Note: Where available, the PDF/Word icon below is provided to view the complete and fully formatted document
Competition aspects of e commerce.



Download PDFDownload PDF

1

30th Annual Conference of Economists

23-26 September 2001 Perth

Competition Aspects of E Commerce

Dr David Cousins Commissioner

Australian Competition and Consumer Commission

2

Introduction *

The development of e-commerce offers a variety of competitive opportunities for business. High technology and Internet companies have not yet delivered the returns that investors once hoped for. However, it is likely that in the longer term e-commerce will significantly change the way in which firms do business, as an increasing number of transactions migrate from the physical to the online world.

From a macro perspective, it is estimated that widespread adoption of e-commerce in Australia could result in a 2.7 per cent increase in national output, and enhance consumption by about $10 billion within the next decade.1 Studies also indicate that in addition to the anticipated increase in competition from international or domestic online sources, e-commerce will also enable businesses to achieve higher levels of efficiency in day to day operations. All of these factors should facilitate higher levels of productivity and consumption.

However, as has become apparent, the development of e-commerce is not without its difficulties. There are a whole range of issues still to be addressed including consumer confidence, allocation of domain name rights, recognition of digital signatures, treatment of intellectual property rights, development of financial payment systems, and application of competition laws. Due to the borderless nature of the Internet, these issues need to be tackled simultaneously on a global and domestic basis.

From a Competition Regulator’s perspective, the New Economy raises exactly the same issues as the offline world, but in new contexts.

Nevertheless, in dealing with online commerce Regulators will be increasingly confronted with some of the more controversial issues in competition law.

Of particular importance in this area is the competitive analysis of network effects. As in other areas, Regulators will need to assess the conduct of competitors - both online and offline - to determine whether the threat of competition may lead to anti-competitive practices. As network effects or externalities may be a key characteristic of many e-commerce activities, it is important to recognise the positive and negative competitive aspects of network effects and how they may be taken into account within the framework of the Trade Practices Act.

A Regulator’s ability to assess competitive conduct issues in relation to an activity characterised by network effects, particularly in New Economy markets, has attracted considerable debate.

This paper identifies and assesses three main concerns raised:

* I would like to acknowledge Ms Vanessa Holliday’s major contribution to this paper. Vanessa is an Assistant Director in the ACCC’s E-commerce Unit, Compliance Branch.

1 NOIE, E-commerce Beyond 2000, Final Report p xi.

3

1. It is sometimes argued that competition regulators over exaggerate the potential market power issues arising in relation to network effects. This, it is submitted by some commentators may be detrimental and stifle pro-competitive ventures.2

2. It is sometimes argued that in New Economy markets, the potential threat of new entry is particularly strong because technology is changing rapidly, therefore, any market power that may arise due to network effects will be short lived.3

3. It is also argued that competition regulators confuse network owners competitive conduct in offering lower prices and otherwise encourage access to, and participation in their networks with predatory behaviour.4

This paper suggests that such concerns are over stated. Network effects raise both pro and anti-competitive considerations which need to be balanced. This will be an increasingly challenging area, as new types of networks form, but the dangers of simply assuming that a network will be pro-competitive are just as great as the dangers of stifling development of new networks in e-commerce.

What is a network effect?

In order to assess the competitive implications of network effects in the New Economy, it is important to identify briefly the economic concept of a network, and a network effect.

A network is a series of connected nodes in which the pieces of the network are complementary to each other.5 Traditional examples of networks include telephone networks, railway lines and gas pipelines. These traditional networks are largely physical networks that involve a direct, physical connection between users. The key

2 Dr Cento Veljanovski, EC Antitrust & the New Economy, Is the EC Commission’s View of the Network Economy Right? European Competition Law Review, 2001 Vol 9 writes: “The application of network effects theory to new economy mergers is overblown and lacks supporting evidence.”

3 Richard Posner, Antitrust in the New Economy, Antitrust Law Journal, [Vol 68, 2001] pp 925 - 943. Posner takes the view that the law is “supple” enough to deal with the New Economy, but that the rapid changes in technology make it difficult to enforce. See also D Teece and M Coleman, The meaning of monopoly: antitrust analysis in high-technology industries, The Antitrust Bulletin/Fall-Winter 1998 pp 801 - 857 at 804 states “ Competition in high technology industries is fierce, frequently characterised by incremental innovation, punctuated by major paradigm shifts. These shifts frequently cause incumbent’s positions to be completely overturned.”

4 See N Economides, Competition and Vertical Integration, Paper delivered at conference, “Competition, Convergence and the Microsoft Monopoly: The Future of the Digital Marketplace, organised by the Progress and Freedom Foundation, Washington DC, Feb 4, 1998 at p 7 which suggests commercial rather than competition reasons for Microsoft’s entry into the Internet browser market. See also D Teece and M Coleman, op cit., p 839 which discusses reasons for adopting high or low prices more generally in high technology industries.

