The Scope of Economic Regulatory Powers: ATCO Electric Ltd v Alberta Utilities Commission, 2023 ABCA 129

In the original ATCO case in 2006, the Supreme Court of Canada addressed the issue of the scope of a regulator’s statutory authority, making a distinction between express and implied powers. As I have written, this distinction makes little sense and it would be better simply to focus on legislative intent, discerned by consideration of statutory text, purpose and context. Earlier this year, the Alberta Court of Appeal considered the impact of the original ATCO case and shed light on the proper approach to determining the scope of regulatory authority: ATCO Electric Ltd v Alberta Utilities Commission, 2023 ABCA 129.

The issue here was the lawfulness of a decision of the Commission denying ATCO the ability to recover costs incurred as a result of a wildfire: according to the Commission, it did not have the authority to permit ATCO to include these costs in the rates charged to consumers of electricity; these were “extraordinary” retirements of assets for which consumers ought not to pick up the tab. The Court of Appeal summarized the Commission’s decision as follows:

To summarize, the effect of the decision under appeal was to disrupt the normal cycle of depreciation, because the loss of assets resulting from the Fort McMurray wildfire was said to be “extraordinary”. The full cost of those destroyed assets in the pool would previously have been recovered over their expected average life, notwithstanding how long any particular asset actually survived, even if its life had been cut short by forces of nature. Because the Fort McMurray wildfire losses were more significant than those predicted in the survival curves and so were said to be extraordinary, the decision under appeal changed the approach by now looking at individual asset survival history. Further, any destruction by fire of an individual asset in a mass depreciation pool meant that the remaining undepreciated capital cost of that asset had to be removed from the pool. The appellant would no longer be able to recover those costs (at para. 10).

The Commission arrived at this view because it took the original ATCO decision and subsequent jurisprudence to limit its flexibility in dealing with destroyed assets (at paras. 13, 19). Subsequent to ATCO, the Commission extended the logic of that decision to all extraordinary retirements of assets (at para. 30). The Court of Appeal considered that this decision, which was a general, policy decision that did not consider any specific disposition of assets or request for a rate to be made, was reasonable: FortisAlberta Inc v Alberta (Utilities Commission), 2015 ABCA 295.

But as the Court of Appeal pointed out, the original ATCO decision dealt with the sale of an asset and the power of a regulator to attach conditions to the sale (at para. 44). This is a much more narrowly drawn power than the rate-setting authority at issue here. Indeed, the rate-setting authority refers to concepts such as a “depreciation” and “prudence” without defining them, leaving them to the regulator to work out on a case-by-case basis (at para. 43). In this regard, the Commission had “broad discretion” (at para. 44) and nothing in the original ATCO decision dictated the outcome of a rate-setting proceeding:

The outcome here was not inevitable or driven by any binding precedent. It is conceded that the destroyed assets were prudently purchased. They were actively being used to provide electrical service when they were destroyed. There is nothing inherently unjust about allowing the utility to recover these costs. It can easily be argued that the risk of fire loss from natural causes should be spread among the customers who benefit from the service being provided, not placed on the shareholders (at para. 46).

Indeed, the original ATCO decision was simply not analogous at all to the issue before the Commission, as it addressed the distribution of profits (as between shareholders and customers) from the disposal of a property, whereas here, the property had been destroyed and there was no profit to distribute (at paras. 52-53). Accordingly, the matter was sent back to the Commission for redetermination (at paras. 61-62).

This is a neat example of the importance of considering statutory text, purpose and context in the determination of the scope of regulatory authority. The Commission had drawn an analogy with a past case rather than looking to its statutory authority. This was an error as it caused the Commission to overlook the breadth of the power it had. Indeed, when the Supreme Court of Canada has considered the scope of the authority to set just and reasonable rates, it has taken a much more expansive view of regulatory authority than it did in the original ATCO decision (see especially Bell Canada v. Bell Aliant Regional Communications, 2009 SCC 40). The point, then, is that broad statutory language provides broad regulatory authority, with narrower statutory language providing relatively narrower authority. In all events, regulators should look first and foremost to statutes, with past jurisprudence relevant only to the extent it deals with the same statute that provides the basis for a proposed decision.

