1 Public
THE COMPETITION TRIBUNAL IN THE MATTER OF the Competition Act, R.S.C. 1985, c. C-34, as amended; AND IN THE MATTER OF an application by the Commissioner of Competition pursuant to section 76 of the Competition Act;
AND IN THE MATTER OF certain agreements or arrangements implemented or enforced by Visa Canada Corporation and MasterCard International Incorporated.
BETWEEN: THE COMMISSIONER OF COMPETITION Applicant - and-
VISA CANADA CORPORATION and MASTERCARD INTERNATIONAL INCORPORATED
Respondents -and-CANADIAN BANKERS ASSOCIATION and THE TORONTO-DOMINION BANK
Intervenors REPLY EXPERT REPORT OF RALPH A. WINTER April 23, 2012
2 Public
I. Overview 1. I previously submitted a report in this matter dated March 12, 2012, concerning the Respondents' Merchant Rules 1 (e.g., the No-Surcharge Rule and Honor All Cards Rule). In my report, I explained the bases for my main conclusions that: (a) the relevant market for the purpose of assessing the competitive effects of the Merchant Rules is the supply of Credit Card Network Services and does not include other payment methods, such as cash or debit cards; and (b) the Merchant Rules influence upwards and discourage the reduction of the prices at which Acquirers supply Credit Card Network Services to merchants and have an adverse effect on competition. 2. On April 10, 2012, the Respondents submitted several reports by experts who disagree with my opinions. In particular, the Respondents filed a report of Dr. Jeffrey Church (the "Church Report") that was limited to responding to the conclusions found in my report. In this report, I will reply to the main assertions made by Dr. Church in his report and explain why, after reviewing Dr. Church's report, my opinions remain unchanged. I have also reviewed the five other expert reports filed by the Respondents and The Toronto-Dominion Bank in this matter, but comment on these reports only in the context of my reply to Dr. Church's report. The sources and documents I rely upon in this report are listed in the footnotes and also in Appendix "B" to this report.
Defined terms herein have the meaning ascribed to them in my report dated March 12, 2012, unless otherwise indicated.
1
3 Public
3. I reply to the following main points arising from the report of Dr. Church: (a) Dr. Church argues that the Merchant Rules are unlike the "typical" or "usual" case of "resale price maintenance" and therefore, section 76 of the Competition Act does not apply to this matter. (b) Dr. Church states that my assessment of the anticompetitive harm resulting from the Merchant Rules is erroneous because it "ignores the implications of the fact that Visa and MasterCard provide network services simultaneously to both acquirers and issuers, i.e., are engaged in a two-sided business". (c ) Although Dr. Church fails to set out what he believes to be the correct relevant market for assessing the competitive effects of the Merchant Rules, he disagrees with my definition of the relevant market and argues that I have made errors in the application of the hypothetical monopolist test in defining the relevant market. ( d) Dr. Church describes a number of purported efficiency rationales for the Merchant Rules, often without citing any analysis or evidence of these rationales. I deal with the issue of efficiency rationales only briefly, as it is discussed in greater detail in the Reply Expert Report of Dr. Frankel dated April 23, 2012. I am in agreement with Dr. Frankel's reply to these comments. (e ) Dr. Church and Dr. Elzinga comment on the effectiveness of surcharges as compared with discounts. Dr. Frankel has also replied to the experts' comments in respect of this issue. I agree with Dr. Frankel's remarks on this issue and for this reason, have not replied to Dr. Church's comments on this particular issue.
2
4 Public
II. Resale Price Maintenance and the Applicability of Section 76 to this Matter 4. Dr. Church makes a number of arguments as to why the price maintenance provision found in section 76 of the Competition Act "cannot" and "should not" be applied to the Merchant Rules. Dr. Elzinga makes similar arguments and my responses to Dr. Church below apply equally to Dr. Elzinga's report. 5. At the outset, I note that although Dr. Church's arguments on this point purport to examine the application of the price maintenance provision from the perspective of "competition policy", they are essentially legal arguments relating to the proper interpretation of section 76. Although I have responded to certain of the points that raise issues of economic or competition policy, questions of legal interpretation clearly fall outside of the expertise of Dr. Church, myself, or any other economic expert. 6. Dr. Church's arguments as to the inapplicability of the price maintenance provision can be summarized in five main points: (a) the Merchant Rules do not meet the "typical" definition of resale price maintenance; (b) the price maintenance provision is inapplicable as there is no "resale" of a product; (c) my approach to examining the application of section 76 to the Merchant Rules employs an alleged improper "reverse causality" by examining whether the adverse effect on competition influences prices upwards, instead of demonstrating that the adverse competitive impact follows from an upward influence on prices;
( d) my interpretation of the price maintenance provision is undesirable from the perspective of competition policy as it would capture all or almost all cases in which the price of an input is influenced upwards due to conduct that adversely affects competition in the market for the input; and
3
5 Public
( e) the Merchant Rules do not limit the price at which either Acquirers or merchants may offer to sell their products.
I discuss each of these arguments below. (a) "Typical" or "Usual" Definitions of Price Maintenance 7. At paragraphs 16 to 19 of his report, Dr. Church examines whether the application of section 76 to the Merchant Rules is "appropriate from the perspective of the economics and policy of price maintenance". Based on this assessment, Dr. Church concludes that the present matter is not like the "typical" or "usual" case of "resale price maintenance". 8. For example, at paragraph 19 of his report, Dr. Church cites a number of "representative definitions" of resale price maintenance in support of his argument that the present matter does not fall into the category of cases "usually" or "typically" considered to constitute "resale price maintenance". 9. As a preliminary point, the suggestion that the Tribunal should decline to grant a remedy on the basis that a case is not consistent with the "usual" case is an "unusual" argument for an economist to advance. Economists provide expert evidence on the competitive effects of conduct, as opposed to arguing that even though conduct may increase prices and impair competition, it should nevertheless escape scrutiny on basis that a legal challenge to such conduct would constitute an atypical case. This is particularly true given that section 76 is a new provision that was enacted in 2009, and has not previously been interpreted or applied by the Tribunal. 10. Although not mentioned by Dr. Church in his report, I acknowledge in paragraph 21 of my first report that the current matter may appear "somewhat unusual" for a section 76 case given that the matter does not involve a physical article, but rather involves a service, and also because the
4
6 Public
upward influence is not in the form of a vertically-imposed price floor. As I state, this matter "is not like a simple case in which a manufacturer places a retail price floor on a pair of jeans". Nevertheless, as discussed below, this matter is not nearly as far from a "typical" price maintenance case as Dr. Church claims. ~-
11. Section 76 employs language that is much broader than the definitions of "resale price maintenance" cited by Dr. Church in his report. If section 76 were intended to be limited (as Dr. Church contends) to forms of vertical agreements that set minimum prices or impose price floors on resellers, then I assume that these limitations would have been incorporated in the language of the provision. Instead, section 76 uses broad language, particularly in the description of the "price maintenance" aspect of the provision. For example, subsection 76(1)(a) includes the phrase "directly or indirectly" as opposed to being confined to "direct" forms of price maintenance. In addition, subsection 76(1)(a) refers to conduct that "has influenced upward" prices, as opposed to confining the section to direct means of setting specific minimum prices or price floors. It is difficult to reconcile Dr. Church's overly narrow interpretation of price maintenance with the actual language of section 76. 12. Further, if one moves away from specific definitions of resale price maintenance found in the various texts cited by Dr. Church and instead focuses upon what has been written by competition enforcement agencies, academics and others regarding the effects of the specific Merchant Rules at issue, the application of the price maintenance provision in section 76 no longer appears at all unusual. 13. For example, in its 2009 complaint against Visa and MasterCard, the New Zealand Commerce Conimission identified the following as among the anticompetitive effects of provisions virtually identical to the Merchant Rules: 5
7 Public
"The Visa MIF [Multilateral Interchange Fee] provisions either alone or together with the other Visa Rules ... have the purpose, effect or likely effect of controlling or maintaining, or providing for the controlling or maintaining of, the [Merchant Service Fees] charged by acquirers for the supply of merchant acquiring services in the Relevant Acquiring Market.
