The economic evidence
670 Much of the attack on the RBA in respect of the economic issues related to the concept of "competition". Dr Pleatsikas, who was called by Visa, considered that an economist would analyse competition, in the context of a market, by referring to the structure of the market and the level of entry barriers. He considered that competition was a fundamental economic concept which necessarily called for consideration of the relevant market. He considered that the relevant market is to be defined with reference to the products that impose a constraint on a firm from raising the price of goods or services above a benchmark level for a significant period of time. In his view, these products comprise the economic substitutes which must to be considered. He observed that the task of defining relevant markets and identifying the products was an appropriate step in assessing how regulatory changes might affect actual competitive conditions. He considered it was an essential task. In his view the minimum economic requirements for having regard to efficiency and competition required a methodology that included five basic analytical elements which all have to be satisfied. These elements do not require perfect information, but they do require sufficient empirical research to provide confidence that competition and efficiency would be promoted by the regulations being proposed. He agreed that there was no economic consensus on these matters. He asserted that the competition assessment framework which he considered necessary for identifying and qualifying the effects on competition was non-controversial among economists. This process involves identifying the products of interest, defining relevant markets, assessing market structure, assessing market power and then assessing how regulatory changes would affect actual and potential competitive conditions. Dr Pleatsikas then discussed the detailed manner in which this methodology should be implemented
671 In applying the suggested five minimum economic concepts, Dr Pleatsikas considers that it is essential to undertake a market definition and also a cost-benefit analysis using empirical evidence. In my view, it is not established that these economic requirements are generally accepted by economists or a majority of them. Nor did Dr Pleatsikas' evidence establish that such a process is essential to an adequate analysis of competition and efficiency. He fairly acknowledged that there is scope for disagreement among economists concerning the importance of the competition and efficiency effects of specific credit card issues identified by the RBA and he accepted that some economists may agree with the Bank's view about competition and efficiency consequences. He accepted that there was no one single method by which all economists would measure the concepts of competition and efficiency in relation to credit card schemes. His five economic requirements are extracted from his own experience and are not specifically identified in any authorities as a cumulative group of essential requirements. He accepted that there is no established body of literature relating to the field of credit card regulation which claims that there is a requirement for the development of specific economic assessment methods to estimate impacts.
672 Dr Pleatsikas agreed in cross-examination that an economist could reach a view that regulation was beneficial in relation to credit cards without carrying out the exercise he suggested in order to determine whether the regulations would promote competition and efficiency. Specifically, he said that some prominent economists have reached a view that interchange fees should be regulated or prohibited totally. In practice, he accepted that economists have reached firm conclusions about the regulation of credit card networks without carrying out the minimum requirements. Indeed, Dr Frankel, a respected economist engaged in economic consulting with Dr Pleatsikas, had formed views concerning the undesirability of a no surcharge rule as a matter of economics without carrying out the empirical analysis which Dr Pleatsikas referred to. Dr Pleatsikas did not doubt that Dr Frankel believed that the conclusions he advocated, in relation to the no surcharge rule and interchange fees, would be pro-competitive. He also acknowledged that a number of economists have come to views contrary to those held by Dr Frankel in equally definitive terms. His position was that it would be necessary to have regard to his five minimum economic requirements to "settle" the question. However, that is not the exercise posed by the statutory language. What the Act refers to is the formation of an opinion by the PSB. It must be acknowledged that the Cruikshank Report (Competition in UK Banking - A Report to the Chancellor of the Exchequer, 20 March 2000) produced in the United Kingdom suggested that regulating interchange fees could improve efficiency and competition, but Dr Pleatsikas considered that the Cruikshank Report was framed at a high level of generality. He agreed that the RBA had material from responsible economists concerning substitutability and that there was economic material which argued that there would not be an undue shift from four-party to three party credit cards arising from the regulation of interchange fees. In particular, he agreed that economist from both sides of the argument had not identified or performed the work necessary to satisfy the minimum economic requirements identified by him in order to reach their conclusions. In effect, Dr Pleatsikas agreed that Visa, MasterCard and the other parties who made submissions to the RBA did not go through the process which he said is essential to make an informed judgment. The economists making these submissions did not feel constrained to adopt his requirement. Dr Pleatiskas did not formulate any specific methodology for assessing the effect of regulation and he acknowledged that the RBA, on the material he had seen, was aware of and adverted to the lack of empirical evidence. In acknowledging the lack of consensus among economists in relation to his suggested minimum economic requirements the following exchange occurred in cross-examination:
"Q. Doctor, what I want to suggest is that no-one, apart from you, appears to have any difficulty in coming to conclusions in this area without applying your minimum economic principles.