5 N Economides, The Economics of Networks, International Journal of Industrial Organization, Vol 14, No 2 (March 1996).

4

characteristic of a physical network is that the value of the network is dependent on the number of people who use it or the nodes it connects to.6 The more people who are connected to a telephone network, the more other people will wish to join. Conversely, a telephone network that has few connections is unlikely to attract many customers and is likely to fail.

In this sense, networks may be said to exhibit demandside economies of scale.

Many products and services which, although they are not physical networks in the same sense as a telephone service, do exhibit some of the same characteristics as physical networks. Therefore much of the theory developed around physical networks, particularly the implications of demandside economies of scale can be applied to these industries.

This is particularly true in the Internet industry and e-commerce. As Shapiro and Varian state, “there is a central difference between the old and new economies: the old industrial economy was driven by economies of scale; the new information economy is driven by the economies of networks”.7 Similarly, the former Chairman of the Federal Trade Commission, Robert Pitofsky makes the point that an essential feature of the New Economy is increased dependence on products and services which are pieces of intellectual property such as computer software, faxes, and Internet services. Such products and services more frequently exhibit network effects as the demand for the product is related to its use by others.8

Where some New Economy networks differ somewhat from traditional networks is that they tend to be “virtual” in nature rather than physical. Some of these high technology networks are in fact physical networks - for example networks of ATM machines and other financial electronic payments systems, and the Internet itself. However, others such as a network of users of computers are virtual networks where users/nodes are not physically linked, but form a network of complementary products. For example, in the case of Microsoft, the operating systems market was said to exhibit network effects because as the number of users of the Windows operating system increased, the more valuable it became as it attracted more software applications. This is sometimes referred to as an “indirect” network effect.9

6 For more detailed explanation of network effects see J Tirole, The Theory of Industrial Organisation, MIT Press, 1989. Also see G Werden, Network Effects and Conditions of Entry: Lessons from the Microsoft Case, Antitrust Law Journal, 69(1) 2001 p 89; C Shapiro and H Varian, Information Rules, Harvard Business Press

7 C Shapiro and H Varian, op cit., p 173.

8 Robert Pitofsky, Antitrust and Intellectual Property: Unresolved Issues at the Heart of the New Economy, speech delivered at the Antitrust, Technology and Intellectual Property Conference, March 2, 2001, available at http://www.ftc.gov/speeches/pitofsky/ipf301.htm.

9 Michael L Katz and Carl Shapiro, Network Externalities, Competition and Compatibility, The American Economic Review, June 1985, Vol 75 no 3, pp 424 - 440 at 424.

5

Such networks tend to be based on intellectual property rights and know how, rather than physical assets. They are more likely to develop through collaborations between independent bodies than Old Economy networks which, at least in the Australian context originally derived from statutory monopolies. Further, because they are virtual rather than physical in nature, there is more scope for competition between networks to arise, whereas physical networks were largely restricted to the geographic area of their physical networks.

What are the competitive issues surrounding network effects?

Network effects may involve both competitive and anti-competitive elements.

One school of thought suggests that competition regulators should be concerned about network effects because network industries are likely to exhibit high barriers to entry.10 It is difficult for a new entrant to attract customers because unless they can connect the majority of customers all at once, they are unlikely to succeed. Where infrastructure investment costs are high and supply side economies of scale are also present, the difficulties faced by potential new entrants are exacerbated.11

Where network effects are considered to be particularly strong, then not only will barriers to entry be high, but it is argued that the market may tend towards a “winner take all” solution where one network becomes dominant and the others fail due to positive feedback. That is, the more that a particular network is used, the more attractive it becomes to existing and new users. This is sometimes referred to as “tipping” or a “snowballing” effect.12 Because the law of diminishing returns may not apply to the same extent as it does in relation to industries with supply side economies of scale, a dominant player may emerge, rather than the traditional oligopolistic outcomes more often observed in traditional industries.13

Where market power is established, there are three main categories of behaviour that may raise competition issues.

First, there may be concerns that an incumbent will exercise market power in the form of higher charges for connection and usage of its network.

Second, exclusionary behaviour by the incumbent including refusal to provide access to users or potential competitors or providing access on a discriminatory basis may raise

10 See Daniel L Rubinfeld, Antitrust Enforcement in Dynamic Network Industries, The Antitrust Bulletin, Fall-Winter 1998 p 862.

11 C Shapiro and H Varian, op cit. p 179.

12 Note however that this is not always the case. It may be that consumer preferences will support different products or standards in a particular area. See Rubinfeld, op cit., who gives the example of co-existence of Sega, Nintendo and Sony.