The Visa Rules prevent or hinder competition in the Relevant Acquiring Market by (inter alia): (a) establishing a "floor price" for the [Merchant Service Fee] charged by Visa Banks as acquirers; ... " 2 14. As an example from the academic literature, Professor Adam Levitin describes the effects of the Merchant Rules (which Professor Levitin defines as the "merchant restraints") as follows: "Vertical price-fixing typically invo Ives a wholesaler and a retailer fixing the price of their own product at resale. Vertical resale price fixing of both maximums and minimums is reviewed under the rule of reason standard; it is not a per se violation .....
No-surcharge rules have aspects of both maximum and minimum vertical price-fixing. If merchants are seen as retailers of credit card services to cardholders, then no-surcharge rules are imposing a maximum price level for credit cards - that of competing payment systems. No-surcharge rules are also imposing a minimum price on other payment systems. Typically, vertical price-fixing involves a manufacturer setting an absolute resale price for its own product. Here, though, merchant restraints link the resale price of a network's card product to that of a competitors' product.
Nondifferentiation rules do the same in terms of fixing the price of higher interchange rate cards and lower interchange rate cards-the maximum retail price for the high interchange rate cards is constrained to being the retail price of the low interchange rate cards, and the minimum retail price of the low interchange cards is the retail price of the high interchange cards. " 3 2 Commerce Commission v. Cards NZ Limited et al., Third Amended Statement of Claim, 2 September 2009 (High Court of New Zealand) [MBJAOOOl_00000468] at paras. 56 and 58.
3 Adam J. Levitin, "Priceless? The Economic Cost of Credit Card Merchant Restraints", 55 UCLA Law Review 1321 (2008) at 1401-02. [Emphasis original]
6
8 Public
15. Visa itself has characterized the "No-Surcharge Rule" as a form of resale price maintenance. In a 2008 submission submitted to the United States Department of Justice, Visa stated as follows: "Visa's prohibition on retailer surcharging is implemented through contracts between retailers and their acquirers. These agreements are in essence maximum resale price maintenance agreements that set the maximum price at which the retailer may resell access to the Visa network at zero. " 4 16. In sum, although the present matter is more complex than the typical case of price maintenance for several reasons, including the nature of the product at issue and the absence of an explicit minimum price floor, in my view, the Merchant Rules do constitute price maintenance (for the purposes of section 76 of the Competition Act). As described below and in my first report, the Merchant Rules influence upwards the price paid by merchants for Credit Card Network Services and have an adverse effect on competition. (b) Resale of Credit Card Network Services 17. In paragraph 23 of his report, Dr. Church argues that section 76 is inapplicable to the Merchant Rules because there is no "resale" of products by Acquirers. However, section 76 does not require that a product be "resold". Rather, section 76 applies to an agreement or other prescribed conduct by a supplier that influences upward or discourages the reduction of the price at which that supplier's "customer" supplies or offers to supply a product within Canada. 18. There is no doubt that Acquirers are customers of Visa and MasterCard. As Michael Bradley, Visa's representative on discovery, acknowledged during his examination:
4 "Visa's No-Surcharge and Non-Discrimination Rules" (May 5, 2008) at 20 [GSSS3806_00023400]. 7
9 Public
"Q .... My understanding is that, and I believe you have already confirmed this, but just so we have it, tha~ the customers of Visa Canada Corporation would be issuers and acquirers as opposed to merchants?
A. That's right." 5 19. MasterCard's representative also testified during the examination for discovery that Acquirers are customers of MasterCard:
20. In any event, Visa and MasterCard do, in fact, supply Credit Card Network Services to be "resold" by Acquirers to merchants. In arguing that there is no resale, Dr. Church states that Credit Card Network Services are "not simply resold, physically unchanged, by acquiring banks to merchants or merchants to their customers" [emphasis added]. 21. It cannot be correct to suggest that for section 76 to apply, resellers must be selling precisely the same set of services - no more and no less - than are supplied by the supplier. This is true for several reasons. First, section 76 explicitly refers to conduct that influences upwards or discourages the reduction of the price that "the person's customer ... supplies or offers to supply or
Examination for discovery of Michael Bradley on December 5, 2011, pp. 52-53 and 59, Qs. 187-89 and 205. 8
10 Public advertises~ product within Canada" [emphasis added], as opposed to "the product" or "the same product" as that supplied by the supplier. 22. Second, as a matter of competition policy, it would severely undermine the effectiveness of section 76 if the jurisdiction of the Tribunal under this provision were to be limited to only those cases where the customer resells precisely the same set of services or articles, without modification, as the services or articles supplied by the supplier. On this basis, manufacturers of cars could require dealers to sell at a minimum price, but avoid the application of section 76 because dealers add and sell floor mats and license plate holders as part of the final sale of the vehicles. 23. Third, Dr. Church's narrow interpretation is particularly inappropriate in this case, as it relates to the supply of Credit Card Network Services. Section 76(3) states explicitly that the provision applies to any person who "extends credit by way of credit cards or is otherwise engaged in a business that relates to credit cards". Notably, suppliers engaged in the business of credit cards are the only type of suppliers expressly identified in section 76. 24. Fourth, this is not like the case of a supplier providing one hidden input, such as electricity, into the manufacture of a product. Rather, the Credit Card Network Services supplied by Visa and MasterCard are the main, primary and critical input supplied to Acquirers. This is supported by evidence on the small proportion of value-added by Acquirers as reflected in the component of Merchant Service Fees typically allocated for Acquirer services. For example, I understand that for ·.:· the period between 2008 and 2011, the portion of the average Merchant Service Fee attributable to TD Merchant Services (an Acquirer) on Visa transactions was no more than approximatelY9
7 9
11 Public
25. Indeed, for many of the witnesses in this matter, the portion of Merchant Service Fees attributable to Acquirers is between 1% and 2% of the total Merchant Service Fee. For example,
26. In summary on this point, a "resale" is not required under section 76 of the Competition Act. In any event, at a fundamental level, Visa and MasterCard do resell Credit Card Network Services to merchants through Acquirers: Visa and MasterCard have created networks to which they provide Acquirers with direct access on certain mandatory terms, including an agreement by the Acquirers to apply and enforce the Merchant Rules against the merchants, who are customers of Acquirers. Pursuant to those agreements, Acquirers, in turn, provide the services that merchants require in order to be able to accept credit cards as a form of payment, including, most importantly, access through the Acquirer to the Visa and MasterCard networks. This network access is the central, fundamental and irreplaceable component of the services provided by Acquirers. (c) Allegations Relating to Reverse Causality 27. Dr. Church asserts in paragraph 27 of his report that my analysis of the application of section 76 is incorrect as it "has a reverse order of causality". Specifically, Dr. Church states that based on his "understanding" of section 76, the Tribunal may only intervene where a "party engages in price maintenance that has had ... an adverse effect on competition" [emphasis added]. Dr. Church goes on to state that "the causality in the economics and competition policy literature