A. Oh, I mean the old joke is, if you ask nine economists a question, you will get 10 opinions. I have no doubt that you can get opinions from economists on a wide variety of issues, even if they haven't studied the issue thoroughly."
673 A similar acknowledgement as to the lack of unanimity between economists in many areas was expressed by Professor Katz, who was called by the RBA. One common thread that emerged from the economic expert evidence in this case is that, at least in the area of credit card regulation of the type involved in this case, there is a lack of conclusive data, methodology and of consensus among economists on many of the central issues raised by the applicants.
674 An important premise for Dr Pleastsikas was that market definition was an essential part of carrying out the minimum economic requirements. However, he agreed that market definition was not universally necessary and that views differed as to when and how it should be carried out. He accepted that in some instances, it may be sufficient for the purpose of defining a market to have regard to products which are substitutable for the product which is being regulated, without delineating the outer bounds of the market. He agreed that in some cases one would only need to perform a limited market analysis. He also accepted that in any particular analysis of a specific market there may be a wide range of views.
675 Dr Pleatsikas said that the RBA had erred in considering only transaction costs in its measurement of efficiency, but in cross-examination agreed with the position that while the RBA focused on costs, it took into account benefits. He agreed that on his reading of the material, the RBA reviewed submissions and concluded that the benefits merchants received in general did not correspond to the costs being imposed on them. He also accepted that the RBA took into account some of the costs and benefits of its reforms, but said that it had not measured and determined other benefits and costs. He did not suggest that experimental models of the nature he proposed were essential to determine efficiency. Nor did he point to any specific model.
676 A major difficulty in Dr Pleatsikas' evidence was that he did not identify any specific method for taking into account other costs and benefits which he contended should be taken into account but he suggested that data should be collected. He could not assist specifically as to how that could be done. He suggested that there be experimental methods, but accepted that none were suggested by those economist who provided reports during the consultation process undertaken by the RBA. He could not specify an appropriate experiment to be carried out and was not aware that anyone had actually designed any reliable experiment of the type he referred to in the context of credit card regulation. He was unaware whether anyone had been asked to do it or had been commissioned to do it.
677 In the absence of any detailed identification or analysis by Dr Pleatsikas of the costs and the effort required to collect data that he suggested should be gathered, it is difficult to give great weight to his proposal. A major difficulty with his evidence was that he was unable to be specific as to how data should be collected, what particular data should be collected and how it should be analysed.
678 He considered that empirical estimates of cross-elasticities of demand need not be precise, but need to be more than simply qualitative. He also said that qualitative assessments may be sufficient in some circumstances to determine whether efficiency has been improved if the results are relatively unambiguous. He agreed that economists would have differing views on the extent to which empirical analyses would be required and accepted that on the material that he has seen, the RBA had reached a conclusion notwithstanding the absence of empirical analysis.
679 By way of empirical analysis, he suggested that it was appropriate to investigate substitution elasticities by collecting data, conducting a survey and conducting an economic experiment. However, he agreed that in relation to credit cards, such empirical techniques had not produced reliable results. He referred to attempts by MasterCard to obtain an empirical analysis in the form of a survey conducted by PricewaterhouseCoopers but in his view this was problematic. It was regarded by one economic expert, Dr Glaser, as being methodologically unsound and flawed. So far as he was aware, the only survey which had been done in relation to credit cards concerned merchant attitudes with respect to surcharging in Sweden, the United Kingdom and the Netherlands and he agreed that those surveys were referred to in the Consultation Documents. He asserted that the questionnaires could be reliably used to predict behaviour, but acknowledged that there was considerable difficulty in survey design and that is not a subject in respect of which he could claim expertise. He had seen no empirical analysis of the nature he suggested could be conducted by Visa or MasterCard apart from the PricewaterhouseCoopers analysis.
680 At one point he took what appears to me to be a surprising position by suggesting that it may be better to rely on an unreliable technique, rather than no specified technique at all. But eventually he agreed with the proposition that economists must acknowledge that available estimate of benefits and costs can be so uncertain that they are unlikely to enlighten.
681 Dr Pleatiskas could not give details of any specific empirical analysis of competition between credit cards networks based on innovation and he had not seen any empirical analysis dealing with the effect on competition of innovation in relation to credit cards. He accepted that the RBA had available information from surveys on consumer and merchant preferences and that the Australian Retailers' Association was a body which could be approached for the purpose of obtaining information about merchant preferences. He proffered no basis for doubting that the RBA analysed and considered the information it received from Visa and MasterCard, as well as issuing and acquiring banks, to the extent they were capable or providing useful information.