13 C Shapiro and H Varian, op cit. p 180 states that “unlike the supply-side economies of scale, demand-side economies of scale don’t dissipate when the market gets large enough: if everyone else uses Microsoft Word, that’s even more reason for you to use it too.”

6

competition issues. Where the incumbent is vertically integrated into upstream or downstream markets to the network, it may have the incentive to refuse access to its competitors in this area. This may have a detrimental impact on competition in those markets. Further, refusal to provide access or interconnection services to a competing network may stifle the development of efficient, innovative services, within the network market itself. 14

Thirdly, inclusive behaviour by the incumbent including the use of exclusivity clauses in access arrangements, predatory pricing or other incentives provided to encourage exclusivity or customer loyalty, and collaborative arrangements with potential competitors which stifle their ability to fully compete against the incumbent network owner may also raise competition issues. 15

Such conduct may result in competitive damage in the market for the network itself, and in particular increase barriers to entry.

Proponents of this school of thought may be particularly concerned about actions that promote exclusivity and increase switching costs as it may actually delay or frustrate the emergence of a new product or technology to threaten an incumbent’s position. For example, one commentator has argued that by requiring users to enter into exclusivity contracts or membership rules, an incumbent may successfully deter the entry of technically superior networks - which may be of particular concern to the long term development of e-commerce. It is argued that where network externalities are present and the value of the network to users is related not only to its own use, but the use of others, that the cost of switching is higher than in traditional markets.16 Another example is the development of non-compatible complementary products which may increase barriers to entry, as a new entrant will need to duplicate a greater number of complementary products in order to compete.

However, where networks are open access, non-exclusive and operating rules are objective and non-discriminatory, competitive issues are less likely to arise.

The alternate view is that networks offer obvious economic efficiencies and social benefits which will outweigh these potential concerns.

14 C Shapiro, Exclusivity in Network Industries, Geo.Mason L Rev. [Vol 7:3 1999] at p 10. See also Daniel L Rubinfeld, op cit. At 862-3 which suggests that a firm may have an incentive to prevent products of rivals achieving compatibility, which may restrict competition even when other products are of a comparable or even higher standard.

15 Albert A. Foer, E-Commerce Meets Antitrust: A Primer, Journal of Public Policy & Marketing, vol 20(1) Spring 2001, 51 - 63; see also D Balto & R Pitofsky, Antitrust and high-tech industries: the new challenge, The Antitrust Bulletin/Fall-winter 1998 p 595 states: “antitrust analysis should proceed very cautiously before requiring networks to open themselves to potential competitors. Not only may compulsory access unfairly penalise a firm or group of firms that has achieved success on the merits, but it may also reduce the incentives of would-be challengers to build a better network.”

16 See C Shapiro, Exclusivity in Network Industries, Geo.Mason L Rev. [Vol 7:3 1999] pp 1 - 11.

7

Some commentators argue that the battle between firms to “win” a particular market is just as intense as normal competition17 and because of this, it is unlikely that networks will attempt to exercise market power by increasing prices.

It is also argued that negative feedback can result in the loss of a market as easily as positive feedback enabled the capture of a market. On that basis, it is argued that regulators should not fear that incumbents will act in an anti-competitive manner in the future, once their network is established because if they were to do so, they would lose the market to a new entrant.18

In this regard, the impact of consumer perception is considered to be extremely important, particularly in rapidly developing high technology markets. Some commentators argue that because network effects are largely driven by consumer expectation of what other consumers will do, it is important for an incumbent network to keep prices low. Otherwise, it is argued consumers will expect that others will leave the network, particularly in high technology areas where there is already likely to be some expectation that a new technology will emerge to replace the existing network and in effect, this becomes a self fulfilling prophecy.19

It is argued that networks achieve significant efficiencies for users which may outweigh the potential risk of competitive detriment.20 Also, they may fulfil a wider social function when they emerge in relation to the development of new products or services -for example, a B2B electronic marketplace or a new computer operating system - in that it may enable whole new areas of opportunity for commercial activities to grow. An example of this is the development of the MP3 peer-to-peer file swapping service where the service has generated interest in the development of other peer-to-peer products, including, it is reported, a TV show swapping service to be released by Sony. Such systems can alter market structures quite drastically in the long term.

17 N Economides, The impact of the Internet on Financial Markets, Journal of Financial Transformation, Vol 1 no 1, 2001 pp 8 - 13.