10
12 Public
goes from the conduct to an effect on competition: first establishing price maintenance and second an effect on competition from the price maintenance" [emphasis added]. 28. It is notable that Dr. Church does not provide a single citation to the "economics and competition policy literature" he alleges support his arguments regarding the direction of causality. It is also notable that Dr. Church has elected to add an additional term ("that") to a paraphrased version of section 76, as opposed to quoting directly from the section itself. Section 76 explicitly states that the Tribunal may issue a remedy where the following two conditions, among others, are satisfied: (i) a person "by agreement, threat, promise or any like means, has influenced upward, or has discouraged the reduction of, the price at which the person's customer or any other person to whom the product comes for resale supplies or offers to supply or advertises a product within Canada"; and (ii) "the conduct has had, is having or is likely to have an adverse effect on competition in a market" Section 76 specifies no particular "order of causality" between these two conditions that must be ·.-established before the Tribunal may grant a remedy. 29. In the present matter, it is clear that the conduct at issue is implementation and enforcement by Visa and MasterCard of the Merchant Rules through agreements with Acquirers. The issues are whether the Merchant Rules "ha[ ve] influenced upward, or ha[ ve] discouraged the reduction of, the ~: price at which the person's customer or any other person to whom the product comes for resale" and whether the Merchant Rules have had, are having or are likely to have "an adverse effect on competition in a market". 11
13 Public
30. Contrary to Dr. Church's claim, section 76 does not require any specific direction of causality between the upward-influence condition and the adverse-competitive-effects condition. The conditions for the Tribunal to implement a remedy are established if the two conditions are met, irrespective of the direction of causality. 31. To take an example, my first report describes two mechanisms for the competitive harm resulting from the Merchant Rules. Under the second mechanism (the "cost externalization theory"), the price for Credit Card Network Services is higher because downstream cash and debit consumers share in the cost of any increase in Merchant Service Fees. Higher prices for credit card customers and, in particular, for cash or debit customers, result from the Merchant Rules. The adverse impact on the ability of merchants effectively to compete for cash and debit customers with prices specifically set for those customers, and the higher prices that result, are both direct results of the Merchant Rules. It does not make sense (and is of no consequence) to impose a direction-of-causation interpretation on this theory. 32. One implication of Dr. Church's interpretation of section 76 is that the Tribunal can consider only those adverse effects on competition that are caused by an upward influence in prices. In addition to being an absurd outcome, such an interpretation is contradicted by Dr. Church's own application of the adverse effect on competition condition found in section 76. At paragraphs 8 and 56 of his report, Dr. Church equates an adverse effect on competition with the creation, preservation or enhancement of market power, a much broader concept than adverse effects on competition that are caused by increased prices. (d) Interpretation of Price Maintenance Provision Not Overly Broad 33. Dr. Church states in paragraph 24 of his report that a point of "particular salience" is that the price alleged to be influenced upward in the analysis in my report is not the "price of a 12
14 Public
downstream firm", such as the price charged by Acquirers to merchants, but "the 'prices' charged by Visa and MasterCard themselves to acquirers". This is simply incorrect. Many of the passages from my report cited by Dr. Church flatly contradict this position. For example, paragraph 22 of my first report states as follows: Section 76, however, is not confined to agreements that specify a price floor. Subparagraph 76(1)(a)(i) applies to any agreement which "directly or indirectly ... has influenced upward, or has discouraged the reduction of, the price at which the person's customer ... supplies or offers to supply . . . a product within Canada." Visa and MasterCard enter into agreements with their customers, Acquirers, which in turn sell Credit Card Network Services to merchants. The agreements between Visa, MasterCard and their respective Acquirers dictate the terms upon which Credit Card Network Services may be supplied by Acquirers to merchants, including contractual clauses that impose and enforce the Merchant Rules. The Merchant Rules are structured so as to eliminate or substantially reduce important sources of competitive discipline on and between Visa and MasterCard. This substantial reduction or elimination of competition between Visa and MasterCard has the effect of influencing upward and discouraging the reduction of the prices at which Acquirers supply Credit Card Network Services to merchants. From the perspective of economics, the upward influence condition and adverse-competitive-impact condition of section 7 6 are met. [emphasis added]
34. My analysis demonstrates that the Merchant Rules influence upwards or discourage the reduction of the price at which Acquirers supply Credit Card Network Services to merchants. Although the adverse effects on competition allow Visa and MasterCard to charge higher prices to their Acquirers, the source of such adverse effects are the Merchant Rules that Visa and MasterCard require Acquirers to impose on merchants. Both of the mechanisms I describe in my first report are concerned with the upward influence on the prices paid by merchants, in the form of Merchant Service Fees, and adverse effects on competition resulting from the imposition of the Merchant Rules on merchants. 13
15 Public
35. Dr. Church also argues in paragraph 29 of his report that the present matter cannot be distinguished from any case where a supplier increases prices of an input to a reseller, such as a supplier of electricity (see Church Report, para. 23) or a supplier of flour to a baker (see Elzinga Report, para. 26). Similarly, in the same paragraph of his report, Dr. Church argues that if my approach is accepted, "any conduct that adversely affects competition in a market for an input could be prohibited as price maintenance, provided the adverse effect causes a higher input price and there is some pass through of the higher input price to the price of the downstream product". 36. There is no question that the increases in Interchange Fees and Network Fees implemented by Visa and MasterCard have the effect of increasing prices to Acquirers and that such increases are passed on to Merchants in the form of higher Merchant Service Fees. In this way, the present matter is similar to increases in prices of the supplier of electricity or flour. However, in making this argument, Dr. Church fails to recognize an important distinction between the present matter and the other cases he cites, all of which merely involve a supplier increasing the cost of an input to a downstream firm. 37. Through the Merchant Rules and their other respective operating regulations, Visa and MasterCard dictate key terms upon which Acquirers may supply Credit Card Network Services to merchants. Unlike the flour supplier or the electricity supplier that impose no restraints at all on the terms upon which downstream customers supply products, Visa and MasterCard have retained control over the terms upon which Acquirers supply services to merchants. Visa and MasterCard require Acquirers to implement and enforce the Merchant Restraints in agreements between Acquirers and merchants. In this regard, Dr. Church appears to acknowledge in paragraph 23 of his report that the Merchant Rules are a form of "vertical restraint" that is applied by Visa and MasterCard through agreements with Acquirers.
