Anticompetitive pricing practices are frequently the subject of complaints to the Competition Bureau.¹ In the five years beginning April 1, 1994, the Bureau received 931 complaints about alleged unfair pricing practices, such as predatory pricing, price discrimination and price maintenance. Complaints are most often heard from participants in the retail gasoline sector but are common in other sectors of the economy as well.²

Anticompetitive pricing practices are dealt with under various provisions of the Competition Act, notably the criminal prohibitions against predatory pricing, price discrimination and price maintenance. As well, the abuse of dominance provision allows the Bureau to apply to the Competition Tribunal for an order prohibiting a dominant firm or firms from engaging in pricing practices which constitute an abuse of their market power. Every time a complaint is made, it is reviewed by the Competition Bureau and may be the subject of a formal inquiry ultimately leading to enforcement action in some cases, including referral to the Attorney-General for criminal prosecution or an application to the Competition Tribunal. Despite the substantial volume of complaints about pricing practices, however, relatively few have been the subject of formal inquiries, even fewer are the subject of litigation and only a fraction of those have been successful.

Some industry organizations and others have expressed concerns regarding the effectiveness of the Competition Act provisions dealing with pricing practices and the Bureau's enforcement of them. Most recently, these concerns were raised in the context of Bill C-235, a private member?s bill which proposed to amend the Competition Act with the objective of better addressing certain forms of anticompetitive pricing activities. The Bill targeted integrated suppliers charging higher prices to independent retailers than the prices at which they sold the same product at retail. Though the Bill has not been passed by Parliament,³ the discussion surrounding it vividly illustrated the level and nature of concerns in various sectors of the Canadian economy relating to pricing activities. As a consequence of its consideration of the Bill, the Standing Committee on Industry resolved to review the pricing provisions of the Competition Act and their enforcement.

In anticipation of the Industry Committee?s review, in June 1999, the Commissioner engaged the authors of this report to conduct an independent study of the provisions of the Competition Act dealing with anticompetitive pricing and their enforcement by the Bureau. The full text of the authors? mandate is set out in Appendix 1. This report sets out the results of our study.

Scope of Review

Part I of this study is an examination of the economic nature of the main types of anti- competitive pricing practices regulated under the Competition Act, price discrimination, predatory pricing and price maintenance, considering the theoretical and empirical literature on these practices. The purpose of Part I is to identify the basic elements of an appropriate competition policy framework to address anticompetitive pricing practices, including both the criteria needed to identify the anti-competitive aspects of pricing practices and the elements of a legal regime capable of providing relief from such practices in an effective manner. In doing so, the challenges to competition policy posed by the dynamic changes currently taking place in the Canadian economy are considered.

In light of the elements of a competition policy framework identified in Part I, Part II of the report examines the provisions of the Competition Act dealing with anticompetitive pricing practices (the ?Pricing Provisions?), including their interpretation by the courts and by commentators. For comparative purposes, the manner in which anticompetitive pricing practices are addressed in the United States and in the European Union are described. Part II also looks at the enforcement guidelines issued by the Bureau in 1992 in relation to predatory pricing and price discrimination in light of the identified elements of a competition policy framework.

Part III examines the enforcement history of the Bureau in relation to the Pricing Provisions focussing on the past five (5) years. A sampling of the Bureau case files were studied and extensive interviews were conducted with Bureau staff to develop a profile of the Bureau?s experience with pricing complaints. Interviews were also conducted with selected stakeholders who had expressed concerns regarding the application of the pricing provisions. On the basis of this data, the report discusses the Bureau?s enforcement activities, including the application of the case weighting criteria used by the Bureau to determine whether and in what manner to proceed with a case. The purpose of Part III is to assess the consistency of the Bureau's enforcement in practice with both the Pricing Provisions and the elements of an appropriate competition policy framework.

In the final part of the report, Part IV, we draw some conclusions regarding the Pricing Provisions and their enforcement and make some suggestions for improvements.

PART I
Developing a Competition Policy Framework for Anticompetitive Pricing

Introduction

Section 1.1 of the Competition Act⁴ describes the purposes of Canadian competition law as follows:

to maintain and encourage competition in Canada in order to promote the efficiency and adaptability of the Canadian economy, to expand opportunities for Canadian participation in world markets, to ensure that small and medium-sized enterprises have an equitable opportunity to participate in the Canadian economy, and to provide consumers with competitive prices and product choices.

Section 1.1 has been interpreted by the Bureau as endorsing the principle that competition law is geared to the maintenance and promotion of competition as a process, and not to the protection of competitors.⁵ Such an interpretation recognizes that a normal characteristic of competition is that some market participants may not thrive or even survive while others prosper because of their superior competitive performance. This dynamic effect of competition is essential to ensure that the efficiency benefits of competition are realized. Reductions in the number of competitors should be permitted in the interests of efficiency where the survivor is a more efficient competitor, the reduction is not caused by anticompetitive conduct and the marketplace after the reduction in competition remains sufficiently competitive, taking into account potential as well as actual competition. Of course, protecting the competitive process will mean protecting competitors in some situations where they are threatened by anticompetitive conduct or their elimination would result in insufficient remaining competition. Distinguishing anticompetitive conduct from acceptable marketplace behaviour and determining what level of competition is sufficient are extremely difficult.

Because the purpose clause of the Competition Act states that competition is to be sought as a way to ensure opportunities for some particular subsets of enterprises, some competitors may legitimately expect broader protection through this law than a single minded commitment to the competitive process based solely on efficiency considerations would dictate. In other words, the purpose clause may be interpreted as expressing an intention to proscribe anticompetitive behaviour, even where the outcome is the removal of a less efficient competitor and sufficient competition remains in the market place. In this way, protecting fair and equitable opportunities for small and medium sized enterprises could lead to difficult tradeoffs with the promotion of efficiency.⁶

In relation to anticompetitive pricing, the challenge of respecting these sometimes competing interests is made more difficult by the evolving understanding of the economic effects and likely incidence of potentially anticompetitive pricing practices. New insights derived from economic theory combined with new empirical evidence regarding actual pricing practices in the marketplace mean that legislators and enforcement agencies must constantly struggle to ensure that competition law rules deal effectively with pricing practices which are destructive of competition. At the same time, they must ensure that rules and enforcement practices do not discourage aggressive competition through ever more efficient market place strategies. Price cutting which is predatory, for example, must be distinguished from pro-competitive price cutting which is beneficial to consumers.

