S.C.R. 654 at para. 34; Dunsmuir, above at para. 54. This presumption applies even where Parliament has enacted full, unrestricted rights of appeal: Edmonton (City) v. Edmonton East (Capilano) Shopping Centres Ltd., 2016 SCC 47,  2 S.C.R. 293.
 Subsection 113(1), interpreted and applied by the Agency in this case, is in the Agency’s home statute and so the presumption applies. It stands unrebutted. Therefore, the Agency’s interpretation and application of subsection 113(1) will be reviewed on the basis of reasonableness.
 By way of confirmation, I note that this Court has adopted the reasonableness standard in a number of similar cases involving similar issues before the Agency: see the cases mentioned in paragraph 26 above.
Now to reasonableness review of the Agency’s decision. Some go about this by forming a view as to what the administrative decision-maker should have decided on the merits, pasting paragraph 47 of Dunsmuir into their reasons—whether the outcome reached by the Agency falls within the range of
“defensibility” on the facts and the law and whether there is
“intelligibility”—and then tossing these labels around to support their conclusion. Some call this disguised correctness.
 Others avoid the sin of disguised correctness but still fall short. They understand that disguised correctness is not genuine reasonableness review. But too often, even in complicated cases that demand a more fulsome treatment, they do not go much beyond asserting conclusions, rather than demonstrating in a substantive way how the administrative decision does or does not accord with the concept of reasonableness.
If reasonableness review is to be legitimate and if it is to appear to be legitimate, it must be conducted in a neutral, substantively rigorous, intellectually honest way, drawing upon the doctrine and sensitive to
“the qualities that make [an administrative decision] reasonable”: Dunsmuir, above at para. 47. To try to do just that, our Court has developed and followed some approaches to reasonableness: see, e.g., Delios v. Canada (Attorney General), 2015 FCA 117, 472 N.R. 171; Canada (Attorney General) v. Boogaard, 2015 FCA 150, 474 N.R. 121; and see Professor Paul Daly,
“Struggling Towards Coherence in Canadian Administrative Law? Recent Cases on Standard of Review and Reasonableness” (online: https://ssrn.com/abstract=2821099) (forthcoming, McGill L.J.). And in doing this—far from freestyling on the matter—this Court has followed the Supreme Court’s pronouncements, attentive to the signals it gives.
What are the pronouncements and signals? In some cases, the Supreme Court tells us that an administrative decision-maker’s ambit for decision-making on a particular question is not
“one specific, particular result” but rather is a
“range of reasonable outcomes” or a
“margin of appreciation”, a range or margin that can be quite broad or narrow depending on the circumstances: Dunsmuir, above at para. 47; McLean v. British Columbia (Securities Commission), 2013 SCC 67,  3 S.C.R. 895 at para. 38. In other cases, the Supreme Court tells us that reasonableness
“takes its colour from the context” and must be
“assessed in the context of the particular type of decision making involved and all relevant factors”: Catalyst Paper Corp. v. North Cowichan (District), 2012 SCC 2,  1 S.C.R. 5 at para. 18; Canada (Citizenship and Immigration) v. Khosa, 2009 SCC 12,  1 S.C.R. 339 at para. 59; Wilson v. Atomic Energy of Canada Ltd., 2016 SCC 29,  1 S.C.R. 770 at para. 22; and many, many others. In other words, certain circumstances, considerations and factors in particular cases influence how we go about assessing the acceptability and defensibility of administrative decisions: Catalyst at para. 18; Doré v. Barreau du Québec, 2012 SCC 12,  1 S.C.R. 395 at para. 54; Halifax, above at para. 44; see also Canada (Minister of Transport, Infrastructure and Communities) v. Farwaha, 2014 FCA 56,  2 F.C.R. 1006 at paras. 88-99.
Looking at this from the perspective of reviewing courts, if the circumstances, considerations and factors differ from case to case, how reviewing courts go about measuring acceptability and defensibility will differ from case to case; in other words, reasonableness will
“take its colour from the context” of the case. Looking at this from the perspective of administrative decision-makers, as a practical matter some in some contexts seem to be given more leeway or a broader
“margin of appreciation” than others in other contexts.