14
16 Public
38. In making the argument that if my approach is accepted "any conduct that adversely affects competition in a market for an input could be prohibited as price maintenance" provided there is pass-through of a higher price, Dr. Church ignores an important fact. The upward influence on prices and the adverse effects on competition in this case result not merely from conduct that adversely affects competition for an input, but are the results of vertical restraints applied on merchants by Visa and MasterCard through agreements with Acquirers. Dr. Church's claim that any anticompetitive conduct in an upstream market could, as a matter of law, be prohibited under section 76 is speculation by Dr. Church, and in any event, is clearly outside the scope of his expertise. He is assuming, for example, that the Tribunal's decision in this matter would create precedent applicable to cases that do not involve a vertical restraint. As an economic expert, I have focused on the economic impact of the rules in this case: in particular, the upward influence of the Merchant Rules on prices and the adverse impact of the Merchant Rules on competition. 39. I do find it surprising, however, that Dr. Church fails to offer any explanation as to why it is allegedly preferable or appropriate, as a matter of competition policy, to permit vertical restraints (such as the Merchant Rules) that increase prices and adversely effect competition to avoid scrutiny by the Tribunal because conduct that adversely affects competition in a market for an input may potentially be captured within the scope of section 76. Dr. Church provides no economic example of a pro-competitive or competitively neutral arrangement that would be improperly captured by the broader interpretation of price maintenance he criticizes. Indeed, Dr. Church cannot provide such an example because even under the broader interpretation of price maintenance, to which he objects, there remains the requirement that the vertical restraint have an adverse effect on competition.
15
17 Public
40. In contrast, there are obvious competition policy concerns resulting from the narrow interpretation of price maintenance advocated by Dr. Church. Following my analysis, the Merchant Rules result in higher prices and a suppression of competition between Visa and MasterCard, which makes the Merchant Rules exactly the type of conduct that competition policy is intended to proscribe. Admittedly, the Respondents' experts disagree with this point. The preferable approach under competition policy is to allow the matter to be resolved based upon an assessment of the economic impact of the Merchant Rules, as opposed to seeking to shield conduct with adverse competitive effects from scrutiny on the basis that the conduct is allegedly unlike that which is seen in a "typical" price maintenance case. (e) Upwards Influence of Price Acquirers Supply Credit Card Network Services 41. Dr. Church states in paragraph 22 of his report that the Merchant Rules "do not limit in any way the price at which either acquirers or merchants may offer to sell their products". 42. Contrary to what Dr. Church alleges, the Merchant Rules do impose restrictions on the prices at which merchants may offer to sell their products. For example, the No-Surcharge Rule requires that from the consumer's perspective, the differential price of credit card acceptance services be zero. Restricting merchants from applying additional charges depending on the form of payment restricts the prices at which merchants may offer to sell their products and services, one component of which is the acceptance of payment by credit card. 43. Dr. Church also alleges in paragraph 22 that "even if the Rules could be characterized as price maintenance, it would be maximum price maintenance, influencing prices downward, not minimum price maintenance which influences prices upward." As I have explained in my first report, the Merchant Rules operate by restricting the price of purchases made using a credit card to the price of purchases made using other methods of payment, such as cash or debit cards. With the 16
18 Public
resulting suppression of competition between Visa and MasterCard, and the externalization of costs to cash and debit customers, the straightforward effect of the Rules is an influencing upwards of the prices paid by merchants for Credit Card Network Services. II. Anticompetitive Harm Resulting from the Merchant Rules (a) Two-Sided Nature of the Market is Incorporated in My Evidence 44. Dr. Church disagrees with my conclusions regarding the anticompetitive harm arising from the Merchant Rules as described in my report. His comments largely centre on the unfounded suggestion that I have ignored the two-sided nature of the market for Credit Card Network Services. For example, at paragraph 35 of his report, Dr. Church states the "difference in assessment arises principally because Professor Winter ignores the implications of the fact that Visa and MasterCard provide network services simultaneously to both acquirers and issuers, i.e., are engaged in a two-sided business". 45. Far from "ignoring" the two-sided nature of the market, my first report explicitly discusses and considers the two-sided nature of the market. Where Dr. Church and I actually differ is on the issue of whether the anticompetitive effects of the Merchant Rules can be defended on the basis that higher Acquirer Fees and, consequently, higher Merchant Service Fees (on one side of the market) allow Visa and MasterCard to allocate more revenue to Issuers (on the other side of the market). I explain this distinction in further detail below. 46. Let me start with a point on which Dr. Church and I agree. Higher prices on the Acquirer/Merchant side of the market, if sustained through the Merchant Rules under the first mechanism described in my first report, lead to greater expenditures by Issuers on issuing activities on the other side of the market. Higher Acquirer Fees being sustained in the market mainly take the 17
19 Public
form of higher Interchange Fees, which are the portion of Card Acceptance Fees that Visa and MasterCard direct towards Issuers and which may be used for issuing activities. The higher Interchange Fees allow Issuers to make higher expenditures on various issuing activities, such as higher rewards offered to cardholders, like air miles, insurance, lower interest rates, advertising and promotion of the credit card, or rebates on purchases. 4 7. Our key point of disagreement is whether the Tribunal must balance these allegedly "offsetting effects" to determine the "net" competitive impact of the Merchant Rules. The answer, in my view, is no. Neither from the perspective of competition policy generally, nor from the perspective of price maintenance specifically, is it necessary to attempt to "net" out any increase in competition on the issuing side of the market from any decrease in competition on the acquiring/merchant side. It is appropriate, in assessing whether the conditions of section 76 are met in this case, to focus on the acquirer/merchant side of the market. 48. My initial report offers the following rationale for focusing upon one side of a two-sided market in assessing the competitive effects of the Merchant Rules. First, as a background point, note that the "balancing" problem that a credit card company faces between low prices for merchants and increased expenditures on issuing activities is the same problem that any firm faces in balancing between low prices and non-price promotions or strategies. I develop this point at paragraph 76 of my first report. Issuing activities fall within the class of non-price demand-increasing strategies, such as promotion and quality enhancement, that virtually any firm engages in, whether in a one~sided or two-sided market. The choice of price versus non-price strategies is analytically identical in a two-sided market to the choice of price in a one-sided market; balancing two sides of a two-sided market is identical to balancing price versus non-price demand-increasing activities in a one-sided market.
18
20 Public
49. Second, in ·competition policy generally; a practice that raises prices as a result of a suppression of competition - whether this practice is a vertical restraint, a facilitating practice leading to less competition, or explicit collusion - cannot be defended on the basis that the excess revenues are being spent productively on non-price dimensions such as promotion, quality- r. i-enhancement or innovation.