The challenge of designing effective competition law rules is further complicated by changes in the marketplace. Canadian businesses function in an ever more competitive environment created by the increasing globalization of business activity, itself greatly facilitated by the liberalization of national and international rules on trade and investment. Just as competition continues to increase, the way in which businesses operate is evolving at an increasing rate. Some sectors are experiencing radical changes in the way business is done and all sectors are being transformed by technology, though some more than others. In terms of competition law, the challenge is to ensure that the law continues to fulfil its objectives as the markets to which it applies develop.

For all these reasons, the design of effective rules to deal with anticompetitive pricing is daunting.⁷ Conceptually, this task requires (1) that one develop correct criteria for identifying anticompetitive activity and (2) that these criteria can be applied to provide remedies in a timely and cost effective way. Most of the remainder of Part I deals with the first aspect, identifying anticompetitive pricing practices. First, it sets out the economic characteristics of the three main types of pricing practices dealt with in this study (price discrimination, predatory pricing and price maintenance), the circumstances in which they are anticompetitive and the existing empirical evidence regarding each type of behaviour. Second, we look at some of the challenges to competition policy analysis posed by the new information economy. Finally, with this background, we turn our attention to the second aspect of the challenge of designing appropriate rules: suggesting the elements of a regime to address the anticompetitive aspects of pricing behaviour and establishing criteria which may be used effectively in practice. Parts II evaluates the current provisions of the Competition Act and their interpretation by the Bureau based on the analytical framework developed in this part. Part III discusses the Bureau?s enforcement experience in light of this framework.

Economic Analysis

Introduction

Most economic analysis of competition policy is concerned with how to protect the competitive process by ensuring that markets function efficiently. The challenge is to design a regime which provides relief where pricing behaviour is actually destructive of efficiency enhancing competition. A pricing practice should be considered anticompetitive, for example, where it creates a risk that future prices and other terms of sale will be less favourable to consumers than they would be otherwise. Pricing practices which are simply part of a competitive process of reducing prices must not be affected.⁸

Even though competition policy is broadly seen as attempting to maintain and promote the process of competition per se, arguments developed in support of a particular policy may be motivated not by efficiency concerns but by a desire to protect groups considered meritorious, socially valuable or particularly vulnerable. As noted above, there are echoes of such considerations in the statement of purposes of the Competition Act. This significantly complicates any attempt to evaluate the impact of the Act and its administration.

We will not deal explicitly with these considerations in the following analysis of the explanations of anticompetitive pricing provided by economic theory because they are largely extraneous to economic analysis, but we will return to them in our assessment of the provisions of the Competition Act dealing with anticompetitive pricing in Part II, since such considerations are explicitly mentioned as part of the purpose of the Act.

Price Discrimination

Price discrimination means charging different prices to different customers, whether other businesses or final consumers, for the same product where the differences in price do not reflect differences in the cost to the supplier of serving the customers.⁹ Three conditions are necessary for a firm to discriminate.¹⁰

The firm must have sufficient market power to set price (otherwise customers charged higher prices would choose to purchase from a competing supplier).
The firm must be able to identify different classes of customers with different levels of sensitivity to the price of the product, or, more precisely, different price elasticities of demand. These differences may arise because of different needs, income levels or uses of the product.
There is limited opportunity for customers to resell to each other. It must not be possible for customers paying a low price to sell to those for whom the product is priced more expensively.

Three different forms of discrimination are discernable.¹¹

First degree discrimination, also known as perfect discrimination, in which each unit is sold for the highest possible price each buyer will pay. Perfect discrimination is unattainable in practice, since it is impossible for the seller to identify and exploit very small distinctions in preferences between customers.
Second degree discrimination, in which demand is partitioned into a number of blocks based on the quantity customers prefer to purchase with different prices being charged for each block (i.e. quantity discounts).
Third degree discrimination, where, based on differing price elasticities of demand, buyers are partitioned into different groups with a price set separately for each group. For example, automobile manufacturers may levy different product markups across product lines, posting the highest markup on luxury vehicles.

Empirical evidence confirming the existence of each form of discrimination can be found relatively easily. Discrimination approaching first degree price discrimination by American colleges was found by Tiffany and Ankrom.¹² Wilson provides several examples of second degree price discrimination, including electrical tariffs varying with hours billed.¹³ Third degree discrimination was analysed by Rosenbaum and Ye¹⁴ in a study showing how economic journal publishers use discriminatory pricing with different subscription rates for institutional versus individual subscribers.

Indeed, price discrimination is commonplace.¹⁵ A major bank is engaging in discriminatory behaviour when it offers no fee banking services to students in order to gain their future loyalty. Major software companies like Microsoft might be considered to be discriminating if they tacitly consented to software piracy in the residential market but actively prosecuted piracy by commercial customers. Slive and Bernhart¹⁶ suggest that the drive of a software company to have its product become the industry standard makes acquiescence to piracy by individual consumers (some of whom are also decision-makers in major corporations) a marketing strategy based on discriminatory pricing.

There are a variety of non-price techniques that can be used to effect price discrimination indirectly through non-price requirements. A classic example is the case of tied sales. At one time, IBM had a monopoly on certain types of tabulating equipment. Different customers valued IBM?s equipment quite differently based on the amount that they used the equipment. However, instead of using price discrimination to get the maximum price that each customer was willing to pay, IBM forced customers to buy tabulating cards from the company, and by charging a price for tabulating cards in excess of their cost, IBM was able to discriminate among its customers according to the intensity of their use of the equipment. Block booking and commodity bundling are other examples of non- price requirements imposed by sellers that succeed in enforcing effective price discrimination.

Price discrimination is not inherently anticompetitive. Indeed, it is very difficult to determine simple indicia of anticompetitive price discrimination. Much depends on the circumstances of each case. Often discrimination may be preferable to a situation in which discrimination is not practised. As noted, in order for discrimination to be possible in economic theory, the discriminating firm must have sufficient market power to set prices and force different persons or groups to pay a higher price than the competitive price. Thus discrimination implies the ability to set prices at a supra- competitive level. The appropriate benchmark for assessing price discrimination is not the competitive price, but the price which the discriminator would charge if it could not discriminate.