18 McKenzie & Lee, How Digital Economies Revise Antitrust Thinking, The Antitrust Bulletin, Summer 2001 pp 253 - 298; Anton & Yao, Standard Setting Consortia, Antitrust, and High Technology Industries, Antitrust Law Journal, Vol 64 [1995] at 258; C Veljanovski, Trade Practices Law in the Network Economy, paper delivered at Trade Practices Committee Workshop, August 2001 at 13.

19 McKenzie & Lee, op cit. p 270.

20 See D Teece and M Coleman, op cit., at p 814 - 5. It is suggested that the benefits of having one platform may be greater than the costs from less diverse platforms. Benefits may include access to a greater range of complementary products and less cost involved in upgrading. See also Januz Ordover and Robert Willig, Antitrust for High-Technology Industries: Assessing Research Joint Ventures and Mergers, Journal of Law & Economics, Vol XXVIII(2), May 1988 at p 312 which suggests that as it is uncertain whether concentration assists or restricts technological advancement, the application of anti-trust laws to restrict concentration may cause harm.

8

Proponents of this school of thought suggest that Regulators are not well placed to distinguish between anti-competitive and pro-competitive behaviour by networks and are more likely to cause harm than good if they try as this may stifle development.

They argue that concerns about perceived high prices for network access are misguided, as they fail to take into account that prices may be set at levels necessary to fund investment to efficiently capture the positive externalities associated with networks.21 It is in fact suggested that caution should be taken in setting prices too low, as this may encourage free riding and discourage future network investment.22

Further, it is argued that exclusionary conduct of an anti-competitive nature is unlikely to occur due to the threat of future competition, and the effects of inclusive conduct is likely to be pro-competitive as it encourages use of the network.

In summary, it appears that in terms of competitive analysis, network effects like supply side economies of scale are a two edged sword. On the one hand, they may provide significant efficiencies. On the other, networks tend toward a dominant position, which may or may not be of a transitory nature which may put them in a position to exercise market power in a number of ways - in particular, to stifle the potential development of competing networks.

Application of the Trade Practices Act to Products and Services Exhibiting Network Effects

The Trade Practices Act (Cth) 1974 (‘the Act’) is the primary piece of legislation which regulates competitive conduct within Australia, and provides the framework for consideration of the competitive implications of network effects in both Old and New Economy markets.

The Act does not prohibit products or services which exhibit network effects as such, or prevent a network growing to a position of dominance due to the popularity of the network. However, the Act may apply to:

1. Conduct of an incumbent network owner that is likely to have a substantial anti-competitive effect or purpose; or

2. Collaborations, mergers or joint ventures between network owners, or network owners and owners of complementary products which are likely to have a substantial anti-competitive effect or purpose.

The Australian legislative framework provides for a range of different regulatory responses to such issues, depending on the nature of the goods or services involved and the nature of the conduct. The main regulatory tools are Part 111A of the Act (Access

21 See J. Gans & S. King, The Role of Interchange Fees in Credit Card Associations: Competitive Analysis and Regulatory Issues, April 2001.

22 See C Veljanovski, op cit., at p 25.

9

declarations and undertakings), industry specific access regimes, and Part IV of the Act (anti-competitive conduct rules including mergers and acquisitions).

Specific Regulation of Essential Facilities

The Act provides that networks which exhibit monopoly characteristics may be dealt with pursuant to Part IIIA (access regulation) or an industry-specific access regime.

An access regime approach is used mainly to address concerns that a monopoly network operator may refuse access to users (ie exclusionary behaviour). It is generally applicable where the underlying network consists of a facility which is uneconomic and therefore undesirable to duplicate, but there is potential for a variety of competitive services to be offered at other functional levels, or part of the network to be offered by a competitor. Therefore, access regulation is used to facilitate competition at those other functional levels.

In some cases, price controls are also used in order to protect end product users from monopolistic practices until such time that effective competition can be introduced.23

By using this form of regulation, such networks are encouraged to develop in order to enable the efficiencies and public benefits of the network to be achieved, but also provide safeguards against anti-competitive behaviour.

Part IIIA - Access Regulation

Certain types of networks, such as rail or road networks may be regulated as essential facility services pursuant to Part IIIA of the Act. Part IIIA provides for the declaration of services provided by means of a facility by the National Competition Council in limited circumstances.24

The main criteria include25:

1. Whether it would be uneconomic for anyone to develop another facility to provide the service; 2. Whether access or increased access would promote competition in at least one market in Australia, other than the market for the service; 3. Whether the facility is of national significance, having regard to the size of the

facility or importance of the facility to trade or commerce; 4. Whether access can be provided without undue risk to health and safety; and

23 Similarly, the Prices Surveillance Act 1983 may also be relevant in monitoring the price of goods and services.

24 Alternately, owners of a service may submit an access undertaking to the ACCC for approval (s 44ZZA). Also, where there is an effective state or territory-based access regime in place in relation to a facility, a service cannot be declared but is regulated under the framework of the relevant state or territory regime (s 44N).