50. To take an example far from the current case, firms engaged in explicit collusion, cannot defend their actions with evidence that excessive revenues are being spent on higher product quality or promotion. Similarly, in a two-sided market context, firms engaged in explicit collusion on one-side of a market cannot defend their actions on the basis that the collusion provides greater revenues on the other side of the market. Nothing in the fact that Visa and MasterCard designate Issuers to perform issuing activities (in a two-sided market structure) rather than undertake them internally (as one-sided firms) changes this competition policy principle. 51. Consider a magazine as an example of a two sided market, the same example adopted by Dr. Elzinga in his report. The two sides of the market in this example are advertisers and readers. Competition policy would not permit a price-fixing cartel among publishers of magazines to set advertising rates, even where evidence were brought that higher prices charged to advertisers might somehow be accompanied by lower costs for readers (e.g. giving away magazines for free). Nothing in the two-sided nature of a market should exempt (or does exempt) firms from laws proscribing conduct with adverse impacts on competition. 52. Competition policy is based on the principle that markets without anticompetitive activity should be relied upon to yield the appropriate mix of price and non-price competition. This principle is as applicable to two-sided markets as it is to one-sided markets. Section 76 specifically, like competition law generally, does not involve a consideration of trade-offs with respect to non-19
21 Public
price expenditures, such as issuing activities, but rather focuses on the anticompetitive conduct at issue. The two-sided versus one-sided aspect of the case is irrelevant to this conclusion. (b) Cost Externalization Effect and the Two-Sided Market 53. In my initial report, I described two mechanisms through which the Merchant Rules have an upward influence on prices and an anticompetitive effect on a market. I referred to the second of these as a "cost externalization effect". To illustrate this effect with a simple example, consider a firm supplying coffee beans to coffee houses. Suppose the coffee houses also sell tea to an equal number of tea drinkers. Suppose further that the supplier of coffee has market power, but that the tea is supplied by a competitive market and is less costly. Now suppose that, like the No-Surcharge Rule in this matter, the coffee supplier imposes a restraint on the downstream coffee houses dictating that its product, once brewed, shall not be sold at a greater price than tea. The result is an immediate upwards influence on the price of tea, as the price of both tea and coffee must now reflect the average cost of inputs for the two drinks. 54. A second upwards influence on prices follows from the fact that the demand facing the upstream coffee supplier becomes less sensitive to price as a result of the restraint. Where there is an equal number of tea and coffee drinkers, downstream coffee drinkers bear only half the cost of a $1 per pound increase in the price of coffee, and tea drinkers bear the remaining cost. The sensitivity, or elasticity, of demand is cut in half. Like any firm facing a less elastic demand, the coffee producer will increase its price. 55. With the two upward effects on prices, tea drinkers face an unambiguously higher price for their drinks. The effect on coffee drinkers, however, is ambiguous. They may face a price that is, on net, lower due to the cross-subsidy (the first effect) or higher due to the coffee producer's incentive to raise price under the restraint (the second effect). 20
22 Public
56. Nothing in this example depends upon tea and coffee being substitutes. Tea drinkers and coffee drinkers may be entirely different people. However, if there is a degree of substitution between tea and coffee, the upward influence on prices is strengthened. Suppose that tea drinkers suddenly discover a taste for coffee, introducing some degree of substitution between the two products. This adds a new benefit to the coffee producer from raising price: with a price increase (under the restraint) some tea drinkers will be induced to switch to coffee. The new effect presents the coffee bean supplier with the incentive to raise prices even further. A small or moderate degree of substitution, in other words, strengthens the cost-externalization effect. 57. These sources of upward influence on prices apply directly to the No-Surcharge Rule, one of the Merchant Rules at issue. Visa and MasterCard each gain from spreading costs across cash and debit customers rather than having the costs paid by their own customers. An incentive to raise prices is induced via this cost-externalization effect, on the part of both Visa and MasterCard. The effect depends upon there being a mix of cash/debit customers and credit card customers within the same sectors (100% of one or the other negates the effect). The theory does not require that cash and debit be substitutes for credit cards, but is strengthened by a small or moderate degree of substitution. 58. The effect of the No-Surcharge Rule in its upward influence on prices under the costexternalization mechanism is strengthened by the incorporation of the two-sided market structure of credit card network services. Instead of simply sharing the costs of credit card transactions, cash and debit consumers can be made to bear all or most of the cost of credit cards. In paragraphs 95 through 102 of my first report, I offer a simple example in which an increase in the Interchange Fee by 10 basis points on the part of the credit card company under the No-Surcharge Rule results in a gain to credit card customers of 5 basis points, due to the rewards and incentives received through
21
23 Public
the use of credit cards. I review and extend this example in Appendix "A" to this reply report in an effort to demonstrate the mechanism by which a credit card company in a four-party system can use a no-surcharge rule to direct the entire cost of a price increase to cash and debit consumers. 59. The example abstracts from the complexities of the actual market so as to illustrate most clearly the mechanism. Nevertheless, the point of the example is valid for actual markets, in spite of their complexity. Under the No-Surcharge Rule, increases in the Acquirer Fee are passed on to all consumers through increases in Merchant Service Fees, as these fees are passed on by merchants via increases in the prices for products. But the credit card company can compensate its own cardholders for these higher prices through increases in the Interchange Fee, since this Fee could be passed on to cardholders in the form of discounts or other benefits. In this way, the two-sided market structure allows the cost of price increases on the Acquirer/merchant side of the market to be focused on cash and debit customers. Thus, two-sided market considerations strengthen the cost-externalization effects. 60. As described in my first report, the cost externalization effect leads to an upward influence on prices to cash and debt customers for three reasons. First, at any given Acquirer Fee, the subsidy from cash and debit customers to credit card customers raises the price of transactions to cash and debit customers, compared to a world without the Merchant Rules. Second, because of the cost-externalization effect, the credit card company's incentive is to raise the Acquirer Fee, especially if (given the two-sided nature of the market) the Fee increase can be focused on the cash and debt customers. Third, if cash and debit are substitutes, to a modest degree, for credit cards at prevailing market fees, the incentive for the credit card company to raise the Acquirer Fee is further increased, because of the resulting substitution of cash and debit transactions towards credit cards.
22
24 Public
. 61. Acquirer Fees, and therefore Merchant Service Fees, are also influenced upwards by this cost externalization, since the second and third effects both relate to the incentive to raise Acquirer Fees. 62. In summary on this point, I have not "ignored" the two-sided nature of the market in examining the adverse effects on competition of the Merchant Rules. In fact, the second adverse competitive effect described in my report (the cost-externalization effect) is actually strengthened if one considers the two-sided nature of the market. (c) Competition Suppression Effect and the Two-Sided Market 63. At paragraph 32 of his report, Dr. Church states Professor Winter's "suppression" theory ("Adverse Effects I") assumes that in the absence of the Rules, competition among acquirers and merchants would result in (widespread) surcharging and this surcharging would steer a substantial volume of consumers to use alternative forms of payment that have lower or no surcharges. Thus, a card platform that lowered its Acquirer Fees might be rewarded by an increase in usage and card membership, according to his theory.
64. This paragraph is accurate but for one statement. The first source of an adverse effect on competition described in my first report (the "competition-suppression effect") does not assume that in the absence of the Merchant Rules competition among Acquirers and merchants would result in widespread surcharging. Surcharging need not be widespread in order to increase competition between Visa and MasterCard. For example, the mere threat of a surcharge or the possibility of surcharging can be an effective means of creating an incentive for Visa and MasterCard to decrease Acquirer Fees.
23
25 Public
65. Although I believe that a significant number of merchants would adopt a surcharging policy, based on the Australian experience (and the fact that in Canada, with higher interchange fees, the incentive to surcharge would be even higher), the theory does not require it. The potential use of surcharging would be substantial, I presume, because otherwise Visa and MasterCard would not be so strongly defending the No-Surcharge Rule. 66. In discussing my competition-suppression effect, Dr. Church states at paragraph 46 of his report that "by focusing only on the acquirer side of the platform, [Professor Winter's] analysis does not recognize the ability for card platforms to increase transaction volume on the issuer side of the platform." This is the central theme of section 3.2 of Dr. Church's report, but it relies upon a mischaracterization of my analysis. My analysis recognizes the accepted economic result that suppression of price competition can intensify non-price competition. In this matter, higher Acquirer Fees are linked to higher Interchange Fees (the largest component of an Acquirer Fee) and therefore, can fund greater issuing activities. As reviewed above, I have incorporated this aspect in my analysis. 67. Dr. Church offers the following conclusion with respect to his competitive impact analysis of the Merchant Rules at paragraph 49 of his report: My conclusion is that with the Rules competition between Visa and MasterCard reduces the network access fee for issuers until network revenues equal network costs. The Rules - even assuming Professor Winter is correct that there would be higher acquirer fees - would not affect market power of Visa and MasterCard. Hence, the Rules cannot have an adverse effect on competition: they do not enhance, create or maintain market power.
