INDU Committee Report

I think it’s … desirable that the predatory pricing provisions of the Act be decriminalized and … at a minimum, the civil law provides a better basis for evaluating the competitive effects of aggressive pricing behaviour, bearing in mind that if aggressive pricing does not transgress the standard and it does benefit consumers. [Tim Kennish, 44:9:30]

I support the decriminalization of section 50 … I support restricting their application to situations in which competition is likely to be lessened substantially. [Donald McFetridge, 44:9:10]

Predatory Behaviour Defined and the Cost Test

Predatory behaviour occurs when a firm temporarily lowers its prices or expands output or capacity in an attempt to deter new competitors from entering the market or to drive out or discipline competitors who are already there. In all three cases, the predator incurs temporary losses in the expectation of, at the very least, recouping them by raising prices later. Prior to the 1980s, most economists regarded predation as extremely rare because the barriers to entry in most markets were thought to be low. Consequently, it was thought that the subsequent high prices required to recoup the losses suffered in the predatory period would not be sustainable in the face of new entrants. Moreover, predation would be very expensive; the "prey" would be aware that the period of lower prices would be costly for the predator and might hold on in the hope of eventual profits (in the case of efficient capital markets), or to see the predator attempt to buy him out. Only in the extremely rare event that the predator had greater and better access to external capital would a predatory campaign pay off; though even a takeover or merger would generally be a more successful way of monopolizing the market. Recent economic research, however, challenges this long-held position on the grounds that predation may be a more frequent occurrence than previously thought.

I disagree with some of my colleagues in the economic profession, in my view, predatory pricing is not a phenomenon that’s never observed, which was the consensus, or almost never observed, which was the consensus among antitrust economists until, let’s say, five or ten years ago. In my view, predatory pricing is an activity that is a common, although not frequent, part of the business landscape. [Roger Ware, Queen’s University, 52:9:45]

The Committee was told that, as simple as the above definition seems, predatory pricing and behaviour is much more complicated to establish in practice. Consider the following contemporary example:

The essence of predatory pricing is selling below cost, under conditions where the only way it makes sense economically for the predator is if, in the future the predator can raise price and recoup those losses. The necessary conditions are the predator has to have a substantial degree of market power, and second, there have to be sufficiently high barriers to entry so that when the predator subsequently having punished the rival ¾ either banished them ¾ or just subdued them, can then raise the price in the future. If there are no barriers to entry, then they can’t raise the price, because if they do entry will come flooding in and drive it back down to the competitive level.

Those are the two necessary conditions for that to occur. That’s the basic idea. Where it gets complicated is that in the very short run in many service industries, airlines being one of them, the marginal cost is near zero. I did an analysis of this and if you take it on transcontinental flights for Air Canada or Canadian Airlines, once they’re committed to the schedule, as they are every quarter, and they have an empty seat, they should accept anybody who is willing to pay more than $40 or $50, because that covers the additional baggage handling, the commission, a little bit of fuel and a meal.

Now, at the lowest fare, the last one I got, they charged, I think, something like $399. That’s very low relative to the unconstrained economy fare, which is well over $2,000, but it’s still well above that very short run measure of cost. … That means that when you get into that kind of a situation, the competitor which has the deeper pockets, … as I say, if you get into a bleeding match, the person with more blood will survive and the one with less will collapse, and that will be the end of the story. [William Stanbury, 47:16:05]

This broad scope in the pricing of services, whereby the marginal cost can approach zero, makes it extremely difficult to distinguish predatory pricing from aggressive price competition. In the case of perishable goods whose marginal cost is often as close to zero as you can get, "selling, for example, inventory or perishable inventory is a good example of a case where selling below cost is a perfectly legitimate business practice. So that’s why it’s so tricky to determine these things" [Roger Ware, 52:10:00]. Furthermore, modern thinking even questions whether the hard-to-define marginal cost concept is the appropriate test of predatory pricing:

Predatory pricing is rare and very difficult to identify in practice because low prices are what competition is all about. What distinguishes an unreasonably low price in the sense of section 50(1)(c) from a competitive price? A traditional test of predatory pricing is based on costs. Professor VanDuzer states that economists generally agree that prices below marginal costs charged by a dominant firm tend to be predatory. A reading of the predatory pricing guidelines might be construed to support that view. I think the view is wrong.