25 Trade Practices Act s 44H

10

5. Whether providing access would be contrary to public interest.

Where a service is declared under Part IIIA or services are subject to an access undertaking, a party seeking access to the service who is unable to reach agreement on the terms and conditions of access may apply to the ACCC for arbitration. This provides for a “light handed” approach where parties seeking access must first attempt commercial negotiations before seeking regulatory intervention.

Whilst not expressly stated, the test of whether a facility is uneconomic to duplicate involves consideration of network effects as well as supplyside economies of scale. For example, in the area of airport regulation, the ACCC has considered that the concept of economic duplication involves consideration of not only the supplyside economies of scale associated with airport infrastructure investment (which is not to say that these are not significant in this area), but also the complementary nature of airport facilities.26 That is, the value of an airport facility also depends on the number of users, routes, and connecting services that utilise the facility. The greater the number of routes, the more likely it is that the airport will attract more connecting flights and therefore more users.

Nevertheless, not all products or services which exhibit network effects will fall within the scope of Part IIIA. First, whilst the term “facility” is not expressly defined within the Act, it has been suggested that it means a physical asset or set of assets which exhibit features of a natural monopoly.27 Where, as in the case of some high technology virtual networks, it is the intellectual property and network effects which may be major cost, rather than duplication of the physical assets as such, it could be debateable whether the “facility” is uneconomic to duplicate. Secondly, it will be a question of fact in each case whether the facility is of national significance.

Further, not all competition issues arising in relation to networks will necessarily fall within the scope of access regulation. Part IIIA does not apply unless it would promote competition in some other market - usually an upstream or downstream market. It does not necessarily provide a solution when the owner of an essential service exercises market power where other markets are not effected. Also, it does not apply in assessing collaborative arrangements or mergers between networks.

Industry Specific Access Regulation

In addition to the general access provisions contained in Part IIIA, there are a number of industry-specific access regimes which are used to regulate access to physical networks including telecommunications, gas, electricity and airport facilities. In most cases, industry-specific regulation has been introduced in relation to former statutory monopoly utility networks.

26 See ACCC, Section 192 of the Airport Act - Declaration of Airport Services Draft Guide, October 1998.

27 See Re Sydney International Airport [2000] ATPR 41-754.

11

The current industry-specific regulatory framework for the telecommunications industry was introduced in 1997. The current regulatory approach is based on the recognition of the high barriers to entry into telecommunications markets due to large sunk costs, the legacy of a historic statutory monopoly and network effects arising from the desire for any-to-any connectivity which reduce the likelihood of new entry.28 These factors are aggravated by consumer switching costs and vertical integration by the incumbent, Telstra into downstream markets.

Part XIC of the TPA allows the ACCC to declare certain telecommunications services. Where services are declared, the access provider must make access available to other service providers and take all reasonable steps to ensure that technical and operational quality is equal to the quality of service enjoyed by the access provider.29 The access provider must, on request provide interconnection. The ACCC has the role of arbitrating access disputes between parties. This may involve terms and conditions of access and interconnection, as well as pricing issues.

Unlike the general access regulation model contained in Part IIIA, the regulatory framework also provides for price controls over certain retail services including local calls and line rental charges. The objective of these measures is to protect the best interests of the community during the transition from a monopoly to a competitive market.30 In addition, competition laws in this area have been strengthened by the introduction of Part XIB.

A similar approach has been taken in relation to the airports, gas and electricity industries. In recognition of the natural monopoly characteristics of airports, a regulatory scheme has been established involving a set of price controls over core aeronautical services (and monitoring of other prices), and an access regime to encourage competition.31 Similarly, an access regime and revenue regulation has been established for electricity transmission and distribution networks.32

Networks regulated under general anti-competitive conduct rules (Part IV)

Where products and services exhibit network effects, but are not necessarily considered to be facilities which fall within the scope of Part IIIA or industry-specific access or

28 Productivity Commission, Telecommunications Competition Regulation, Draft Report, March 2001 p 26.

29 The criteria for declaring a telecommunications service are somewhat different to the criteria for declaring an essential service pursuant to Part IIIA of the Act. S 152AL(3) provides that the ACCC must be satisfied that the making of a declaration will promote the long term interests of end-users.

30 The ACCC recently reviewed current price control arrangements. See ACCC, ACCC Review of Price Control Arrangements, February 2001.

31 See Airports Act 1996 (Cth).

32 See National Electricity Code (NEC). Also note that the regulatory responsibility for the electricity transmission and distribution networks is divided between the Commonwealth (with the ACCC as regulator) and the States.