If price discrimination simply results in expanding a market, an increase in welfare will result. If we assume that some groups of consumers would not ordinarily purchase any product at the price a nondiscriminating seller would charge if it was restricted to a single-pricing strategy, these groups would be better off if the seller was willing and able to sell them product at a lower price. Price discrimination of this kind can increase total output and welfare.¹⁷ The low price buyers are better off and those buying at the price which would otherwise be charged are no worse off.

Discrimination may allow a price discriminating firm to extract a much larger portion of the consumer surplus than what would be the case if discrimination were not permitted. To the extent that the discriminator charges more than it would if discrimination were prohibited, discrimination will impose a loss on the consumers paying the higher price. This loss is transferred to the price discriminator, however. It is not a dead loss to society. How one views the distributive effect will be affected by various factors, including whether the discriminator also discriminates by selling below the price it would charge in a single price world and whether there are efficiencies associated with the discrimination. For example, through discriminating, the discriminator may be able to expand production to a more efficient level. In short, the consequences of the discrimination are difficult to characterize in the abstract.¹⁸

Charging different prices to different customers may be justified in certain circumstances, in which case it is not truly price discrimination. A transaction or information cost difference associated with selling to different customers, will justify charging them different prices. So, for example, charging more to a customer buying low volumes, or less to a high volume buyer may not be discriminatory where the volume discount is justified by cost differences. Differences between the prices a supplier charges to affiliated and non-affiliated distributors may be justifiable on a cost basis, if the transactions costs of dealing with non-affiliated distributors are higher. In a similar way, if different prices are charged at different times, as a result of changes in input costs or demand shifts, or price differentials are transitory, perhaps because they are responding to price changes by a competitor or other market exigencies, the differences are not discriminatory.¹⁹

Discrimination may be anticompetitive if it is engaged in as part of a policy of predatory pricing, where the object is to eliminate or discipline a competitor or deter market entry. This particular form of discriminatory pricing is discussed in the next section.

In some circumstances, price discrimination may be induced by buyers. There are at least two variants of such discrimination. A customer may demand that a supplier charge higher prices to another customer with whom it competes, perhaps because it is unhappy about the low pricing policy of the second customer. Alternatively, a large customer with market power may be able to extract special non-cost justified discounts from a supplier, putting competitors of the customer at a competitive disadvantage. Concerns about this second form of buyer induced discrimination was the main reason for the introduction of the Canadian price discrimination provisions in 1935.

The likely incidence and effect of discrimination of the second situation type has been questioned by some commentators.²⁰ Dunlop, McQueen and Trebilcock have expressed skepticism regarding the advantages to a large customer of extracting large, non-cost justified discounts from suppliers. They argue first that, where the supplier market is competitive and supplier profit margins are low, large non-cost justified discounts to a large customer may threaten the existence of weaker suppliers. If suppliers leave the market, the resulting concentration at the supplier level is not in the large customer?s interest. Their second point is that, if suppliers grant large discounts to one large customer, it may signal the benefits of bargaining and encourage other customers to seek such benefits. It may even draw new large customers from other markets. Such consequences would significantly reduce the benefit to the large buyer of seeking the discount in the first place.²¹

In terms of competitive effect, if new players enter the large customer?s market and successfully obtain the same discounts, the result may be a general lowering of costs to and greater competition amongst firms in the large customer?s market. As a consequence, consumers may benefit from lower prices and increased sales.

Where price concessions are sought by a large customer, there may be anticompetitive effects in the market in which the customer sells. Vigorous price competition in that market by the other customers of the discriminating supplier will be constrained because they will be at a competitive disadvantage relative to the large customer. The nature of the competitive impact will depend upon whether adequate competition remains in the market.

Based on the foregoing discussion, economic analysis does not disclose discrete factors which are sufficient to determine when price discrimination is anticompetitive. In order for price discrimination to occur the discriminator must have market power,²² meaning, in part, the absence of opportunities for customers to change suppliers. As well, customers must not be able to engage in arbitrage by reselling to each other. Price differences justified by differences in transaction or information costs of supplying different customers are not discriminatory. Nor are price differences which are a temporary expedient or a defensive competitive response.

Any competition law provision designed to address anticompetitive price discrimination should be restricted to price discrimination thus defined. Some competitive effects test will also be necessary because the existence and nature of any anticompetitive effect will depend upon the particular circumstances in which discrimination occurs in each case. Assessment of the competitive effects of discrimination will be difficult, imposing a need for significant data and difficult microeconomic forecasts of demand and other variables

Predatory Pricing

Predatory pricing occurs where a firm temporarily charges particularly low prices in an attempt to deter market entry by new competitors, to drive out existing competitors, or to discipline competitors.²³ By setting very low prices in the face of a new competitor, an incumbent firm can prevent market entry. Either the threat of a low pricing policy or the reputation of the firm for predatory behaviour can be sufficient disincentives to deter new market entrants. When predatory pricing is undertaken to discipline competitors, the behaviour is intended to demonstrate the ability of the dominant firm to inflict losses on unruly competitors, who may themselves be engaged in price cutting or other practices that the dominant firm may be concerned about. In all these cases, the predator incurs temporary losses during its low pricing policy with the intention of raising prices in the future to recoup losses and gain further profits. Determining when predatory pricing has occurred is complex and difficult. The main problem is that low pricing is commonly complained about by firms struggling to compete but it is hard to distinguish predation from aggressive competition.

Prior to the 1980's, predation was regarded by economists as likely to be rare. This view was based on the assumption that to become an economically rational strategy for a firm there must be a reasonable prospect of recouping losses after a successful low pricing campaign and that prospects for recoupment are low in the absence of high barriers to entry. If high prices were charged by a supposed predator after successfully eliminating or deterring competitors from entering a market with low barriers to entry, others would enter to take advantage of the high prices and the price would not be sustainable. Likely new entrants might include the market participant who had exited or someone else who had acquired their plant and equipment.