 In some of its cases, this Court has tried to identify the circumstances, considerations and factors that can affect the outcome of reasonableness review. Sometimes other appellate courts have joined this effort: see, e.g., Mills v. Ontario (Workplace Safety and Insurance Appeals Tribunal), 2008 ONCA 436, 237 O.A.C. 71 at para. 22.
 In this case, what is the context from which reasonableness takes its colour? What circumstances, considerations or factors affect reasonableness review?
 One important factor is the existence of a binding judicial pronouncement concerning subsection 113(1) of the Act. Unless the administrative decision-maker, here the Agency, can distinguish the precedent in some reasonable way, it constrains the interpretive options available to the Agency and affects our evaluation of reasonableness: Canada (Attorney General) v. Canadian Human Rights Commission, 2013 FCA 75; 444 N.R. 120 at paras. 13-14; Canada (Attorney General) v. Abraham, 2012 FCA 266, 440 N.R. 201 at paras. 37-50; Farwaha, above at para. 95.
 In this case, the binding judicial pronouncement is a decades-old decision of the Supreme Court. It stands for the proposition that when interpreting and applying subsection 113(1), the Agency must assess the reasonableness of the parties’ conduct in light of the facts disclosed by the evidentiary record:
Apart from statute, undertaking a public carrier service as an economic enterprise by a private agency is done on the assumption that, with no fault on the agency's part, normal means will be available to the performance of its duty. That duty is permeated with reasonableness in all aspects of what is undertaken…and it is that duty which furnishes the background for the general language of the statute. The qualification of reasonableness is exhibited in one aspect of the matter of the present complaint, the furnishing of facilities: a railway, for example, is not bound to furnish cars at all times sufficient to meet all demands; [all they must do is provide] a reasonable service. Saving any express or special statutory obligation, that characteristic extends to the carrier's entire activity. Under that scope of duty a carrier subject to the Act is placed.
…The duty being one of reasonableness how each situation is to be met depends upon its total circumstances. The carrier must, in all respects, take reasonable steps to maintain its public function; [its obligation] must be determined by what the railway, in the light of its knowledge of the facts, as, in other words, they reasonably appear to it, has effectively done or can effectively do to meet and resolve the situation.
(Patchett, above at pp. 274-275.) In developing acceptable and defensible jurisprudence concerning subsection 113(1) of the Act, the Agency must work within the standards set by Patchett.
 Another context colouring reasonableness review in this case is the nature of the Agency’s decision and the nature of the Act. The Agency’s decision lies at the very bullseye of its regulatory know-how and mandate, the very reason why Parliament has vested the Agency with jurisdiction over the merits of cases like this and has left us with just a reviewing role.
 When the Agency interprets subsection 113(1), it legitimately draws upon its regulatory experience, its knowledge of the industry and its expertise in the transportation sector, guided by the standards set by Patchett, above. Provided the Agency adopts a defensible interpretation of subsection 113(1) and a defensible methodology or test for determining reasonable conduct, and provided it applies these things in a manner that is alert and responsive to the evidence before it, this Court must refrain from second-guessing. The reasonableness of the parties’ conduct based on the particular facts disclosed by the evidentiary record—the factually suffused merits of the case—is a matter very much within the ken of the Agency, not us. See, by way of illustration and analogy, National Corn Growers Assn. v. Canada (Import Tribunal),  2 S.C.R. 1324 at pp. 1347-48, 74 D.L.R. (4th) 449 and Council of Canadians with Disabilities v. VIA Rail Canada Inc., 2007 SCC 15,  1 S.C.R. 650 at para. 104.
What did the Agency do in this case? First, it applied a test—which it calls an
“evaluation approach” (at para. 10).
 Under this evaluation approach, the Agency asked itself three questions (at para. 10):
1. Is the shipper’s request for service reasonable?
2. Did the railway company fulfil this request?
3. If not, are there reasons which could justify the service failure?
(a) If there is a reasonable justification, then the Agency will find that the railway company has met its service obligations;
(b) If there is no reasonable justification, then the Agency will find that there has been a breach of the railway company’s service obligations and will look to the question of remedy.
CN submits that the Agency’s
“evaluation approach” is unreasonable: see CN’s memorandum of fact and law at paras. 22 and 25. I disagree.