12

price regulation, they may still fall within the scope of the anti-competitive conduct provisions of the Act (Part IV).

Conduct that may fall within the scope of Part IV includes:

Allegations of Misuse of Market Power (s 46)

Section 46 of the TPA prohibits a person who has substantial market power from using its market power for a proscribed purpose. Proscribed purposes include eliminating or substantially damaging a competitor, preventing entry into a market, or preventing or deterring a person engaging in competitive conduct.

In the context of network industries, this may involve a range of issues including refusals to provide access to a potential user, or discrimination against certain types of users; refusals to allow for interconnection with competitors; predatory behaviour; and restrictions on the ability of users to deal with competitors which may restrict the development of competing networks33. That is, both inclusive and exclusionary conduct may fall within the scope of s 46.

Mergers and other forms of Collaboration between competing networks and potential competitors (s 45 and 50)

Section 45 of the TPA prohibits contracts or other arrangements which are likely to result in a substantial lessening of competition in a market. This includes arrangements between competitors that are likely to result in a fixing, controlling or maintaining of price which are deemed to substantially lessen competition pursuant to s 45A. For example, the setting of interchange fee agreements which are likely to result in the fixing, maintaining or controlling of prices may raise issues under s 45 and s 45A of the Act.

Also, where members of a network agree to exclude other persons from using the network this may also raise issues under s 45.

Mergers and acquisitions of networks that are likely to result in a substantial lessening of competition in a market are prohibited under s 50.

Exemptions

There are two kinds of “exemptions” which are of particular relevance in network industries.

First, where conduct may otherwise breach Part IV (other than s 46), the parties may seek authorisation of the conduct which will provide protection against legal proceedings under the Act. The ACCC will authorise the conduct if it is satisfied that

33 This may also raise issues of exclusive dealing pursuant to s 47 of the TPA which prohibits refusal to deal, or dealing on condition that the acquirer of goods or services does not deal with the supplier’s competitors.

13

there are public benefits that outweigh the anticompetitive detriment, or, in the case of a merger application, the proposed acquisition would result in such a benefit to the public that it should be allowed.34 This is the primary provision which currently enables the ACCC to take into consideration public benefit arguments, including efficiencies in assessing collaborations to form a network, or the conduct of an existing network.

Second, s 51(3) of the Act exempts the imposing of or giving effect to a licence or assignment of a patent, registered design, copyright or trade marks from the operation of Part IV - unless the conduct falls within s 46 (misuse of market power), or s 48 (retail price maintenance). This may restrict the application of the Act in cases where a network uses its control over intellectual property rights to create or exert market power (unless it can be shown that it has engaged in a misuse of market power).

Summary

Under current Australian laws, networks, particularly physical networks which have developed out of statutory monopolies, may be subject to access regulation. This will cover situations of “exclusions” which may stifle the development of competition of related markets, and to a degree enable some horizontal competition to an incumbent by providing for the development of interconnection agreements.

Virtual networks emerging the New Economy appear more likely to raise issues under general competition laws, as they are not necessarily characterised as facilities which are uneconomic to duplicate. Also, general competition laws are more flexible in addressing issues of “inclusive” conduct aimed at stifling the development of competitors.

Recent Cases

Australian Financial Payments Systems

Financial payment systems are an example of an area where competition issues have emerged within the context of a network environment.

Network effects are a characteristic of payment systems. The value of an ATM network increases as more ATM tellers are introduced, because this encourages more customers to use them, which in turn, encourages the development of more ATM tellers. Equally, as financial institutions issue more credit cards, merchants benefit from obtaining access to a wider range of customers and will be more willing to participate in credit card schemes which in turn will increase the attractiveness of credit cards to customers.

In a purely theoretical setting it would be expected that the development of such networks would benefit society. However, in 1996, the Wallis Report35 found

34 Section 90.

35 Financial System Inquiry Report (1997).

14

significant market power issues arising in payment networks in Australia. Key issues identified included:

1. Interchange fees charged for credit card use at the wholesale level were being passed on by Merchants in the form of higher prices for goods and services. As the fees were not transparent, consumers were not responding by moving to other payment forms. In fact, this may limit the development of new payment systems.

2. The relative bargaining power between major card acquirers and regional banks was uneven, so regional banks had difficulty gaining access to networks as acquirers.

3. The rules of international credit card associations were not transparent and could limit membership to the existing range of financial institutions.

Similarly, in 1997 the ACCC raised concerns in the course of considering an authorisation application from the Australian Payments Clearing Association (APCA) in relation to proposed rules for the Consumer Electronic Clearing System. It had concerns that inequality of bargaining power between member institutions in ATM and debit card payment networks in negotiating interchange fees would place some at a competitive disadvantage.