McFetridge canvasses a variety of other theoretical arguments which have been made suggesting that predation is unlikely. Significantly, predation is a very expensive strategy. Not only must the predator finance losses on the sales that it would otherwise have made, it must also service the increased demand generated by its below cost pricing, which, in turn, increases the losses it incurs. If the victim knows that predation is costly for the predator and that the predator will be able to make supra-competitive profits if it leaves the market, the victim has an incentive to either stay in the market until the predator inevitably gives up or try to negotiate with the predator to buy it out. Indeed, the prospect of such a buy out should attract capital market participants to invest in the victim, allowing it to hold out until such a buy out offer is made. Customers and suppliers of the predator and the victim have an incentive to prevent the successful development of the predator's monopoly and may assist the victim to survive.²⁴ Where capital markets operate in such an efficient manner, likely it would be cheaper for the prospective predator to buy out the victim up front than to engage in predation and then buy out the victim.²⁵

More recently, some sophisticated theoretical claims have been made suggesting a wider array of circumstances in which predation may be a rational strategy. These claims are based on models which acknowledge that information and capital markets are not perfectly efficient. The predator and the victim will not have complete information regarding each other and the victim may not have access to sufficient capital to survive the period of predation.

The likelihood of successful predation is claimed to be enhanced under the so called "long purse" theory. Under this theory, predation is more likely to be successful where the predator has better access to capital than the victim. Creditors of the victim will terminate funding or refuse to advance additional funds needed to finance the losses arising during the period of predation due to the weakening of the victim?s financial position caused by the predation. Since predation will also hurt the financial performance of the predator, at least during the period of predation, the predator must have sufficient financial resources to avoid problems with its own creditors.²⁶ In order for there to be such differential access to credit, capital markets must be imperfect. Otherwise, as suggested above, the victim should be able to obtain financing to survive the predatory campaign.

Once "long purse" predation has eliminated current market participants, an explanation must be provided as to why recoupment is possible.²⁷ The most common theoretical claim is that predation is used to create a reputation for toughness which deters future entry, allowing the predator to recoup its investment in predation.²⁸ Predation is less costly because, in effect, initial acts of predation insulate the predator from competition by creating a strategic barrier to entry in the market. Since the cost of predation is reduced while the prospects for recoupment are enhanced, predation should be more likely. This explanation is most compelling in circumstances where the predator is active in multiple markets. If predation in one market creates a reputational barrier to entry in all markets in which the predator participates, the predator may be able to engage in supra- competitive pricing in all markets after a successful predatory campaign in one of them.

Rasmussen²⁹ and others have suggested other theoretical scenarios for predation based on signalling and signal jamming. Incumbent dominant firms may successfully predate by sending signals about the profitability of entering a market which creates a barrier to entry. Lowering price upon entry may be interpreted by prospective entrants as a signal either that demand is weak or that the predator's costs are so low that they can afford to reduce prices. In either case, the intended message may be that there is no prospect of profitable entry. An identical price response by an incumbent firm may be used to send false signals. A firm may lower prices to a level below its costs even when demand is strong in order to falsely signal that demand is weak or that its costs are lower than they actually are. Such ?signal jamming? may be used to keep firms from entering the market by making profitable entry appear more difficult or impossible.³⁰

The plausibility of profitable predation on any of these theories depends upon certain assumptions being fulfilled. In general, the managers of the predator must have incentives to engage in a predatory strategy if the reputation for predation is to be credible. This will mean that their compensation is not linked to short term share value and that they will be protected against losing their jobs during the unprofitable predation period. In a recent study of U.S. firms convicted of predation, Lott concludes that there is no evidence of the management incentives or entrenchment necessary for the theoretical claims to be plausible.³¹ He found, instead, that firms accused of predation tied management compensation to short term profits. Nevertheless, to the extent that pricing decisions are made outside the management group or are not effectively monitored by management, this conclusion does not negate the possibility of predatory strategies being adopted.

The effect of pricing will depend upon its relationship to the costs of current and potential competitors of the predator and its expected duration. In order to discipline a competitor, a predator may need only to drop the price substantially for a short period. It may not need to drop prices below the competitor?s costs to discipline the competitor for discounting or engaging in some other practice that the predator does not like. In order for predation to successfully put a competitor out of business, however, prices may need to be pushed down to a level below the competitor's cost and be maintained at that level for some time. Similarly, in order to deter entry, the potential entrant may need to believe that prices will be below its cost. If the price is above its average total cost, a market participant may still make profits and there will be an incentive to enter. Pricing below a participant's average variable cost will mean there is no incentive to enter and, in the absence of barriers to exit, will cause existing participants to exit. When the price charged falls between average total cost and average variable cost, the effect will depend upon other circumstances in the market.

Nevertheless, even though the effect of predatory pricing is a function of the costs of competitors of the predator, pricing can only be considered predatory where it is below some measure of the predator's cost. Where the predator's costs are lower than those of its current and potential competitors, even a price below the competitor's average variable cost may be profitable, even profit maximizing, for the predator.³² Preventing the alleged predator from taking advantage of its lower cost structure would punish the efficient. Consequently, tests for predation focus on the predator's costs.

The rule suggested by the prominent U.S. antitrust scholars Areeda and Turner is the most noteworthy attempt to formulate a price/cost rule. They assert that a price at or above marginal cost is not predatory and that a price below marginal cost is predatory.³³ Because of the difficulty in measuring marginal cost, Areeda and Turner suggest average variable cost as a proxy in most circumstances.³⁴ Others argue, however, that prices between average total cost and average variable cost may be predatory. Joskow and Klevorick³⁵ suggest that, where prices are in this ?grey zone,? predation may occur where the alleged predator is dominant and the barriers to entry are high enough that post-predation recoupment is feasible. Joskow and Klevorick also suggest that consideration should be given to the dynamic effects of competition on costs including the effects of changing technology. In short, cost evidence alone, typically, is not conclusive when price is in the grey zone.

There may also be business justifications for pricing in the grey zone which are not predatory. Under normal competition, prices may fall below average total cost when firms are seeking to enter a market or expand, where demand is declining or growth is slower than expected or there is excess capacity in the market.³⁶ As well, it is quite naive to presume that an incumbent firm should sit passively in the face of a new aggressive entrant. More reasonably, one may expect the incumbent firm to increase output and reduce prices as a way to prevent its market share from being eroded.