682 In substance, Dr Pleatsikas' evidence reinforced the conclusion that there is considerable room for differences of opinion on many central issues concerning the regulation of credit cards among economists and that none of the empirical studies he suggested as desirable had been identified or carried out in this area. Nor had any been funded or formulated of which he was aware.
683 In my view, Dr Pleatsikas' evidence does not establish or substantially assist to establish the conclusion that there is an obligation on the RBA to follow the methodology which he suggested or to gather the empirical data which he indicated ought be obtained. It falls far short of establishing any error in failing to perform exercises or gather possible data which might be useful or reasonably appropriate to make the determinations under challenge, particularly having regard to the practice which economists have adopted in formulating views with respect to the regulation of credit card systems.
684 MasterCard called two expert witnesses Professor von Weizsacker and Dr Veljanovski. Professor von Weizsacker's evidence can be summarised as follows.
685 Professor von Weizsacker was concerned with an examination of competition and efficiency and the need for formal market definition together with the likely impact of the no surcharge rule and the need for interchange fees. He had impressive academic qualifications and experience as a regulator over many years in the field of economic regulation.
686 In cross-examination he accepted that price competition among payment systems was not a matter he had investigated, particularly in any Australian market for payment systems and he had no empirical basis for his assertions. Specifically, he accepted that he, himself had carried out no empirical analysis about the question of whether a cost-based interchange fee would lead to a possibility of four-party credit card systems ceasing to be viable. He did not recall any submissions made by issuing or acquiring banks or other participants in the credit card systems as expressing a view that the cost-based interchange fee would threaten the viability of the system. Underlying his evidence was a view that there was likelihood of entry into the credit card market by retailers acting alone or collectively, although he had not investigated the likelihood of this in Australia.
687 One central difficulty with adopting the evidence of Professor von Weizsacker was that he was committed to a particular, idiosyncratic view that favoured a relatively unique "replacement test" for defining a market. He agreed that market definition was a tool and viewed the credit card market as being evolutionary. In Australia this conclusion is open to doubt. His preferred replacement test has not been accepted among economists generally. He believes that if an economist came to a conclusion that the market in credit cards is not evolutionary but has matured or stabilised, a different measure to that in his replacement test might be appropriate. The replacement test he suggests involves an extremely broad field of analysis including matters such as tourism and developments overseas. He said that his replacement test was not a standard test anywhere. Additionally, he was of the view that if issues are complicated, experts can disagree and one could expect disagreement among experienced and eminent experts.
688 In relation to the no surcharge rule, he considered that removing the rule might be pro-competitive. His criticism of the Interchange Standard was that there was sufficient pressure on merchant service fees and in his view, as a matter of policy, there was no need to regulate interchange fees.
689 In my view, taken overall, the evidence of Professor von Weizsacker clarified to some extent the discussion of the concepts of competition and efficiency, but his views were by no means conclusive in relation to the matters under consideration and did not substantially assist the applicants' case.
690 MasterCard also called Dr Veljanovski. It became evident during his cross-examination that his experience and expertise was limited to a consideration of the concepts of competition and efficiency and the need for formal market definition. The evidence of Dr Veljanovski was largely theoretical. He had not carried out the work which he suggested ought to have been done. Nor could he express a view whether such work had ever been attempted anywhere in the world. In particular, with respect to the claimed failure of the RBA to measure total surplus in relation to its reforms, despite his assertions, it was not part of his brief to investigate the task, nor review the literature on the matter.
691 In relation to his assertion that the RBA had simply dismissed the relevance of network effects and failed to subject them to serious economic or factual investigation, he accepted that the RBA was aware of the existence of the potential justification of the interchange fees. He did not have regard to the submissions made to the RBA and made no attempt to determine whether it would be possible to obtain empirical information. Nor did he conduct any empirical analysis. He was aware that Professor Katz had prepared a report canvassing and analysing the various economic issues, but did not rely on it because he was not instructed to do so. He took a relatively narrow view of his instructions and fell back on this narrow view as a basis for not taking into account a number of important matters which were put to him in cross-examination. He retreated to this position on several occasions under cross-examination. There are significant deficiencies in his evidence.
692 Although he did not consider whether the report of Professor Katz, who was called by the RBA, amounted to a serious economic investigation concerning the interchange fee, Dr Veljanovski had not conducted any inquiry which might assist him in forming a definitive view that the RBA had failed to examine the interchange fee.