The Agency’s evaluation approach is reasonable. It is a practical, useable test that captures both the essence of and much of the detail in subsection 113(1) of the Act. Far from imposing impossible burdens upon carriers like CN, as CN suggests, it suitably reflects the Supreme Court’s holding in Patchett that the carrier’s duty
“is permeated with reasonableness in all aspects of what is undertaken.” Take, for example, the centrality of reasonableness in the first question—whether the shipper’s request for service is reasonable. In my view, CN’s real concern is not so much with the evaluation approach, but rather with the manner in which the Agency has applied it to the evidence in this case.
 In following the evaluation approach, the Agency must be careful. The evaluation approach is just a practical and general test or methodology for analyzing and applying the standards set out in subsection 113(1) of the Act. It is not the wording of subsection 113(1) itself. The meaning of the wording always governs, not the tests or methodologies the Agency has fashioned in its jurisprudence. Therefore, if a party wishes to submit before the Agency in a particular case that subsection 113(1) requires that the evaluation approach needs to be tweaked, modified, followed or applied differently, the Agency must consider the submission in an open-minded way.
The first question under the evaluation approach—whether the shipper’s request for service is reasonable—reflects the Agency’s view that only bona fide, reasonable requests by shippers for traffic on the railway can meet the requirement in subsection 113(1) of the Act that there be
“traffic offered for carriage on the railway.” For example, the railway company’s obligations under subsection 113(1) are not triggered by groundless, outlandish requests.
In this case, the focus of CN’s attack on the Agency’s decision is on its handling of this first question. As is shown by its notice of appeal, it says that the Agency was too trusting of Emerson’s say-so that it needed railcars. Put another way, the Agency assumed that a request by Emerson satisfied the statutory requirement under subsection 113(1) that there be
“traffic offered for carriage.”
There are statements in the Agency’s reasons that, if plucked out of context and read in isolation, could support CN’s attack. For example, at one point in its reasons, the Agency states, without elaboration, that
“by placing orders according to CN’s policy and ordering system, [Emerson] properly triggered CN’s level of service obligations” (at para. 27). This bald statement was said in the context of the Agency’s finding that Emerson did not have to anticipate and notify CN about the increase in its demand for transportation services that would result from the bumper crop (at para. 26).
The real question behind CN’s submission is what amount and sort of evidence a shipper like Emerson must bring forward in order to trigger the carrier’s obligation to
“receive, carry and deliver the traffic.” In other words, using the words of the evaluation approach and the Patchett standard, what amount and sort of evidence must a shipper bring forward to demonstrate that its request for service was reasonable? This called for an interpretation of subsection 113(1) in light of its text, its context within the legislation, and the purpose of the legislation—the methodology set out in cases such as Re Rizzo & Rizzo Shoes, above and Bell ExpressVu, above.
 The Agency did not explicitly follow the text-context-purpose approach. From the standpoint of clarity, it might have been better had it done so. But the Agency’s observations concerning subsection 113(1) and its analysis of how it applies to this case reflect these very matters. It did demonstrate an appreciation of the text, context and purpose of subsection 113(1) and it viewed these things—as it must do—through its particular regulatory lens. In examining whether Emerson had demonstrated that its request for service was reasonable, it brought to bear its regulatory experience, its knowledge of the industry, its understanding of how transactions between shippers and their customers come about and are documented, and its overall expertise in the transportation sector. These things are largely beyond the ken of the Court and, thus, are matters on which the Agency is given a wide margin of appreciation.
A specific instance of the Agency drawing upon these things is seen in its ruling concerning the significance that can be drawn merely from the fact that a shipper has requested railcars. It observed that in the context of this industry and its normal practice, a request for railcars by itself is some evidence of the need for carriage or, in the words of the Agency, a request will
“generally” (i.e., not always) show that cars are needed for carriage, subject to the consideration of contrary evidence such as
“bills for demurrage charges” (at para. 28).
 The Agency’s reasons on this point grappled with the evidence before it, though not necessarily in a way CN likes, finding that Emerson had adduced enough evidence to prove that its request for service was reasonable.
Thus, I disagree with CN’s submission that the Agency in effect simply accepted Emerson’s say-so about its need for railcars. Nor does the bald statement in paragraph 27 of its reasons—
“by placing orders according to CN’s policy and ordering system, [Emerson] properly triggered CN’s level of service obligations”—stand alone.