Williamson suggests an output restriction rule focussing on the behaviour of the dominant firm when new firms enter the market as a test for predation.³⁷ He suggests that if post-entry output levels are higher than pre-entry levels, the dominant firm has acted in a predatory way. This is a practical rule in the sense that it is easy to determine if a firm has expanded output, but it is debatable whether the simple fact of reacting to new entrants by putting the pressure on is a satisfactory test of predatory pricing given the business rationales for doing so described above.

Some commentators³⁸ have suggested that evidence of intent is useful to distinguish true predation from procompetitive price cutting. The difficulty with such an approach is that it is often impossible to produce reliable evidence of intent. On the one hand, the language of the market place is not precise and aggressive competition may be expressed in language which sounds predatory. On the other hand, sophisticated business people may be able to disguise intent effectively. It will often be the case that no intent evidence is available. Because of these concerns, other commentators have disputed the value of intent evidence.³⁹

A final consideration relating to the anticompetitive effect of predation is that the magnitude of the effect will depend on the degree to which a competitor who is eliminated contributed to competition in the market. Where the eliminated competitor contributed to competition in a significant way, such as by being especially vigorous, efficient or innovative, its elimination by below cost pricing by a predator, would increase the loss to society associated with the predation.

Few would disagree that predation is anticompetitive in its effects, but its existence in practice is often debated. In a recent U.K. study, over a 10 year period, it was found that only six (6) cases of predation were initiated and in just three (3) was anticompetitive conduct found.⁴⁰ One must be somewhat careful about inferring the absence of predatory activity from such studies, however. An absence of cases, may reflect the allocation of enforcement priority to other types of anticompetitive behaviour, evidentiary challenges in assembling a predation case, problems in designing effective legal rules to address predation or some combination of these factors.⁴¹

Prosecutions and private litigation based on allegations of predation are more plentiful in the United States. Nevertheless, in a review of U.S. court cases in which companies have been successfully prosecuted, some commentators have concluded that many firms had been wrongfully convicted even using a relatively relaxed test for predation: any pricing below average total cost.⁴² Other studies of individual cases, however, have concluded that predation occurred.⁴³

To summarize, the basic indicators of predation may be identified as follows, though none is conclusive.

Market power defined by reference to market shares and barriers to entry. In the absence of market power, the prospect of recouping the costs of a predatory campaign is small.
A policy of selling at prices below some measure of the predator?s cost.
- Where sales are at prices below average total cost and the predator has no pro-competitive explanation, such as
  - meeting competition or changes in demand conditions; or
  - excess supply.
- Where sales are at prices below average variable costs.
Evidence of predatory intent.

This simple listing raises but does not address, the challenge of how each of these indicators may be used in practice. While it is simple to state that market power is needed to make predatory strategies credible, the assessment of market power is inherently problematic. To begin with there are complex issues associated with determining the relevant product and geographic market. As well, while categories of barriers to entry may be readily identified, such as sunk costs and economies of scale, how precisely they may be measured and how to evaluate them is the subject of debate. When are sunk costs high enough to constitute a barrier to entry? Should sunk costs arising from efficiencies of the dominant firm be counted? More difficult still are assessments of barriers derived from slippery concepts such as reputation.

Even an assessment of costs is difficult. In principle, the relevant costs should be the anticipated marginal costs of the predator throughout the period of predation. Given the difficulty of determining marginal costs in practice, average variable costs are often used as a proxy. Even average variable cost, however, may be difficult to ascertain in practice. In U.S. judicial decisions, the determination of costs has been described as more difficult than the market power analysis.⁴⁴

Finally, as noted previously, evidence of subjective intent to predate is both hard to come by and often ambiguous. Though some have suggested it is the only way to distinguish predation from competition, its reliability is questionable and so, as an independent basis for imposing liability, it is often deficient. In some cases of true predation, it will not be obtainable. In cases where there is no prospect of recoupment, intentional predation will not have an adverse effect on consumers. Indeed, consumers will benefit from low prices during the unsuccessful predatory campaign. The only risk is that an effective competitor in the market will be eliminated.

Price Maintenance

Price maintenance occurs where a firm tries to set a minimum price at which another firm can sell its product. It is one of the most pervasive restraints in the market place.⁴⁵ Resale price maintenance may take place vertically, such as between a wholesale supplier and a retailer which resells the supplier's products. It may also be part of a horizontal arrangement between competitors who agree to impose resale price maintenance on resellers of their products.

The economic rationale for prohibiting vertical resale price maintenance under competition law is that it lessens competition by restricting the ability of the retailer to compete on price. It leads to higher prices for consumers and higher margins for retailers, and, in the process, protects inefficient retailers that would not prosper in a truly competitive environment. Resource misallocation may also result where retailers direct excessive resources to non-price competition, such as attention from sales staff, delivery speed and after-sales service. In the absence of price maintenance, competition would be more likely to eliminate less efficient retailers and lead to price and cost reductions in the long run.⁴⁶ Where price maintenance is implemented by a supplier solely in response to pressure from one of the supplier?s large customers seeking to eliminate the low pricing policies of competitors of the customer, the only purpose may be to protect the large customer from price competition.⁴⁷ On the other hand, efficiency is typically served by freedom of contract and many commentators have suggested that vertical resale price maintenance should be permitted, at least in some circumstances.

Horizontally, price maintenance is anticompetitive where it takes the form of an agreement among suppliers to fix their prices. Where suppliers agree not to fix their own prices but to impose resale price maintenance, the anticompetitive effects are also easily identified. Competition among customers is effectively precluded. As well, the practice reduces uncertainty in the market and facilitates collusion among suppliers. In the absence of price maintenance, competing suppliers are uncertain whether low retail prices reflect the decisions of competitors or retailers, and therefore have greater incentive to lower their own prices. In an environment of price maintenance, one would expect suppliers to reach joint profit maximizing levels because they know that retail prices reflect competing supplier prices rather than lower prices resulting from retail competition.

Why do businesses engage in price maintenance?⁴⁸ There are at least three common types of economic rationales suggested.