693 A large part of his criticism of the RBA process was based on the approach that market definition, as is commonly carried out under Part IV of the TPA, was essential to an analysis of competition. However, he agreed in cross-examination that market definition was an analytical tool and he accepted that it was appropriate to analyse competitive effects. He accepted that it is not essential in all cases for a market definition analysis to be conducted when seeking to ascertain, from an economic viewpoint, whether a particular action will be pro-competitive. Under cross-examination, he agreed that he had expressed the view in a published article that in relation to the relevance of markets in European Community competition law, economists do not generally pay much attention to the definition of a market as most advanced text books would readily show. Rather he considered that those texts usually proceed directly to discuss and identify market power. His view was that market definition was only essential in regulatory and competition proceedings. In summary, this aspect of his claims was based on the assertion that the approach in the present case should be analogous to proceedings for breach of the TPA in relation to the analysis of competition law.
694 He rejected the competition benchmarks suggested by the RBA as being necessary or sufficient for the public interest, but in cross-examination accepted that they were related to questions of competition or efficiency. He agreed, in particular, that an important feature of a competitive market was that consumers would have sufficient information to enable them to make an informed choice and that if prices are distorted or not clear, consumers could make wrong consumption decisions. That is to say he accepted the desirability of transparency. He did not dispute that merchants should be free to set prices for customers that promote the competitiveness of their business. He was unaware of network effects in Australia in relation to credit cards and did not know whether network effects existed, but spoke in terms of a substantial or high degree of probability that they do exist. He agreed that prices charged by financial institutions to consumers who use payment instruments should reflect the relative costs of providing those instruments, as well as demand conditions and agreed with the proposition that less restrictive barriers to entry are more likely to produce a competitive market in the absence of any supervening factor. He accepted that any restrictions on the entry of institutions to a payment system should be the minimum necessary.
695 In my view, the evidence of Dr Veljanovski does not materially advance the applicants' case and, whether taken either alone or cumulatively with the evidence of the other experts, his evidence does not support a conclusion that, from an economic point of view, the RBA had fallen into reviewable administrative error or had failed to perform its functions under the legislation.
696 The RBA called two expert economists to give evidence in relation to the questions of competition and efficiency and other economic considerations.
697 The first was Professor Katz. His evidence was to the following effect.
698 He agreed that the word "efficiency" has a particular meaning for economists and that in his opinion, the expert witnesses for the applicants expressed the somewhat narrow view that "efficiency" was to be measured by application of the concept of total surplus. This concept was frequently, but not always, used in economic modelling and empirical studies generally. He agreed that the term "total surplus" meant the sum of consumer surplus and producer surplus. By consumer surplus, economists mean the difference between what a consumer is willing to pay for a good or service and what he or she actually pays. Producer surplus refers to the amount of income a producer receives in excess of what it would require in order to supply a given number of units of goods or service. In this way, total surplus can be regarded as a consumption benefit measured in dollars minus the cost of production. Such an approach requires that costs and benefits be taken into account.
699 Professor Katz also referred to the concept of Pareto efficiency in discussion pointing to the concept of efficiency as central to economic theory. Pareto efficiency in particular, was an optimal situation and was an important part of modern theory. This concept was considered in the material obtained or produced by the RBA.
700 Professor Katz is an eminent economist with advanced academic qualifications and considerable practical experience. He had substantial experience of credit card systems. His specialty is in the economics of industrial organisations, which includes the study of antitrust and regulatory policies. He has co-written a standard text on Microeconomics. He is recognised as one of the pioneers in extending the theory of network effects to competitive settings. His practical experience consists of consulting on the application of economic analysis to the issues of the antitrust and regulatory policy and he has served as the consultant to the US Department of Justice and the Federal Communications Commission in relation to antitrust and regulatory policy in telecommunications, especially with respect to network effects. From January 1994 through to January 1996, he was the chief economist of the Federal Communications Commission and from September 2001 to January 2003, he was Deputy Assistant Attorney-General for Economic Analysis at the US Department of Justice. He has written numerous articles and has carried out extensive research into network effects. After reviewing relevant documentation and reading reports of the experts of the applicants, he reached the following conclusions.
701 Professor Katz considered that the expert material raised a number of questions about the magnitude of economic effects. He has not identified any errors in the economic logic that underlies the RBA's central conclusion that industry practices may lead to the use of credit cards even in instances where it would be more efficient to use alternative payment mechanisms. He does not consider that the economic evidence demonstrates that the policies proposed to be implemented would harm competition and efficiency. He saw the experts as raising numerous objections to the process which the RBA used to analyse and choose among policy options, but he does not agree that the alleged flaws in the process make it difficult, or impossible, to predict that the measures will, on balance, significantly promote competition and efficiency.