In September 2000, the ACCC and the Reserve Bank of Australia released a joint study on Debit and Credit Card Schemes in Australia36. It was found that interchange fees for ATM services were double the average cost of providing services, and credit card interchange fees were set significantly above cost. No surcharge rules in credit card schemes prevented price transparency and ensured that other consumers subsidised the cost of credit card payments. Credit card schemes were found to limit access, excluding all institutions other than deposit-takers.

During the same period, the ACCC also took action against one bank, on the basis that in jointly setting interchange fees for credit cards it was likely to be in breach s 45A of the Trade Practices Act.

The results of the study and the s 45A action suggest that networks may raise market power issues. First, the membership restrictions imposed by the credit card schemes illustrates that not all networks will want to provide access to everyone even though it may benefit the network - they may want participation at the retail level, but not at the wholesale level because this may endanger their individual market positions37. Second, this case shows that not all networks can be relied on to charge low prices to encourage usage. That is, while direct prices to consumers for use of credit card facilities appeared to be low, in fact, high prices were being charged to merchants and subsequently

36 Reserve Bank of Australia & ACCC, Debit and Credit Card Schemes in Australia, A Study of Interchange Fees and Access, October 2000.

37 A common complaint raised by firms seeking access to a network is not that they are not allowed to use the network, but that they are limited to being retail customers. They are not offered access at wholesale prices.

15

passed onto consumers in the cost of products. Third, whilst on the one hand, interoperability between member financial institutions enabled them to compete against each other, it also provided opportunities for collusive price fixing and via interchange fees restricted the level of competition between them.

While some arguments have been made that inter-change fees are a necessary and efficient mechanism in developing credit card services38, it is questionable whether such high fees were necessary once the network was established.

A longer term consideration which emerged from analysis of the financial payments system was that the above conduct not only imposed higher costs on the community as a result of the exercise of market power, but also restricted the development of competitive payment instruments. As the cost of credit cards to consumers was subsidised through higher charges to merchants, pricing signals regarding the cost of credit facilities was distorted. This again, illustrates one of the key concerns regarding the exercise of market power in the New Economy - that instead of encouraging competition, networks may be able to stifle innovation and efficiency.

As a result of the issues arising in this area and to further public interest, the Reserve Bank of Australia has formally designated credit card schemes in Australia as payment systems subject to its regulation under the Payment Systems (Regulation) Act 1998.39 Under this regime, the Reserve Bank will establish standards for the setting of interchange fees and a regime for access to the credit card system.

The Microsoft Case40

Internationally, competition regulators have looked at the implications of network effects in a number of matters. Probably the most prominent case in this area has been the Microsoft matter, which has generated debate on a wide range of issues and is still not yet finally decided. In that case, the US Department of Justice took action against Microsoft alleging that by a range of actions it had used its market power in the market for Intel-based PC operating systems to stifle competition in that market and the emerging market for Internet browsers. A key element of the Trial Judge’s decision against Microsoft was that it derived the necessary market power to engage in this conduct due to network effects. It was held that Microsoft had a dominant market share in Intel-based PC operating systems, and that market share was protected by the so called “applications barrier to entry”.41 That is, it was held that the market for Intel-based PC operating systems exhibited indirect network effects because the value of Microsoft’s operating system increased as more applications were developed for that system. This made it difficult for any other

38 See J Gans & S King, op cit.

39 See Reserve Bank of Australia, Designation of Credit Card Schemes in Australia, Media Release, 12 April 2001.

40 US v Microsoft Corporation, 87 F. Supp. 2d 30 (DDC 2000) (appeal pending).

41 See Conclusions of Law, p 3.

16

operating system to attract significant consumer demand, thus enabling Microsoft to retain its position.

Conduct which was alleged to have raised issues included the imposition of technical and contractual ties between Microsoft’s Windows operating system and Internet Explorer. It was argued that via this conduct, Microsoft was constraining the ability of other software developers including Netscape from entering the market for Internet browsers, and ultimately, the operating systems market itself.

Accordingly, Microsoft’s conduct may be viewed as predominantly “inclusive” behaviour to restrict the development of competitors.

This case demonstrates the divergence in views regarding the competitive implications of network effects.

Critics of the Microsoft decision suggest that Microsoft’s market power is transitory due to the rapid development of technology and that its conduct has consumer benefits which should be taken into consideration in assessing the case.42 In particular, critics of the decision argue that Microsoft’s behaviour served to increase usage of an innovative new product, Internet Explorer, and provided convenience to users by offering the new product with the Windows operating system. However, it could also be argued that the way in which Microsoft went about tying the two products was beyond what was necessary to achieve this outcome and therefore had the effect of restricting the development of innovative new services.43

It is likely that similar issues may arise in assessing the conduct of a product or service exhibiting network effects in Australia under s 46 of the TPA. If the provider of a network seeks to leverage off its network to introduce new complementary products this may provide efficiencies and social benefits, but at the same time, could be used in order to preserve market power and to stifle competitive innovations.