68. Dr. Church's assertion may be valid in an implausible world where no one uses cash or debit and higher Interchange Fees are passed by Issuers to cardholders on a one-for-one basis, 24
26 Public
through (for example) larger consumer discounts to cardholders. In the real world, however, customers do use cash or debit and not all of the Interchange Fees are passed through to cardholders in the form of discounts. Rather, Interchange Fees are also spent on promotion and other activities that are not substitutes for discounts for cardholders. I am not aware of any economic theory that suggests that in a duopoly in which price competition is suppressed, non-price competition in dimensions such as promotion will be so intensified as to neutralize the impact of the suppressed price competition. 8 69. In addition, evidence reviewed in my first report indicates that network fees by MasterCard and Visa have increased substantially in the period since these companies were restructured in
Moreover, these figures do not include changes in other components of MasterCard's total network fees. Nothing in the evidence, in short, indicates that the higher prices from the suppression of price competition under the first theory in my report have been allocated entirely to greater expenditures on issuing activities via increases in the Interchange Fee.
8 George Stigler, in a classic economic article, "Price and Non-Price Competition," was the first economist to address the issue of whether non-price competition would offset a suppression of price competition. (In Stigler's article, the suppression of competition takes the form of collusion on prices, but the analysis extends to our context.) Stigler refers to the "common belief of economists that price competition is much more effective in increasing output and reducing profits than non-price competition" (p.151) and traces this common belief to reasonable assumptions on technology and demand. See: George Stigler, "Price and Non-Price Competition," Journal of Political Economy, Vol. 76, No. 1 (Jan. - Feb., 1968) pp. 149-154.
25
27 Public
III. Market Definition 70. I begin my reply to Dr. Church's assessment of my analysis on market definition by noting that although Dr. Church has criticized my approach to defining the relevant market, he fails to set out his own view on the appropriate relevant market. Specifically, nowhere does Dr. Church state that he agrees with the broad market definition proffered by Visa and MasterCard that includes credit cards and all other methods of payment, such as cheques, money orders, payments through text messaging and electronic fund transfers. Nor has Dr. Church provided any analysis that would support the definition of the relevant market alleged by Visa and MasterCard. 71. In terms of his critique of my market definition, Dr. Church argues that I have improperly applied the hypothetical monopolist test to the product defined as credit card network services. Dr. Church objects to my application of the test on two grounds: (a) my calculations in arriving at a critical loss are incorrect, even within the framework I adopt (his paragraph 76); and (b) I pose the incorrect question in applying the test (his paragraphs 73 and 74) by focusing on whether a hypothetical monopolist _"could" impose a small but significant and non-transitory increase in price ("SSNIP"), as opposed to whether the hypothetical monopolist "would" impose such an increase. I will discuss these critiques in this order. 72. On the first of these critiques, my point in applying the hypothetical monopolist test to the set of products consisting of credit card network services (as a candidate for a relevant product market) is simple. Acquirers are the immediate customers of these services, but acquirers are intermediaries who pass on the services to merchants. A monopolist producer of credit card network services, by raising its price by 10 basis points (which I take as a SSNIP) above the competitive level would more than double its profit per unit. This monopolist producer would therefore have to lose more than half its demand to other transactions methods in response to a 10 26
28 Public
basis point price increase in order not to profit from the price increase. In other words, the "required loss" is 50 percent (the "BECSL" in the notation that Dr. Church uses). 73. Demand for credit card network services is obviously nowhere near this elastic. Therefore, the relevant product market is no larger than credit card network services. Specifically, in response to a 10 basis point (0.1 percent) increase in the pnce of credit card network services, the monopolist supplier will certainly not lose 50 percent of network volume, which is the amount necessary to render the price increase unprofitable. 74. The data underlying this calculation consist of a 5 percent SSNIP and (roughly) an 81 percent margin of network fees over cost. Using these two data points in a standard formula for the required loss yields, in Dr. Church's calculation: 9 Required loss= (0.05) I (0.05 + 0.81) = 0.058, or 5.8 percent. 75. Dr. Church's calculation of the loss required for a hypothetical monopolist not to profit from a SSNIP, and therefore the amount sufficient for the relevant market to be expanded beyond credit card network services, is only 5.8 percent - much smaller than my estimate of more than 50 percent. 76. However, Dr. Church errs in using 81 percent as a margin of the total price, or total Acquirer Fee, over marginal cost. In fact, as explained at paragraph 62 of my report, 81 percent is MasterCard's margin from its total Network Fee, not its total price. In order for Dr. Church's calculations to be correct, he must be consistent, using the same base for the 5 percent SSNIP as for the margin. His calculations are not consistent in this regard because he applies the 5 percent SSNIP to the Network Fee, but applies the margin for the total Card Acceptance Fee. Correcting See J. Church and R. Ware, Industrial Organization: A Strategic Approach (2000) McGraw-Hill at 609. 27
29 Public
his error, and assuming (for the sake of this analysis) that competitive network fees are no more than 12 basis points and total competitive Acquirer Fees to be approximately 200 basis points, a corrected calculation yields a margin of approximately (12/200)*(0.81) = 0.05. 10 Plugging this into the formula above yields the 50% figure that I use in my first report: Required Loss = (0.05) I (0.05 +0.05) = 50 percent. 77. Again, however, rather than plugging the data into a formula it is simpler to point out that because the hypothetical monopolist would more than double its profit per unit from a SSNIP, it would have to lose more than half its demand not to profit from the SSNIP. Consequently, in order for the market to be broadened beyond Credit Card Network Services, demand in response to the 10 basis point increase would have to drop by more than half. The hypothetical monopolist test is easily met in the present matter, as the required loss to make a price increase unprofitable is extremely large. 78. I tum next to Dr. Church's second, conceptual critique of my application of the hypothetical monopolist test. In my report, I pose a standard formulation of the hypothetical monopolist question: could a hypothetical monopolist profit from a small but significant and non-transitory increase in price (SSNIP) above competitive levels. I take 5 percent to represent small but significant, as is conventional. 79. Dr Church's first objection to my application is that I ask whether the hypothetical monopolist could profit from a 5 percent price increase above competitive levels. But in theory the right question is slightly different: would the hypothetical monopolist, in its profit-maximizing choice, choose a price increase of at least 5 percent above the competitive level. IO Talcing a smaller value for the total competitive Acquirer Fee (160 basis points) still yields a corrected margin of (12/160)*(0.81) = 0.06 and a required loss of (0.05) I (0.05+0.06) = 45 percent, for a hypothetical price increase of (0.05)* 160 = 8 basis points. This is not inconsistent with figures discussed in my expert report (a required loss of more than 50 percent for a price increase of IO basis points).