702 Professor Katz does not consider that formal delineation of the relevant market is essential to making economic projections regarding the competitive and efficiency effects of policy options. Nor does he consider that the policy analysis undertaken by the RBA is not economically valid because it does not develop explicit numerical estimates of certain economic parameters and projections of competitive and efficiency effects. He disagrees with the proposition that relevant issues have not been raised or considered by the RBA as part of its policy design and assessment process. He refers to the RBA's extensive consultations.
703 In relation to the Surcharge Standard, Professor Katz considers that there is general agreement among economic experts and the RBA that the extent of surcharging is likely to be limited and that the Standard includes a mechanism to limit the potential for merchants to set inefficiently high surcharges. With respect to the Access Regime, he says that in his view, the applicants' experts do not identify actual problems with policy towards access to the schemes, but simply raised concerns about problems that could possibly arise under what they consider to be potential interpretations of the Regime.
704 With respect to the Interchange Standard, Professor Katz considers that the applicants' economists have used an unrealistic and idealised method of maximising the economic efficiency as measured by total surplus which makes use of complete and perfect information about the consumer and merchant benefits derived from the use of payment mechanisms.
705 Professor Katz prepared a commissioned report for the RBA in relation to the economics of credit cards and prior to preparing his expert report in these proceedings he prepared a number of statements in proceedings brought by the United States Department of Justice against Visa and MasterCard. He was involved in consultations with the RBA during the consultation and formulation process of the measures under challenge in these proceedings.
706 Professor Katz's evidence is attacked on the basis that he was partisan to the case of the RBA because of his prior consultancy work and, in particular, his involvement in the action brought by the United States Department of Justice and that the whole of his evidence should be rejected. There is no substance in this submission.
707 The applicants point to the role of Professor Katz in assisting the RBA in relation to his instructions and his performance of systematic economic analysis and examination of the broader issues identified by the RBA. They refer to the empirical work which he considered in his report to the Department of Justice in the United States proceedings and also refer to an empirical analysis to which he had regard in preparing that report. There was a suggestion of indirect economic interest. Specific reference is made to steps taken by Professor Katz in considering data with respect to industry personnel, consumer preferences, merchant acceptance, long term trends in the use of payment instruments and data on the structure of credit card and charge card industry in relation to market power. Professor Katz gave some evidence as to when it is appropriate for governments to act or regulate payment instruments.
708 I found Professor Katz to be a frank and open witness who made concessions where it seemed appropriate in cross-examination and I could see nothing in his evidence which indicated a pre-commitment, prejudgment, bias or entrenchment in a particular position as distinct from having relatively firm views in relation to some issues, based on his own opinions and substantial experience. Having regard to the content of his evidence and his responses under cross-examination, I am not satisfied that the submission that his evidence should be rejected in the basis of bias or unreliability has been made out.
709 Professor Katz indicated that he would rely on the tools of economic decision theory that governments should only act when they believe that the expected benefits are greater than the expected cost. In his view, the effect of economic decision-making theory is that the government does not require definite evidence, but it should have the expectation that benefits are greater than cost. Professor Katz expressed what he said is a widely held view among American economists that regulation costs may often exceed the costs of the problems the regulation is trying to address. Visa raised the fact that he expressed an opinion that government intervention is usually only warranted in situations of market power and that in the absence of market power there is limited rationale for government intervention. Professor Katz was of the opinion that in the context of regulation, the question is not whether the pricing methodology is optimal, but whether that method is likely to lead to better bargaining results in the absence of guidelines. He also expressed some concern with regulation and its effects.
710 Professor Katz points out that total surplus is not the only appropriate measure of efficiency. He expressed the view of the use of total surplus as a measure of efficiency assumed that the distribution of income is socially optimal. The applicants pointed to the fact that in preparing his commissioned report and his expert report to the Court, Professor Katz examined economic effects that can go beyond competition and efficiency. For instance, his commissioned report dealt with the distributional concern that non-credit card users would be paying for credit cards.
711 Professor Katz considered that economists recognised the fact that market definition is not necessary or sufficient to understand the competitive effect of various policy questions.
712 Professor Katz referred to the concept of market power as being the market share of a participant in the market and the number of substitutes that exist for a firm's products. He agreed that there was debate among economists in relation to credit cards concerning the identification of externalities in credit card networks and whether the credit card market is competitive or whether certain participants were able to earn excessive amounts. He said there was also debate as to whether the no surcharge rule or its abolition is welfare enhancing and as to the purpose, effect and magnitude of interchange fees.