This might therefore seem a strange case for the Court of Appeal to also address issues relating to past jurisprudence but it nonetheless dealt with an important point about the continuing force of its previous decisions. In Vavilov, the Supreme Court made some radical changes to Canada’s administrative law framework. For instance, on a statutory appeal the standard of correctness henceforth applies on questions of law. Previously, the reasonableness standard would often apply, even on a statutory appeal on a question of law. Plenty of appellate jurisprudence (especially in the field of economic regulation) was based on consideration of the reasonableness of administrative decisions. Given, however, that reasonableness review has become synonymous with the concept of multiple possible interpretations, what is the precedential force today of a previous Court of Appeal decision holding a regulator’s interpretation of law to be reasonable given that the standard of review now applicable is correctness?

The Court of Appeal suggested that such decisions should be considered to be presumptively binding:

Vavilov should not be read as automatically displacing all the case law decided under the pre-Vavilov standard of review regime. For one thing, even if a decision was reviewed for “reasonableness” and found to be reasonable, that does not mean that the reviewing court did not also agree that the decision was “correct”: see for example FortisAlberta at paras. 130, 134, 137-38, 148. Further, decisions like FortisAlberta were themselves based on binding decisions like Stores Block. If the tribunal decision in FortisAlberta had failed to follow binding precedent it would not have been reasonable. The importance of stability in the law means that after Vavilov binding precedents of this Court should presumptively be regarded as continuing to be binding, notwithstanding the change in the standard of review analysis: see Vavilov at paras. 18, 143-44.

As Nigel Bankes has observed, Vavilov does not actually speak clearly to this issue. The cited passages in Vavilov deal with stare decisis generally and the relevance of the Supreme Court’s past precedents for selecting the standard of review and applying the reasonableness standard.

Is the Court of Appeal right? It seems to me that there are two distinct concerns at play here that have to be carefully untangled.

First, in some instances, a court applying the reasonableness standard will have indicated that a decision was not only reasonable but correct, or alternatively that a decision was unreasonable because there was only one acceptable, possible outcome. The Court of Appeal is right, in my view, to say that these decisions should be treated as binding even though they were decided on a reasonableness standard.

Second, however, on other occasions a court will have upheld a decision as reasonable without indicating what the correct decision would have been. The court’s conclusion might have been based, in whole or in part, on a previous decision that, in turn, might have been reasonable (but not necessarily correct). I think the underlying concern here is that there may be a chain of reasonable decisions which because they have strayed too far from statutory text, purpose or context might be incompatible with a court’s view of the correct interpretation of the statute. Rightly, the Court of Appeal is concerned that liberal application of the correctness standard would throw the authority of many past regulatory decisions into doubt. Notice that the stare decisis analysis in Vavilov was directed to the more abstract issue of the selection and application of the standard of review, whereas the Court of Appeal is concerned with the concrete problem of whether past substantive decisions on regulatory powers continue to bind. There is significant potential for upheaval here. Hence the insistence on a presumption of continuing bindingness.

Professor Bankes cogently argues that the judges are “incorrect” to insist on a presumption of continuing bindingness but I am more sanguine. Once the Court of Appeal’s concerns have been unpacked, it is easier to understand why the judges thought the presumption is appropriate. Now, it might be that the idea of a presumption is unhelpful, because as Professor Bankes notes, it prompts immediate questions about whether and in what circumstances the presumption could be rebutted. For my part, leaving presumptions out of it, I would re-word the Court of Appeal’s advice as follows: where the correctness standard applies, courts have the last word on what a statutory provision means, but in giving the last word, courts should be mindful of settled expectations and practices generated by jurisprudence upholding past decisions as reasonable. Whatever about stating this in terms of a presumption, the Court of Appeal is quite right, in my view, to insist that the change to the correctness standard should not be taken as automatically displacing past jurisprudence applying reasonableness review.

This content has been updated on October 6, 2023 at 11:42.