28
30 Public
80. The distinction between the "could test" and the "would test" is relevant only if there is evidence to suggest that a hypothetical monopolist would maximize profit by increasing price to a level, say, 3 percent above the competitive price level. In this case, the "could test" might be met in the sense that if the monopolist raised price by 5 percent, profit would increase. But this test would be inappropriate because in fact the monopolist would raise price by only 3 percent. The "actual" price increase that the hypothetical would implement is relevant; the profitability of a 5 percent price increase is not. 81. Dr. Church offers no evidence that the profit maximizing price is less than 5 percent, nor does he offer any other evidence to demonstrate that the technical "would versus could" distinction is at all meaningful in this case. The distinction between the "would" and "could" issue is what Dr. Elzinga would describe as a "red herring". 82. As explained above, substantial increases in the price for credit card network services result in very small decreases in demand. As a consequence, a price increase of only 3 percent above competitive levels would not be profit-maximizing. Increasing the price only 2 percent more, to 5 percent, would significantly increase margins, with a very small decrease in demand. The "could versus would" issue simply does not arise in this case. 83. As noted above, Dr. Church, in his response to the market definition conclusion that I reach, does not defend or even offer a conclusion of his own on the definition of the relevant market. However, we can easily start with Dr. Church's concept of the appropriate price in the credit card market and identify where his assumptions would lead. 84. Dr. Church concludes at 63 that in a two-sided platform market: 29
31 Public
The price for a transaction is the sum of the two network access fees paid to the card networks (acquirer and issuer access fees). This is price for using a platform to complete a transaction. The assessment of market power by Visa and MasterCard and defining the relevant market depends on the profit margin of Visa and MasterCard. This is the difference between the price to complete a transaction and the marginal cost of completing a transaction.
85. If one follows the approach suggested by Dr. Church, the "price" is the sum of the competitive network fees, which we can assume here (for simplicity) to be at the current levels of approximately 12 basis points. A SSNIP, or 5 percent increase in these fees, would be less than 1 basis point. Could anyone seriously suggest that a monopolist would not benefit from raising network fees 1 basis point above the competitive level of 12 basis points? Surely the answer must be "no". Under Dr. Church's concept of price, a hypothetical monopolist of credit cards would surely benefit from raising price by a SSNIP above the competitive level. There would be no basis for extending the relevant market beyond credit card network services. Dr. Church's proposed approach yields even stronger support for credit card network services as a relevant market than my approach. IV. Purported Efficiency Rationales for the Merchant Rules 86. Dr. Church offers, with little or no analysis, several pro-competitive or efficiency explanations for the Merchant Rules. The first of these explanations starts with the following assertion, drawn from Dr. Elzinga's report: "The Rules insure that holders of a credit card or type of credit card will not be discriminated against by merchants. [footnote omitted] Merchants cannot discriminate against a network's cards by declining
30
32 Public
or discouraging their acceptance or charging an additional fee for their use." u
87. Consumers using a credit card, especially a premium credit card, are being provided a service that may well involve additional cost for the merchant. By "discrimination", Drs. Church and Elzinga appear to mean the circumstance where a customer bears the cost of the additional services they elect to use; in this case, the use of a premium credit card that costs the merchant more than other payment methods. To other economists, payment for additional services is a normal feature of competitive markets, rather than a form of "discrimination". It is certainly not inefficient. Quite the opposite - prices for payment by specific credit cards allows efficient choices of payment methods by customers. This, indeed, is the function of the price system. 88. Including surcharging as part of a price system ensures that the benefits from a decision (here the use of a particular credit card) exceed the costs (here the merchant's costs as reflected in the surcharge). The Merchant Rules suppress the price system and impair its ability to direct efficient choices by the consumer, especially among credit cards. The concept of "discrimination" adopted by Dr. Church and Dr. Elzinga is completely at odds with the understanding of this term in economics. > ' . 89. Dr. Church also alleges that the Merchant Rules prevent "free-riding" by merchants and the "hold-up" of consumers. For example, Dr. Church states at paragraphs 52 to 53 of his report: If the Rules are not in place, consumers are susceptible to renegotiation by the merchant on the basis of the form of payment they present. The merchant may not accept their credit card or may raise the price based on their type or brand of credit card.
11 Church Report, para.52. 31
33 Public
Merchants ... have an incentive to free ride on the investments made by the card network and merchants that abide by the Rules (or their equivalent). The merchants who engage in hold up benefit from the increase in demand from appearing to accept all forms of payment and certainty regarding the price, but are able to increase their profits by switching to a less costly form of payment to them or adding a surcharge.
90. I interpret Dr. Church's theory to mean that merchants will advertise prices without a surcharge, attract consumers who think that the prices advertised can be obtained with any method of transaction (including any credit card) and then surprise consumers by either: (a) switching consumers to a less costly form of payment; or (b) adding a surcharge. 91. If a merchant surcharges a premium credit card based on cost (e.g., stating "You can buy this product for $100 with your standard credit card or $101.50 with your premium card.") they would be providing consumers with the option to obtain the additional service of transacting with a premium card at cost. This cannot in any sense be described as "bait-and-switch", "hold-up" or "free-riding". 92. Charging an additional premium for a service (using a credit card) that the consumer believed was being provided for free (at the time of making the decision of where to shop) could in a theoretical economic model be described as a hold-up. The consumer, having invested in the costs of traveling to the store, is vulnerable to the store's decision to change the expected price upwards because the costs of shopping are sunk and cannot be recovered. But as a matter of reality, this theory is far-fetched. It is inconsistent with the economics and normal functioning of retail markets. 93. In the real world, retailers who attempted to hold-up and annoy consumers for the sake of an extra I or 1.5 percent of price would gain a reputation as unsavory places to shop - losing the 32
34 Public
entire markup (often more than 40 percent) of those customers deterred by the reputation. Although Dr. Church appears to believe that this is a valid option for merchants, he does not provide any explanation as to why merchants do not engage in such "hold-up" strategies today. For example, we do not see retail merchants extracting additional revenues by informing customers arriving at the store that they, in fact, will have to pay an extra fee for parking. Indeed, parking at retail stores, tends to be low cost - not unexpectedly high. 94. Moreover, the experts present no evidence whatsoever that this type of market failure -hold-up, free-riding or bait-and-switch - is a problem in countries where surcharges are allowed under the law. In Australia, where surcharges have been allowed since 2003, 64 percent of the 2,294 merchants surveyed by East & Partners in December 2010 had either adopted surcharges or were planning to apply surcharges. Furthermore, the proportion of merchants surcharging in Australia has been increasing steadily over time; and in some sectors such as hotels and transportation, surcharging is even more common. This is evidence suggestive of the potential frequency of surcharging in Canada if the Commissioner's application is granted and surcharging is allowed. In fact, surcharging is likely to be more frequent in Canada because the average merchant fee is likely to be several times the average merchant fee in Australia: the interchange fee has been regulated in Australia since 2003 to not more than a weighted average of 60 basis points. As surcharging becomes a more common practice, consumers cannot continue to be surprised, whenever they pull out a credit card to pay for a transaction, by the requirement to pay for using a credit card. The "bait and switch" argument of the Respondents' experts is simply not plausible. 95. Dr. Church has not developed his efficiency theory for the particular business practice at issue in this case. Nor has he presented a single piece of evidence in support of the theory. Indeed, he appears to emphasize that hold-up is a particular concern for "merchants that do not expect
33
35 Public
repeat business" (para. 53). Dr. Church offers no evidence that merchants who expect no referrals or repeat business are even a significant percentage of the retail sector. Dr. Church offers no substantiation to demonstrate why this theory should apply to credit card surcharging. Even in sectors where one might expect limited repeat business, such as travel and tourism, the rise of popular rating websites and referral services (such as TripAdvisor) has drastically limited any private benefit to merchants from holding-up customers. Also, Dr. Church fails to provide any explanation 1JS to why his concerns cannot be addressed by s:imply requiring merchants to properly disclose surcharges to customers. 96. The lack of theoretical development or evidence supporting Dr. Church's efficiency hypothesis is a serious concern with his conclusions. He has offered some comments criticizing the theories of the Merchant Rules as influencing prices upwards and adversely affecting competition, but I have outlined in this reply why these comments are without merit. Competition law cases generally come down to theory and evidence of efficiency versus anticompetitive explanations of a business practice. Dr. Church's report leaves the Tribunal with little or no guidance in terms of evidence for an efficiency explanation of the Merchant Rules.