713 In relation to interchange fees, he agreed that it was not possible to conclude definitively that interchange fees are harmful based on the fact that literature in the area is ambiguous and any firm conclusion requires important assumptions. He also agreed that at the time of the Joint Study, there was only a small body of economic literature on the role of interchange fees in payment systems and that the literature on the economics of credit card associations was limited.
714 Professor Katz accepted that the RBA needed to obtain information to assess the efficiency and competition impact of its proposals and to base its analysis on facts so far as possible. He also considered that it was important to take into account the costs of obtaining information in relation to the benefits that might accrue from such information. Professor Katz was not aware whether the RBA had attempted to quantify the impacts of its regulations and could not recall seeing any quantification of the costs in consumer benefits associated with the use of credit cards, or the regulatory costs, including the administrative and compliance cost.
715 Professor Katz considered that the methodology of the RBA was based on the conclusion that competition between three and four-party schemes would force the three party schemes to lower their interchange rates, but did not evaluate any underlying studies. He could only point to limited data to indicate that the RBA had made any attempt to quantify the impact of its reforms.
716 In relation to the no surcharge rule, Professor Katz accepted that the market outcomes may not be fully optimal, although he had concluded that economic theory provided reasons to believe surcharges can serve as an important means of internalising network externalities, but the market outcome may not be fully optimal in the presence of frictionless surcharging.
717 Professor Katz accepted that studies in some countries had shown that there was relatively little impact from the abolition of the no surcharge rule, but pointed out that the RBA thought that one of the reasons for such a lack of impact was the interchange fee. He accepted that there could be indirect impacts from the abolition of the no surcharge rule. In his commissioned report, he recognised that the no surcharge rule may have no effect and may not matter and that this should be considered in coming to a conclusion whether it ought to be prohibited. He agreed that one of the RBA's concerns was that non-cardholders would pay higher prices because merchants would be passing to them the costs of credit cards, but he did not disagree that if the merchants were to charge differential prices for different payment instruments the costs of administration would be exorbitant.
718 Professor Katz agreed that the overall benefits were an important factor in the setting of interchange fees and concluded in his commissioned report that in some circumstances, it is optimal to use interchange fees to rebalance the costs and benefits enjoyed by the two sides of a card-based transaction.
719 Both in Professor Katz's commissioned report and in his statement filed in the proceedings, he agreed that rebate expenditures are economic costs from the issuers' perspective and that they should be taken into account in determining issuing margins. He also agreed that loyalty programmes should be included when calculating the profit or margin of the issuer. This view has an impact on the calculations made by the RBA, although the RBA had a different view as to the way loyalty points should be treated. It did not consider they were a resource cost and therefore decided not to include them in the definition of eligible costs.
720 Professor Katz considered that if the result of the regulation was a shift from credit cards to charge cards, there may be some negative efficiency effects. He considered that in the United States, one of the closest economic substitutes to credit cards were charge cards and agreed that in Australia charge cards are probably the closest economic substitutes for credit cards, but he did not carry out the relevant study to enable him to finally commit to that as an expert. He was of the opinion that the analysis offered by some experts was incomplete. Because of his specific experience and speciality in network economics I found his evidence of greater assistance than that of the experts called for the applicants.
721 Professor Farrell was also called by the RBA to give expert evidence on the issues of efficiency and competition. He is Professor of Economics at the University of California at Berkley and an Affiliate Professor of Business and chair of the Competition Policy Centre. His view was that no rigidly defined methodology is necessary for an economist to undertake a proper economic consideration of the concepts of efficiency and competition.
722 I found Professor Farrell to be a frank, open and measured witness and given his practical experience and excellent economic qualifications I found his evidence to be most helpful.
723 Professor Farrell's academic qualifications were of the highest order (indeed as were the qualifications of other expert witnesses), and he had considerable practical experience and expertise in econometrics and this was apparent from his evidence. He adopted a careful, measured approach to questioning and made appropriate concessions where necessary. He was assiduous to ascertain the assumptions behind questions and did not simply respond off the cuff. He was not defensive or argumentative. In instances of conflict with other witnesses I prefer his evidence. Although, he agreed with some matters put to him in cross-examination, he remained firm in his opinions regarding central issues such as the absence of a need for market definition and the absence of a requirement for a formally structured methodology in the context of credit card regulation. He also adhered to his view with respect to the absence of any compelling data to establish the applicants' case and also the practical desirability or perceived necessity of taking action without ascertaining all possible data which might be gathered. He emphasised the range of differences of opinion between economists on the relevant issues. His approach struck me as realistic and well-reasoned and his conclusions were reached after careful thought. There was no suggestion of bias towards the RBA.