 This is all seen in paragraphs 28-36 of the Agency’s reasons:
 With respect to CN’s allegation that [Emerson] should have to provide evidence of actual delivery commitments and arrangements, the Agency considers that demonstrating a reasonable request for service does not require a shipper to strictly document each and every transaction it makes in respect of the acquisition/production and the subsequent sale/use of the goods that shipper intends to ship. In the context of the transportation of grain, demand for rail cars is inextricably linked to demand for grain. Grain shippers order cars because they have grain to move and customers to purchase it. In the absence of any evidence showing that a grain shipper has ordered cars that it was not in a position to load and release to the railway company for carriage, for instance bills for demurrage charges, the Agency will generally conclude that if a commercial grain company orders cars, it is to move grain to market.
 As noted above, transportation is a derived demand and the purpose of section 113 of the [Act] needs to be understood in its broadest context….
 In the context of the grain industry, considering the manner in which that commodity is traded, it would be unreasonable for the Agency to require a shipper to produce, for each tonne of grain that the shipper intended to ship, the contractual arrangements showing that the grain was purchased and subsequently re-sold. This would render the availability of remedies for a level of service breach contingent on the shipper being in a contractual breach with its business partners.
 Evidence of contracts between the grain shipper and its grain suppliers and customers may constitute convincing evidence that the shipper had grain to move in the cars it ordered from the railway company. However, the fact that contracts have not been produced in respect of each car ordered from the railway company will not necessarily mean that the shipper failed to prove a reasonable request. Other elements of evidence may demonstrate to the satisfaction of the Agency that the shipper had grain to move in the cars it ordered from the railway company.
 In this case, [Emerson] filed letters from two of its customers complaining about delayed deliveries. [Emerson] also provided copies of its forward sales contracts with producers.
 Further evidence on the file indicates that at the system level there was a bumper crop in 2013 and that more grain had to be moved than cars supplied. This is consistent with CN’s need to ration cars in the first place; the demand for cars exceeded the supply of cars.
 The agency is of the opinion that this shows that the supply of grain existed and that there was a demand for [Emerson’s] grain to be delivered. Therefore, [Emerson] had a legitimate demand for grain transportation services.
 CN did not provide any evidence that rebutted [Emerson’s] evidence. While CN alleges that [Emerson] engaged in “tactical ordering,” CN did not provide any evidence to demonstrate that [Emerson] ordered more cars than it actually needed to move the oats available to it for sale to its customers.
 The Agency notes that in more than one week during the complaint period, [Emerson] ordered more cars than the capacity of its siding. CN suggests that this indicates that [Emerson] ordered more cars than it required. The Agency is of the opinion that the fact that [Emerson] ordered more cars in a week than can be delivered in a single spot only proves that [Emerson] placed orders that would require CN to serve [Emerson] more than once in the same crop week; it does not prove that [Emerson] ordered more cars than it required given the supply and the demand conditions in the market.
Based on the record before it, including the bumper crop of 2013-2014 and the availability of grain, the Agency concluded (at paras. 38-39) that
“on the balance of probabilities” Emerson
“had a legitimate demand for service” or, in other words, it
“would have had grain to ship had it received the cars it ordered.” Under subsection 113(1) this triggered CN’s obligations. Again, based on the record before it, the Agency concluded that CN did not fulfil its obligations (at paras. 40-46). Following the reasoning in its October 3, 2014 Dreyfus decision—a decision mindful that, as the Supreme Court said in Patchett, CN’s duty under subsection 113(1) of the Act is
“permeated with reasonableness in all aspects of what is undertaken” and a decision that this Court upheld as reasonable in Dreyfus, above—the Agency found that CN breached its level of service obligations to Emerson during Emerson’s complaint period (at paras. 47-75).
I turn now to CN’s submission that the Agency reached an unreasonable result by, as it put it in the second ground of its notice of appeal,
“treating unfulfilled car order requests in a given week as constituting cumulative ‘traffic offered for carriage’ under [subsection 113(1) of] the Act in subsequent weeks, months and years.”