First, in relation to purely vertical price maintenance, several efficiency explanations are possible. Suppliers may want to encourage resellers to compete on demand determinants other than price, such as service. The retail market may not provide the optimal level of service the supplier desires because of a ?free rider? problem. Discount shops may free ride on the efforts of full-service retailers that provide important pre- and post-sales service on technically complex products such as computers or electronics. Resale price maintenance ensures that resellers have an incentive to offer important consumer services because they are precluded from competing on price.⁴⁹

Suppliers? incentive to engage in this practice would have to be based on a belief that the increase in demand resulting from enhanced service would more than offset the reduction in the level of demand associated with the higher maintained price. While service levels may be specified in a contract, resale price maintenance may have lower transaction and monitoring costs and, consequently, be more efficient.⁵⁰

Another efficiency explanation is that suppliers, such as those in the designer clothing industry, may want to maintain a certain image of their product, which can be damaged by the item being discounted or used as a loss leader. Again, the supplier must be able to conclude that the prestige associated with the higher price means that preventing discounting is profit maximizing. Suppliers may also introduce resale price maintenance to encourage resellers to expand the number of outlets, to provide retailers with an incentive to promote a supplier?s product or to ensure retailers sufficient margins to cover the retailers? cost of quality certification for the supplier?s products.

A second type of rationale is that a cartel among suppliers may be facilitated by an agreement to impose resale prices. By fixing resale prices, monitoring the cartel agreement on wholesale prices would be facilitated, especially where wholesale prices themselves could not be observed easily. As well, suppliers may enter into a cartel for the sole purpose of fixing resale prices as discussed above. A third rationale is that resale price maintenance could be a feature of a cartel at the resale level. Retailers threatened by the entry of discounters may band together to get suppliers to fix resale prices to prevent successful entry by discounters. Marvel and McCafferty describe how the National Association of Druggists in the United States, an ardent proponent of price maintenance, instructed its members to put one supplier, Pepsodent, ?under the counter? when the company discontinued price maintenance in California. As a result, Pepsodent virtually disappeared in California and saw its sales drop by 40 per cent nationally. Pepsodent responded by re-instituting price maintenance.⁵¹

On the basis of their recent study of price maintenance in the United States, Deneckere, Marvel and Peck aptly state that, "...resale price maintenance has the curious status of being both per se illegal and widely practised."⁵² Nevertheless, there is a dearth of empirical evidence on the competitive effects of resale price maintenance which has been lamented by those who concern themselves with its study.⁵³ The existence of price maintenance is not in doubt, but its welfare effects are in question. This is largely because of the numerous reasons for which price maintenance is employed.

In their work described above, Marvel and McCafferty argue that price maintenance is anti- competitive when it is used by manufacturers to gain or maintain monopoly power through control of the distribution system.⁵⁴ Empirical studies, however, suggest that price maintenance associated with such cartels is rare. This is the conclusion reached by Ippolito in her analysis of price maintenance cases brought before the American courts from 1976 to 1982.⁵⁵ Although her review of the evidence found that no single theory was capable of explaining price maintenance activities, her study suggests that purely vertical agreements between suppliers and customers are most common and, consequently, there was scope for the efficiency explanations provided by the service enhancement, brand image and other theories. Competition authorities must be capable of unpacking the various uses of price maintenance in particular cases to identify and measure the anticompetitive effects, including their effects on efficiency.

While all horizontal price maintenance arrangements may be anticompetitive, the determination is more complex for purely vertical arrangements because a determination must be made as to what consumer welfare would be in the absence of price maintenance. Nevertheless, it is possible to identify some of the economic indicia of anticompetitive vertical price maintenance as follows:

The person implementing price maintenance (the ?Supplier?) has market power, a characteristic of which is limited opportunities for customers to change suppliers;
The Supplier does not have an efficiency based justification, such as a desire to increase service or prevent brand impairing practices, which would include loss leadering or misleading advertising; and
The Supplier was induced to implement price maintenance in relation to one customer by another customer who competes with the first.

As the foregoing discussion indicates, the presence of factors 1 and 2 are necessary conditions for price maintenance to have an anticompetitive effect, though the significance of the effect will depend upon the particular circumstances in each case, imposing a need for significant data and difficult microeconomic forecasts of demand and other variables.

Resale price maintenance induced by a large customer, referred to in factor 3, will have an anticompetitive motive where no efficiency justification described in factor 2 is present. Nevertheless, a large customer will only be able to coerce a supplier into implementing price maintenance if the customer has market power itself. Also, resale price maintenance in response to pressure from a large customer will only have an adverse impact on the customer whose prices the Supplier seeks to maintain if factor 1 is also present. Where a supplier does not have market power but, price maintenance is imposed on all suppliers in an industry by a cartel of suppliers or retailers, the result could be anticompetitive.

Challenges of the New Economy

The Canadian economy has become increasingly competitive as a consequence of globalization, due, in part, to the ongoing process of trade liberalization. As well, in certain sectors the channels of distribution have substantially changed. The emergence of ?big box? retailing and internet distribution are both a response to and a cause of increased competitiveness. Even more fundamentally, the economy is currently undergoing a radical transformation; it is becoming more and more knowledge-based and increasingly innovation-driven. These features of the new economy may require a rethinking of competition policy in relation to anticompetitive pricing.

The old economy was driven by manufacturing and tangible commodities and was focussed on the allocation of existing material resources. The new economy is knowledge-based and technology- driven; it is geared to innovation, to the creation of new use-values, products and services.

Competition in the new economy is distinguished from the old by the pace of technological change. Formerly, many firms tended to take their milieu as given; technologies, institutions, preferences as fixed. They tried to optimize its results within this context by choosing the ?right? technology and the ?right? product mix. In many sectors, the role of innovation was relatively small.

Today competition in most sectors has grown due to globalization and accelerated technological change, forcing enterprises to embody a philosophy of continuous improvement and innovation, to become learning organizations in order to remain competitive. To do so, they require an organizational flexibility that they did not previously possess.⁵⁶ Increasingly, learning is the source of wealth-creation and in order to maximize learning opportunities, firms must engage in higher degrees of cooperation. The new economy depends on new modes of collegiality, alliances and sharing of knowledge among firms.

In terms of competition policy, increased competition and technological change mean that, more than ever, efficiency must be considered from a dynamic point of view.⁵⁷ The impact of behaviour on efficiency must be assessed in light of the continuous change generated by intensified competition and innovation which characterizes some sectors, such as information technology.⁵⁸

At the core of the new economy are certain principles which may have profound implications for competition policy.