724 Professor Farrell's view was that no specific, defined methodology was mandated as a matter of economic principle and in particular, he saw no basis for requiring adherence to the five element methodology favoured by Dr Pleatsikas or the defined process of Dr Veljanovski. He considered it likely that no two witnesses would be in complete agreement on any particular methodology as being essential. He also considered that the economic issues raised by the rules and practices of Visa and MasterCard were complex and "subtle". He pointed out that economists routinely analyse difficult issues relating to competition and efficiency using whatever evidence and methodologies are appropriate to the task at hand and considered that there was no obvious reliable way to collect some of the quantitative information that was claimed to be essential, such as elasticities with respect to interchange fees and pass-through rates.
725 Professor Farrell adhered to his view that in order to make the challenged determinations, it was not essential to define a market because he considered market power was the critical concept, which according to Professor Farrell, is the ability of a firm or groups to harm consumer welfare and economic efficiency through its actions. In his view, a firm lacks market power in circumstances where, if it fails to offer consumers a fully competitive level of benefits, consumers will readily turn elsewhere for such benefits. He considers that an assessment of market power is often used as a proxy to reduce the necessity for investigations of firms' behaviour in cases, for example, of a proposed merger. In his view, when assessing whether market power exists, an important consideration is the elasticity of demand in respect of the firm's products or services. This is a measure of the strength with which demand responds when price or some other aspect of the product changes. In a case of high demand elasticity, consumers will go elsewhere at the onset of a price increase. Despite extensive cross-examination, Professor Farrell remained firm in the view that it was not necessary as a matter of economics for the present exercise to delineate the parameters of the market for payment services. His view was that a precise market definition is not warranted because there is, in his view, a continuum in which some products are relatively close substitutes, contributing a good deal of demand elasticity and others which are relatively distant substitutes, contributing only a little. He pointed out that the relationship between market power and market share is most straightforward in markets for homogenous goods such as gold or particular types of wheat, for example, in which competitors' strategies largely amount to decisions relating to the level of production. This can make market definition relatively simple. However, it is technically more complex and more intuitive to consider differentiated product competition where the products are imperfect substitutes for each other. In his view, there are imperfect links between measures of market share and market power and it is common in competition policy to apply tentative presumptions or rules of thumb based on market share which requires market definition. However, good economic practice requires an awareness that, in some cases, these presumptions are only a pointer where fuller examination is more or less likely to define market power. In his view, Professor von Weizsacker's discussion of the replacement test suggests that a departure from the standard market definition approach may, in some instances, be appropriate and the fact that Professor von Weizsacker makes the argument is seen to cast doubt on the claims of Dr Pleatsikas and Dr Veljanovski that conventional market definition is essential.
726 Professor Farrell also points out the complexity of market analysis when he says that even a firm with a small market share may be able to exercise significant unilateral market power in certain industries. He refers to the difficulties which arise if one tries to define a market in the context of the present case. These include the issue of the strongly overlapping membership and governance of four-party schemes in Australia and whether those schemes are to be treated as competitors in defining a market and identifying competitors in the market. Conventional market definition demands "yes and no" answers and these issues illustrate the difficulty of doing this.
727 Professor Farrell refers to the contention that interchange fees serve what may be described as a balancing role. He agrees that in certain circumstances they can fulfil both this role and the closely related function of improving incentives for card holding and card use. However, he considered that the discussion and analysis of this issue by Professor von Weizsacker was incomplete because he did not properly address the harm to non-credit card customers from paying interchange fees since they may ultimately share the cost as a consequence of general price increases to all purchasers.
728 In his view, no surcharge rules lead to economically inefficient systems in two ways. First, when a merchant has reason to charge different prices for different payment methods but does not, the common price he will choose is likely to be a compromise price, with the result that the no surcharge rule has the effect of raising the price to the group that would otherwise have faced the lower price. The second inefficiency to which he points that may stem from the no surcharge rule is a socially unacceptable incentive for customers to use credit cards as the means of payment. He points out that Professor von Weizsacker recognises the argument that card schemes may increase prices to cash customers and reduce prices to card customers, but Professor von Weizsacker does not investigate this question.
729 Professor Farrell considers that the adverse effect on non-card customers of merchants as a consequence of the no surcharge rule is a negative externality. He considers, but does not affirmatively assert, or conclude, that prohibition of the no surcharge rule can address harm to non-credit card users and thus produce some possible efficiency improvements. In relation to interchange fees, he explains why limiting interchange fees may address the harm to non-credit card users if prohibiting no surcharge rules does not and goes on to explain why interchange fees set by Visa and MasterCard are unlikely to be socially optimal.