At paragraphs 65-68 of its reasons, the Agency found that CN was justified in some delay in delivering cars to Emerson during the complaint period but that CN was not completely relieved of its service obligations concerning the traffic Emerson had offered for carriage. After all, subsection 113(1) of the Act requires a carrier to receive, carry and deliver
“all traffic” offered for carriage
 The Agency studied CN’s service over the entire complaint period and found that the proportion of cars waybilled to cars ordered decreased over time. Although CN could justify some delay in delivering cars, it could not justify the indefinite delay in service for what turned out to be 40 percent of Emerson’s traffic. In support of this finding, the Agency adopted a particular view of CN’s obligations under the Act and then applied that understanding of the Act to the facts before it (at paras. 65-66 and 68):
 While the [the fact that a] service request of a given shipper is unexpected or differs from historical patterns may justify some delay in delivering cars ordered, they must nonetheless be delivered and moved by the railway company. The term “without delay” in paragraph 113(1)(c) of the [Canada Transportation Act] needs to be interpreted in the context of the specific circumstances of each case. When faced with an unexpected demand for service, especially if the railway company did not have sufficient lead time to react, paragraph 113(1)(c) of the [Canada Transportation Act] will be interpreted as providing a railway company a reasonable amount of time to fulfill the service request in question.
 However, this does not mean that a railway company can invoke a lack of lead time indefinitely and in all cases. Railway companies must, under the [Canada Transportation Act], ensure that they allocate sufficient resources on an ongoing basis, to furnish adequate and suitable accommodation for the carriage, unloading and delivering of the traffic offered for carriage on the railway. The lack of lead time to plan does not nullify a railway company’s obligations to receive, carry and deliver, without delay, the traffic ordered.
 The Agency finds that the factors led to some delay in delivering cars to [Emerson] during the complaint period; however, CN did not establish a justification for the overall poor level of service it provided to [Emerson] over the complaint period. Specifically…[Emerson] experienced a pattern of poor service such that CN’s service to EMI deteriorated from the beginning of the complaint period to the point that for four weeks, beginning in Week 36, [Emerson] received only 49 percent of all the cars it had requested up to that point in time. By the end of the complaint period, CN had not delivered 40 percent of the total number of cars requested by [Emerson] during the complaint period.
From this understanding of the Act and these facts, the Agency concluded that
“CN breached its level of service obligations to [Emerson] over the complaint period” (at para. 70).
The Agency’s approach to subsection 113(1) was to look at the situation globally, alive to all the circumstances of the case, and to assess overall whether CN had fulfilled its obligations under the Act over a period of time, bearing in mind that, in the words of Patchett, those obligations are
“permeated with reasonableness.” This approach is supportable on the wording of subsection 113(1) as understood under the Patchett standard. The wording of subsection 113(1) does not require the sort of week-by-week examination that CN has urged upon us. A global examination of whether the carrier has fulfilled its obligations may have much to commend it where, as here, cars are being ordered and delivered every week and shortfalls are periodically occurring.
 Overall, in this case, the Agency reached factually suffused conclusions founded upon the evidentiary record and readings of subsection 113(1) of the Act consistent with an acceptable interpretation of the provision and the Supreme Court’s decision in Patchett. CN has not demonstrated that the Agency’s decision suffers from the sort of indefensible flaw or blatant mischaracterization that struck at the foundation of its decision in Viterra, above. For the foregoing reasons, the Agency’s decision in this case is acceptable and defensible on the facts and the law and, thus, reasonable.
C. Proposed disposition
 I would dismiss the appeal with costs.
FEDERAL COURT OF APPEAL
NAMES OF COUNSEL AND SOLICITORS OF RECORD
APPEAL FROM THE CONFIDENTIAL DECISION OF THE CANADIAN TRANSPORTATION AGENCY DATED JULY 10, 2015, NO. 14-06408
STYLE OF CAUSE:
CANADIAN NATIONAL RAILWAY COMPANY v. EMERSON MILLING INC. AND CANADIAN TRANSPORATION AGENCY
PLACE OF HEARING:
Vancouver, British Columbia
DATE OF HEARING:
September 19, 2016
REASONS FOR JUDGMENT BY:
CONCURRED IN BY:
APRIL 18, 2017
Douglas C. Hodson, Q.C.
For The Appellant
For The Respondent,
EMERSON MILLING INC.
FOR THE RESPONDENT, CANADIAN TRANSPORTATION AGENCY
SOLICITORS OF RECORD:
MacPherson Leslie & Tyerman LLP
For The Appellant
DLA Piper (Canada) LLP
Vancouver, British Columbia
For The Respondent,
EMERSON MILLING INC.
Canadian Transportation Agency
FOR THE RESPONDENT, CANADIAN TRANSPORTATION AGENCY