1. The centrality of technological change to economic growth.

Although it is not a new belief that technological change drives economic growth,⁵⁹ the increasingly rapid pace of technological developments is reinforcing and entrenching the place of technology at the economic centre of growth. This leads to recognition of the importance of innovation to economic growth, and, correspondingly, the significance for public policy of ensuring the optimal conditions for innovative activity. Competition policy must now give more serious consideration to the role of innovation as the lifeblood of the economy.⁶⁰

Recognition of the interplay between innovation, economic growth, and competition policy began to appear in the United States in the mid 1990's.⁶¹ Recently speeches made by heads of the antitrust divisions of both the Department of Justice⁶² and the Federal Trade Commission⁶³ confirm this approach. The importance of innovation is also being recognized in Canada,⁶⁴ but, unfortunately, the Competition Act does not consistently permit such considerations to be addressed.⁶⁵ The Competition Bureau has dealt with some aspects of innovation in its Strategic Alliances under the Competition Act⁶⁶ and the Merger Enforcement Guidelines. Intellectual property rights are the subject of the draft Intellectual Property Enforcement Guidelines published for comment in June 1999. No overarching policy based on the primacy of innovation has been developed.

2. Increasing returns and low or zero marginal cost.

Increasing returns are present when unit costs of production decrease as the firm produces additional product.⁶⁷ Increasing returns are characteristic of many sectors in the new economy,⁶⁸ particularly in industries like information and communications technologies. Firms in such industries may need to incur massive sunk costs in developing a new product, such as a new software program, but, after the first unit is sold, the marginal cost of supplying an additional unit is equal to, or at least very closely approaches, zero. This implies that the challenge lies not in fulfilling demand, but rather, in creating demand for the new product.⁶⁹ One need only think of the phenomenal marketing activities of Microsoft prior to the release of Windows 98 to understand this point.

Increasing returns also operate in industries characterized by economies of scale. ?Superstores?, selling huge volume, benefit from substantial reductions in unit costs the more units they sell. Superstores are able not only to spread their costs over more units but also to obtain large discounts from their suppliers based on the magnitude of their purchasing power. As a consequence, they are capable of turning a profit while significantly undercutting the prices of small independent retail competitors.

3. Monopoly rents are short-lived or non-existent.

In order to protect consumers and ensure that competition is maintained, one of the objectives of competition policy is to control the activities of dominant firms which constitute an abuse of their market power. In some sectors of the new economy, however, even a monopolist, such as a firm offering the latest innovation in software, is not likely to retain this position for long and the corresponding need for competition law enforcement is reduced.

Incumbent firms are unseated rapidly as their products are displaced by different products which meet altogether new needs in the marketplace. Because demand is limited in time, incumbent firms are only likely to earn normal profits because demand subsides before sufficient time has passed to reap super-normal earnings.⁷⁰

4. The new importance of standards

In the new economy, gaining a large market share quickly as a way to establish the dominance of a standard may be an increasingly common business strategy. Standards are an example of network effects. A network effect occurs when the value of a product increases with the number of users. Software is the classic example. The more people using a particular word processing software, the more valuable it becomes. Where a product becomes an industry standard such network effects are substantial. Where standards become a crucial feature of competition, efforts to establish the dominance of a standard, including the use of low pricing, may be part and parcel of competition.

These four (4) factors may necessitate a re-thinking of competition policy. In relation to pricing practices, legitimate efficiency enhancing competition through low pricing practices is likely to become more pervasive, particularly in industries characterized by high rates of innovation, increasing returns and where the prospect of establishing the industry standard may have substantial benefits. At the same time, again in particular industries, technology is driving down barriers to entry, both through innovations in marketing and distribution, such as internet sales, and by creating low cost ways of participating in business. Publishing is an example of a sector where costs of setting up a business have been dramatically reduced, at least in some niches, by technology. Improved access to information is reducing barriers to entry in all markets. In reducing barriers to entry, technology is expanding the scope of geographic and product markets themselves. When one combines declining barriers to entry with increasing threats to dominance in some markets from new products and technologies, the likelihood that dominance can be exploited to injure competition is substantially reduced.

At the same time, a characteristic of an innovation driven market is that the innovator will be dominant, at least for a time and that there may be efficiencies associated with dominance. Richardson, for example, has suggested that larger firms or even dominant firms may show superior innovative performance due to advantages of economies of scale, as well as valuable cumulative learning experience such firms may have gained through prior innovative successes.⁷¹ It is important that competition law enforcement take into account these aspects of competition in the new economy. This means that competition authorities should increasingly emphasize dynamic over static efficiency goals in their enforcement analysis. Dynamic efficiency recognizes, for example, that innovation is essential to efficiency and that the establishment of a standard may be beneficial to consumers and, in any event, that any standard will not be sustainable in the long term since standards themselves are a significant site of competition.⁷²

In the new economy, competition is becoming more, not less, intense. Increasing numbers of small and medium sized independent firms will go out of business, replaced by new firms, many of which will themselves fail. The challenge of accurately identifying and taking enforcement action against anticompetitive pricing behaviour will become more daunting and the Competition Bureau will need to be vigilant to ensure that its enforcement policies are both informed by and sensitive to the exigencies of the new economy.

General Considerations Regarding a Competition Policy Framework for Anticompetitive Pricing

Creating substantive rules and procedures to ensure that effective remedies are provided for anticompetitive pricing practices is challenging. As discussed above, such practices raise difficult issues in terms of how to distinguish behaviour which is destructive of competition from that which is efficient and procompetitive. Even if you are able to define appropriate criteria for determining the behaviour against which action should be taken, there remain substantial additional problems relating to how to take enforcement action against it. Gathering evidence as well as creating procedures which will give timely and appropriate relief in a cost effective way are very difficult. In this section, we focus on the first set of challenges: designing rules to identify anticompetitive pricing practices. The other issues we will take up in the balance of the report.