Date: April 23, 2012 RALPH A. WINTER
34
36 Public
Appendix "A" 1. An example is helpful to illustrate the mechanism by which the cost of an increase in the Acquirer Fee can be focused on cash and debit consumers. In the paragraphs below, I review and extend the example introduced in paragraphs 95 through 102 of my first report. 2. Consider first the impact in a four-party credit card network of a 10 basis point increase in the Interchange Fee when merchants are free to surcharge and all consumers use credit cards. Note that the impact of the Interchange Fee on the Acquirer/Merchant side of the market can be offset by a discount to cardholders on the Issuer/Cardholder side of the market. Acquirers pass on the 10 basis point increase in the Interchange Fee to Merchants in the form of a 10-basis point increase in the Merchant Service Fee. Merchants, in turn, pass on the increased Fee in the form of a 10 basis point increase in surcharges on credit cards. On the Issuer/Cardholder side, the 10 basis points increase in the Interchange Fee is passed on to Cardholders as a 10 basis point increase in the discount on their credit card account. The change in the Interchange Fee has no impact in this simple example of a "perfect" two-sided credit card market. Figure 4A from my first report, which illustrates this effect, is reproduced below for simplicity. 3. Consider what happens when, as in the real world, the market includes some customers paying with cash or other methods of payment. For example, suppose that half of the customers pay with cash, there is a no-surcharge rule in place and (for simplicity) there is only one credit card company. In this example, the 10 basis point increase in the Merchant Service Fee is passed on in the form of a 5 basis point increase in merchant prices to consumers. (Figure 4B, reproduced below.) Cardholders are 5 basis points "better off', assuming that the Interchange Fee increase translates directly to benefits for cardholders.
35
37 Public
4. As noted in my first report at paragraph 102, the 5 basis point benefit need not be left with the cardholders, but instead can be transferred to the credit card company, such as Visa or MasterCard. To make this point explicit, consider a simultaneous increase of 10 basis points in the Interchange Fee and 10 basis points increase in the Acquirer network fee. (I illustrate this change in a third figure below, which I label as Figure 4C.) The total Acquirer Fee increases by 20 basis points, which is passed on fully to merchants. Merchants, under the No-Surcharge Rule, pass on the increased Merchant Service Fee with an increase in the price of products by 10 basis points. Cardholders, facing a 10 basis point increase in product prices and a 10 basis point increase in their monthly discount are on net unaffected by the fee increase. Cash and debit consumers bear the entire cost of the fee increase. The net result of the combined increase in the Interchange Fee and the Network Fee is a transfer from cash and debit consumers to the credit card company. Figure 4A: Impact of a 10 basis point increase in the Interchange Fee (without cash or debit customers)
interchange fee Issuers IE + 10 bp ! + 10 bp discount Cardholders retail prices I + 10 bp 36
Acquirers t + 10 bp in merchant fee I .. !·. ~ Merchants
38 Public
Figure 4B: Impact of a 10 basis point increase in the Interchange Fee (with 50% cash or debit customers)
interchange fee Issuers jc + 10 bp ! + 10 bp discount retail prices Cardholders + 5 bp Figure 4C: Impact of a 10 basis point increase in the Interchange Fee and in the Acquirer Fee (with 50% cash or debit customers)
Credit Card Company / le interchange fee Issuers + 10 bp ! + 10 bp discount retail prices Cardholders I + 10 bp 5. The point of the example is not to present a fully realistic description of the impact of changes in fees in a four-party credit card network. The two-sided nature of the market is "perfect" in the example but not in the real world. The point of the example is that accounting for the two-sided nature of the market actually strengthens the cost-externalization effect described in my report by allowing the credit card company to focus the cost of its price increase entirely on cash
37
Acquirers I l + 10 bp in ~' merchant fee Merchants
"bp; n netwa<k fee Acquirers I l + 20 bp in merchant fee ).: Merchants I
39 Public
and debit customers, instead of sharing the cost between credit card users and users of cash and debit. 6. This example takes to the extreme the assumption that Drs. Church and Elzinga insist upon: that prices of both sides of the market must be considered in an assessment of the competitive impact of the Merchant Rules. If reality is less extreme, such as where discounts to cardholders only partially offset the increase in Interchange Fees, the cost-externalization effect is still strengthened by the two-sided nature of the market, but not to the same degree.
38
40 Public
Appendix "B" - SOURCES AND DOCUMENTS RELIED UPON IN REPORT Jeffrey Church and Roger Ware. Industrial Organization: A Strategic Approach (2000), McGraw-Hill.
Commerce Commission v. Cards NZ Limited et al., Third Amended Statement of Claim, 2 September 2009 (High Court ofNew Zealand) [MBJA0001_00000468].
Examination of Michael Bradley held on December 5, 2011.
Joshua R. Floum, Jonathan Gleklen and Timothy Muris. "Visa's No-Surcharge and Non-Discrimination Rules", May 5, 2008 [GSSS3806_00023400].
Levitin, Adam J. "Priceless? The Economic Cost of Credit Card Merchant Restraints", 55 UCLA Law Review 1321 (2008).
Stigler, George. "Price and Non-Price Competition," Journal of Political Economy, Vol. 76, No. 1 (Jan. - Feb., 1968), pp. 149-154.
39
IN THE MATTER OF the Competition Act, R.S.C. 1985, c. C-34, as amended;
AND IN THE MATTER OF an application by the Commissioner of Competition pursuant to section 76 of the Competition Act;
AND IN THE MATTER OF certain agreements or arrangements implemented or enforced by Visa Canada Corporation and MasterCard International Incorporated.
BETWEEN: THE COMMISSIONER OF COMPETITION Applicant - and-
MASTERCARD INTERNATIONAL INCORPORATED
Respondents -and-THE TORONTO-DOMINION BANK THE CANADIAN BANKERS ASSOCIATION
Intervenors
REPLY EXPERT REPORT OF RALPH A. WINTER
Davies Ward Phillips & Vineberg LLP Suite 4400, 1 First Canadian Place Toronto, Ontario M5X lBl
Kent E. Thomson (LSUC #24264J) Adam Fanaki (LSUC #38208L) Davit D. Akman (LSUC #44274R) Tel: 416.863.0900/Fax: 416.863.0871
Department of Justice Canada Competition Bureau Legal Services Place du Portage, Phase I 50 Victoria Street, 22nd Floor Gatineau QC KIA OC9
William Miller (LSUC #14443V) Tel: 819.953.3903/Fax: 819.953.9267
Counsel to the Commissioner of Competition
Public CT-2010-010 THE COMPETITION TRIBUNAL
VISA CANADA CORPORATION and
APRIL 23, 2012