I think there are examples of predatory pricing but this set of conditions priced below marginal costs, is not sufficient, even if a firm is dominant. Consider just one example, Amazon.com. This firm was founded in 1995, has yet to price above cost, and yet has a stock market value of over $20 billion. It’s pricing less than cost, but it’s not engaged in predatory pricing. Through low prices, it’s investing in a future market share as a new innovator. [Ralph Winter, University of Toronto, 48:9:20]

So there is a temporal aspect to pricing that may not be properly accounted for in the current cost test of predatory pricing. Moreover, the example where below-cost pricing is not predatory pricing was further extended to apply to simple goods such as a razor and razor blades or a number of other complementary products. Apparently, pricing razors below their accounting measures of cost makes good economic sense when it leads to greater sales of razor blades and ultimately greater firm profit. In this case, what should be compared to today’s price is today’s average variable cost minus the present value of the firm’s expected increased gross margin per unit in the future that is attributable to the low-pricing policy. Needless to say, when this last bit of information is gathered by the investigator the "prey" will have given up the struggle. Clearly, economic theory, as a practical guide to enforcement of predatory pricing, leaves something to be desired.

Predatory Behaviour and its Legal Treatment

Predatory pricing is a criminal offence under section 50(1)(c) of the Competition Act. Several elements must be established before an offence is proven. The alleged predator must be engaged in a business and have adopted a policy of selling products at prices that are unreasonably low. Both the "policy" requirement and the "unreasonably low" price requirement have raised difficult issues of interpretation. With respect to a policy, one of four requirements must be met: (1) it must have the effect or tendency of substantially lessening competition; (2) it must have the effect or tendency of eliminating a competitor; (3) it must be designed to substantially lessen competition; or (4) it must be designed to eliminate a competitor.

Professors VanDuzer and Paquet clearly recognize the many problems with the current predatory pricing provision:

Predatory pricing … is really by far the most difficult kind of anticompetitive behaviour for which to work out appropriate rules. The basic provision in the Act currently is a criminal offence, and ... I think the main problem with that provision is that it’s very vague. It’s not at all clear what unreasonably low pricing means. We’ve had very few cases that have interpreted the provision to provide us with any guidance as to exactly what the provision means. The consequence is it’s very difficult to use this provision as a reliable guide to distinguish aggressive competition that results in the reduction of prices from predatory pricing. [Anthony VanDuzer, 14:15:40]

Consequently, its application and effectiveness is suspect:

One of the problems is that they set a very high standard. In order to have prices that are unreasonably low under the guidelines, you have to be able to establish that the party who is the alleged predator, who’s conducting this low-pricing campaign, has enough market power that, after they finish their low-pricing campaign and either put somebody out of business or discipline them or deter them from entering into the market, they will be able to raise prices unilaterally to a level above the level that would apply if it were a competitive marketplace, and recoup all the losses they incurred during the predatory low-pricing campaign, as well as, obviously, some additional profit. The difficulty with that sort of standard is in most circumstances it means you have to make a prediction about how the market’s going to work. And as we all know, … it’s extremely difficult to prove beyond a reasonable doubt — this is a criminal standard — that this test is going to be met. [Anthony VanDuzer, 14:15:40]

The bulk of the testimony provided to the Committee was that the current predatory pricing provision is out-of-date. A typical comment was:

The point is that the law, the way it’s drafted today, was drafted in 1935. It essentially was designed to deal with supermarkets and corner stores, to be honest about it, and it was to deal with goods. Well, we’re in a world now where most of the sales and production is in services and some of those services have the characteristics I mention in which case this law is absolutely useless. We need a new predatory pricing law. [William Stanbury, 47:16:05]

Moreover, it is counterproductive:

With respect to our law, what I’ve said is that certain provisions of it are emasculating the better provisions ... Sections 79 and 78 relating to abusive dominance works well, but the old pricing provisions back in 50 and 51, which have never been enforced, make people think those are the law in that area rather than 79 and 78. Thus the focus … gets back to these old prolix, unenforceable things that I would like to see off the books because then I think people would concentrate on using them in 78 and 79. [Warren Grover, Blake, Cassels & Graydon, 46:10:50]

Finally, as it is currently worded, the predatory pricing provision offends the overriding spirit of the Competition Act, which is to preserve the process of competition and not competitors specifically:

I have a final point on predatory pricing. If I can have one impact before the Committee, I hope that this is it. We should eliminate four words from this section. These four words are: "or eliminating a competitor." Competition policy is about protecting competition, not about protecting firms against competition. … No firm should face the obligation to price high enough to protect its rivals. It’s enough that we have a substantial lessening of competition condition ... The protection of rivals should not be a criterion. Effectively, that is what the phrase — these four words "or eliminating a competitor" implies.