The central debate regarding the design of competition law rules is what behaviour should be simply prohibited per se as opposed to dealt with on the basis of all the circumstances of each particular case to determine if there is, in fact, an anticompetitive effect, sometimes referred to as a rule of reason approach. A rule of reason approach examines and balances various factors relating to the competitive effect of an activity to determine if the activity is an unreasonable restriction on competition. Such an approach requires a determination of the relevant market and a consideration of its structure and competitive conditions, including barriers to entry, in order to ascertain whether the person engaged in the activity has market power. Such an approach also requires an assessment of the purpose and effect of the activity as well as any efficiency based explanations for it.⁷³ Before discussing what type of approach is appropriate for dealing with anticompetitive pricing, it is useful to make some general observations on the choice between the per se and rule of reason approaches.⁷⁴

A per se approach has certain advantages. It provides clear guidance to business people, consumers and their advisors regarding what is prohibited, allowing them to readily comply with the law and to seek relief when it has been violated. A per se approach also facilitates enforcement activity because the elements of the behaviour which must be proved are clearly set out and do not depend upon complex and ultimately contestable microeconomic arguments regarding the effect on competition.

The main disadvantage of a per se approach is that it tends to be either under or over-inclusive or both. The difficulty of defining precisely the type of behaviour to be prohibited means that per se rules will often not catch all anticompetitive behaviour, will sweep in behaviour which is procompetitive or both. The latter is a more serious problem, not only because it punishes precisely the behaviour that competition policy seeks to promote, but also because it has a chilling effect in the market place. It discourages all businesses from engaging in procompetitive behaviour where there is a risk that they will contravene the law.

The over and under inclusion problem is largely resolved by the rule of reason approach which requires an intensive enquiry into the effect on competition before any behaviour is found to be contrary to the law and a remedy available. The result should be a more accurate assessment of what behaviour is anticompetitive and the chilling effect on procompetitive activity should be substantially attenuated. At the same time, of course, the certainty with which business people and their advisors can know whether particular behaviour will be the subject of successful enforcement action is substantially reduced as well. In addition to its lack of predictability, the rule of reason approach is much more expensive, since data on costs, output and profits and microeconomic forecasts of demand and other variables will be produced by both sides in any proceeding and the adjudicative process, inevitably, will be complex and drawn out.

The choice between these two approaches in relation to a particular category of behaviour must depend, in part, on the plausibility of creating relatively accurate bright line per se rules. This in turn depends upon the economic understanding of the behaviour. In circumstances where a particular behaviour may be either pro or anticompetitive depending on the circumstances, it is hard to justify a per se rule. As the economic understanding of anticompetitive pricing practices has evolved, it is now clear that in many cases there are procompetitive explanations for price discrimination and price maintenance and it is difficult to distinguish procompetitive price cutting from predation. The very fact that our economic understanding is continuing to evolve counsels against a per se standard. Also, a rule of reason approach would permit competition law adjudication to take into account the changes taking place in the Canadian economy including their differential impact on various industries. Moving to a rule of reason approach is consistent with the long term trend here as well as in the United States.⁷⁵

In the Canadian context, a second general issue in connection with designing competition law rules is whether anticompetitive conduct should be a criminal offence or subject to a civil sanction, that is, subject only to an order of the Competition Tribunal prohibiting the anticompetitive conduct or otherwise seeking to restore competition. Given the uncertainty associated with the rule of reason approach, applying it where the consequence may be a criminal conviction with the associated stigma seems inappropriate. To this consideration, one may add the practical observation that it is extremely difficult to prove the conclusions of economic analysis required under a rule of reason approach to the criminal standard of proof, beyond a reasonable doubt.

There are several other factors in favour of a civil approach. Price discrimination, predation and price maintenance are not inherently criminal activities. None involves the moral turpitude associated with conspiracy or bid-rigging. Traditionally, the constitutional basis for the federal government's competence to legislate competition law has been its criminal law power and, as a consequence, when they were created originally the legislative provisions dealing with anticompetitive acts made them criminal offences. More recently, the courts have adopted a broader notion of scope of the criminal law power and acknowledged that federal jurisdiction in relation to competition law may be found under its authority to enact laws relating to trade and commerce as well.⁷⁶ Now there is no impediment to dealing with anticompetitive pricing as reviewable under the civil approach.

A practical reason for doing so is that the civil burden of proof is on the balance of probabilities. Such a burden is much more appropriate, given the inherently contestable nature of rule of reason cases, than the more absolute criminal law burden. As well, the Competition Tribunal includes economic experts with the requisite skills to make difficult assessments regarding competitive effect. This may be particularly important in relation to predation, where the line between anticompetitive low pricing and aggressive competition is a fine one.

Finally, dealing with anticompetitive pricing civilly would add a measure of consistency in the way in which alternative kinds of business strategies are dealt with. There is no obvious reason for dealing with vertical pricing practices criminally, while subjecting other anticompetitive practices occurring in a vertical context to civil review only. Such an argument is most compelling with respect to price discrimination. There are non-price strategies which are functionally equivalent to price discrimination which are only reviewable civilly, either under the abuse of dominance provision or one of the discrete civil provisions. Tied selling and other non-price adjustments to the terms of trade may be used as an alternative to price discrimination to increase the costs to particular purchasers.

Notwithstanding the general advantages of a rule of reason approach, to the extent that there are specific aspects of price discrimination, predation or price maintenance which may be identified using bright line criteria as being never anticompetitive, or always anticompetitive, a per se approach is indicated in the interests of predictability and certainty. The economic analysis set out above suggests that few aspects of each category of pricing practice may qualify for such treatment.

With this background, in Part II we review the existing provisions of the Competition Act dealing with anticompetitive pricing, discussing both the provisions themselves and the limited case law considering them, as well as the Bureau?s published enforcement guidelines for price discrimination and predatory pricing.

Table of Contents | Part II

Common menu bar links

Institutional links

Resources For

The Organization

Proactive Disclosure

Anticompetitive Pricing Practices and the Competition Act

Anticompetitive Pricing Practices and the Competition Act
Theory, Law and Practice

TABLE OF CONTENTS

Introduction

Background

Scope of Review

PART I
Developing a Competition Policy Framework for Anticompetitive Pricing

Introduction

Economic Analysis

Introduction

Price Discrimination

Predatory Pricing

Price Maintenance

Challenges of the New Economy

General Considerations Regarding a Competition Policy Framework for Anticompetitive Pricing

Common menu bar links

Resources For

The Organization

Proactive Disclosure

Anticompetitive Pricing Practices and the Competition Act

Anticompetitive Pricing Practices and the Competition Act Theory, Law and Practice

Anticompetitive Pricing Practices and the Competition Act
Theory, Law and Practice