730 Professor Farrell rejects the notion that formal empirical studies should be carried out. The general proposition that it is better to know more rather than less may be accepted, but questions arise as to the purpose for which a specific empirical investigation project is needed and whether it is feasible and whether the results are likely to be trustworthy. In his opinion, Dr Pleatsikas has given no reason to expect that any study could convincingly disentangle, statistically or econometrically, the effects of changes on the four-party scheme interchange fees from other variables. Professor Farrell is concerned as an expert in econometrics with the difficulty of avoiding statistical bias. He considers that such a study would have little precision and would not be useful. He agrees that there is a compelling reason to study cross-elasticities with respect to interchange fees, but points out that no good opportunity is evident and none is suggested by Dr Pleatsikas. In this case, to insist that the RBA should not act without such a study amounts to insisting that the RBA should not act at all. In his view, the economics of choice and taking action under conditions of uncertainty compels no such result. He also takes a similar view with respect to Dr Veljanovski's suggestion that the RBA should have investigated the translation of merchant service fees into retail prices. In his view it would be desirable to know this. He disagrees with the suggestion by Dr Pleatsikas that the RBA should have studied the sources of consumer and merchant heterogeneity with respect to the use of payment mechanisms. He considers that this lacks any compelling reason. In his view, it is not optimal for regulators to rely on an approach which implies that they should not act if they lack information that would be nearly impossible to obtain.
731 In the view of Professor Farrell, a fundamental insight of microeconomics is that when incentives are misaligned, economic inefficiencies are likely to result and economists can conclude that outcomes will be inefficient and can infer which incentives should be adjusted even without explicit cost-benefit analysis or a structured empirical investigation along the lines suggested by the expert witnesses for the applicants.
732 In cross-examination Professor Farrell agreed that usually, one needs to consider substitutable products when dealing with issues of competition. That is to say other products that in ordinary usage and, potentially in technical usage, could be in the same market. What requires a great deal of information is quantifying the substitutability. Thus to make an accurate prediction of price impact of a merger you would need data about substitutable products and about the degree of substitutability.
733 Professor Farrell considers there is good reason to think that the no surcharge rule, while it may not in the usual sense increase or decrease competition, could have an effect of distorting competition and causing firms to compete not on overall value, but on some other criteria such as the differential price they offer to card holders. There is no need for an exact understanding of who are the other players or products in the market to reach that conclusion. The access regime is essentially a matter of intervening to create more competition and that would be much more likely to be desirable if there was not adequate competition from other payment schemes.
734 According to Professor Farrell, one of the matters the RBA ought to have considered was whether changing the price signals for credit cards required change in the price signals for other payment instruments, including charge cards. It may be helpful to have an idea of the quantum of shift away from credit cards to debit cards if that was the purpose, but he added that perhaps this is not necessary.
735 Professor Farrell agreed that, in theory, there was level of interchange fee which could be included in the price for all customers which would not harm non-credit card customers, namely, where level of interchange was equal or proximate to the level of merchant transaction benefits. An empirical study would need to be conducted to determine whether merchant service fees exceed convenience benefits to merchants. He did not agree however, that it must follow that non-card users would not be harmed so long as interchange fees do not exceed convenience benefits to merchants.
736 Professor Farrell's evidence is that it is very difficult to ascertain the level of the merchant transaction benefits to perform this comparison. It will vary considerably across merchants. Benefit in economics is not necessarily determined by the cost.
737 Professor Farrell said that these elements are fundamentally different. He agreed that if the merchant service fee does not exceed or is equal to the merchant convenience benefit relative to the alternative payment mechanism, there will not be a pass-through to customers because there is no additional cost. He repeated that there were problems with obtaining an accurate estimate having regard to the variation among merchants, making it extremely difficulty to gauge. That was not an approach taken by the RBA because of its difficulty. The position as to the merchant benefit is a fairly complex question and it cannot be simplified.
738 The import of the discussion by Professor Farrell is that if the regulation seeks to effect a move from credit cards to debit cars or charge cards, then it would be useful to have a quantitative figure. There is no evidence that any attempt has ever been made to make such a quantitative assessment and none was pointed to by the applicants in the hearing. The applicants simply relied on the qualitative assertion that if a hypothetical balance were achieved there would be no need to pass through additional cost. In my view, this submission of the applicants does not advance the applicants' case in relation to the no surcharge rule or the interchange fee.