A reading of this section of the Act leads firms who have been unable to survive market competition into believing that they have a valid predatory pricing claim. It leads to cases which are frivolous from an economics point of view and which will ultimately be decided in favour of the defendant because courts have well-developed predatory pricing cases, at least from U.S. law. The wording of the Act is very misleading, and I would suggest it has been very costly in encouraging predatory pricing claims. [Ralph Winter, 48:9:20]

A number of witnesses questioned the view that the problem lies with the legislation, suggesting rather that it rests with the Commissioner of Competition and his reluctance to pursue legitimate predatory cases:

One point that I want to stress is that I believe that the lack of successful cases against predatory pricing in Canada is not a result of any weaknesses in legislation. I think either under the existing section 50(1)(c) or under section 79, many successful cases could have been brought over the last 20 or 25 years. The real reason for a lack of enforcement was because the majority of the economics profession, who had some influence on the Competition Bureau, didn’t believe that predatory pricing was a significant problem. It was believed to be extremely rare, or rare to non-existent. If you look at any industrial organization textbook dating from ten years ago or so, you’ll find that that’s exactly what it will say. Now, there’s been a bit of a change of view by the economics profession over the last ten years. Certainly a substantial number of economists now believe that predatory pricing is quite possible. As I said, it’s a common, but not frequent part of the landscape. Enforcement has not caught up with this change in the economic analysis of this practice. [Roger Ware, 52:9:50]

Many experts appearing before the Committee advocate shifting predatory pricing infractions from the criminal to the civilly reviewable section of the Act. The following comment was typical:

On the predatory pricing provisions, I like the idea of taking them out of the criminal law. They could either … just leave it to [ section] 79 as a case of abuse or you could create a new civil provision on predatory pricing. [Tom Ross, 46:9:20]

One witness noted that the Competition Tribunal has already confirmed this fact:

My preference, in fact, would be to eliminate section 50(1)(c) altogether and to use section 79 of the Act, the abuse of dominance provision, to cover predatory pricing, because I generally feel that the civil provisions are better and a more efficient way of dealing with most any competitive practices and predatory pricing is covered by section 79 under abuse of dominance. In the NutraSweet case … there was an allegation of predatory pricing against NutraSweet, although the Competition Tribunal rejected it, but they did in that rejection agree that predatory pricing cases could be brought under section 79 of the Act. [Roger Ware, 52:9:50]

However, the Commissioner of Competition, the Canadian Bar Association and some members of the Canadian Chamber of Commerce oppose this suggested change; they believe the criminal status best deters egregious anticompetitive conduct and favours more enforcement resources. They believe the double layer of protection (sections 50(1)(c) and 79) against predatory pricing is more appropriate at this time.

The Committee has reservations about this position, believing that it is very likely that "if you don’t use it, you will lose it." There is simply insufficient case law to validate the deterrent effect of section 50(1)(c). Understandably, however, the Commissioner questions the logic of the claim that the low number of trials is an indication of the law’s ineffectiveness:

I would caution you against measuring the Bureau’s effectiveness in terms of the number of cases it brings before the courts or the Competition Tribunal. Formal cases are only one element in a continuum of instruments used to encourage compliance and success is reflected in the level of competition in the economy, not in the number of cases that are brought to the courts. [Konrad von Finckenstein, 9:9:20]

The Committee cannot, however, simply ignore the predatory pricing provision’s inactive and ineffectual history, which includes only two contested cases and these are more than two decades old; nor can the Committee ignore the fact that "decriminalizing … predatory pricing … would make these provisions more enforceable as the Crown would no longer have to meet the criminal burden of proof, which is beyond a reasonable doubt" [Paul Crampton, 53:15:45]. The Committee finds the status quo disquieting.

In fact, the economic theory of predatory pricing is further fraught with complex and confusing factors that will prove problematic for all but the most seasoned competition law expert. As one witness said: "The difficulty of identifying true predatory pricing means that the judgment in predatory pricing cases should be rendered by a specialized tribunal and not by courts" [Ralph Winter, 48:9:20].

For all these reasons, the Committee believes substantial change to the legal treatment of predatory pricing may be required. The Committee does not favour any one proposal at this time, but believes that a two-track approach, whereby the Crown should pursue egregious predatory behaviour under the criminal provision of the Act and the Commissioner of Competition should pursue any other predatory behaviour under the civilly reviewable section of the Act, merits further study. The Committee, therefore, finds that:

The Government of Canada, after consulting with stakeholders, should consider amending paragraphs 50(1)(b) and 50(1)(c) of the Competition Act by replacing the phrase "or designed to have that effect" with the phrase "and designed to have that effect." In this way, the criminal predatory pricing provisions would require evidence of both "pricing below cost" and the intent of "lessening competition or disciplining or eliminating a competitor."

And

The Government of Canada, after consulting with stakeholders, consider adding a new predatory pricing provision in the reviewable civil section of the Competition Act, possibly to be made applicable to the abuse of dominant position provision (section 79). The Government of Canada should also give consideration to ensuring that both the alleged predator has "market power" and the practice in question would "lessen competition substantially." Consideration should be given to introducing new enforcement guidelines for predatory pricing under the abuse of dominant position provision.

And

The Government of Canada should study the impact of amending section 78(i) to state: "selling products at a price lower than average variable cost for the purpose of disciplining or eliminating a competitor."