WorldCom/MCI

The European Commission has also taken in account the competitive implications of network efficiencies in a number of cases. These include WorldCom/MCI44 where the European Commission raised concerns about the merger between two Internet backbone operators. It was concerned that as the size of WorldCom/MCI’s Internet network grew and it became more valuable for customers to have access to it that it would achieve a dominant position due to the forces of network effects. This would place it in a position to exercise market power against potential competitors,

42 See Richard McKenzie & Dwight Lee, Op cit.; Prof N Economides, US v MS and the Future of the US Computing Industry, May 5, 2000 at http:/www.stern.nyu.edu/networks/usvms.html.

43 See R Gilbert and M Katz, An Economists Guide to US v Microsoft, Journal of Economic Perspectives - vol 15 No 2, Spring 2001 at p 35 queries why, if Microsoft’s objective was to increase the value of its personal computer platform, it wouldn’t support well made browsers other than Microsoft itself.

44 MCI/WorldCom, Case IV/M. 1069, OJL 166/1 (1998).

17

particularly in relation to interconnection. Particular concerns included the ability to terminate peering arrangements (effectively treating competing wholesalers as customers) and degrade the quality of links between its network and competitors - causing more customers to shift to WorldCom/MCI.

Again, this decision has been criticised on the basis that it over emphasises the dangers of network effects giving rise to market power issues. In particular, one commentator has argued that the termination of peering arrangements is not necessarily evidence of foreclosure, but is consistent with economic efficiency and the need for a network operator to protect itself from free-riding.45

Summary

The above examples suggest that market power issues may arise in relation to a number of different networks.

Inevitably, in assessing these issues the conduct of network products or services may contain both pro and anti-competitive elements. As is seen in Microsoft, MCI/Worldcom and the treatment of interchange fees in credit card payment systems, the impact of network effects and the merits of each case are hotly debated.

Nevertheless, given that failure to act in such cases poses significant risk of anti-competitive detriment, particularly in matters where it appears that the effect of market power may be to stifle innovation and the development of new products, it would be unwise to dismiss substantive competition issues without careful consideration.

The Future - E Commerce Networks

These themes are likely to re-emerge in relation to new e-commerce networks which are developing, particularly B2B electronic marketplaces. Another area of e-commerce where network effects are relevant is in collaborative processes to develop standards for e-commerce.

B2B Electronic Marketplaces

B2B electronic marketplaces (B2Bs46) are essentially electronic platforms which allow users to trade with other businesses in a standardised manner. Beyond that, marketplaces can differ dramatically, as they may be;

• Open systems, free to any and all users or closed systems which require users to join up as subscribers or members;

45 See Cento Veljanovski, Trade Practices in the Network Economy, Paper Delivered at The Trade Practices Committee Workshop, 17 - 19 August, 2001 p 25 - 26.

46 Although the term B2B is commonly used when discussing B2B electronic marketplaces, technically the term B2B denotes a much broader range of e-commerce activities.

18

• Focusing on a particular product or industry (verticals) or a wide variety of products used in many industries (horizontals); • Providing facilities for catalogue purchases or an actual price mechanism (eg operate a spot market), or both depending on the product; • Operated by independent commercial organisations, a joint venture between

competitors, or an existing industry association; • Operated for profit or non-profit; • Providing actual trading capabilities (ie click on a product and send the message

through to the seller), or catalogue-only capabilities (ie see the sellers catalogue, but complete trade off line); • Providing payment and settlement services or not; • Offer to integrate trading into buyers and sellers internal IT systems, or require rekeying of trading information.

There are numerous examples of B2Bs developing both at the domestic and international level. Domestically, examples include: corProcure and Cyberlynx, horizontal buyer-driven catalogue and reverse auction B2Bs established to trade indirect goods and services; PeCC, a vertical buyer-driven platform for the pharmaceutical industry; Ausmarkets and Yieldbroker, vertical seller-driven trading platforms for wholesale bonds and other OTC products market; and EANnet, an independently owned electronic platform for the grocery industry. Internationally, examples include Covisint, the vertical buyer-driven trading platform for the auto industry, Myaircraft.com, and GlobalNetExchange (the global retail B2B).

A B2B provides users with a one stop shop to trade with each other. As more users join, the B2B becomes increasingly valuable to existing users because they will have access to a larger number of buyers and sellers. This provides a better opportunity to compare price and other terms of trade in order to obtain the best deals. Also, as more users join up to a particular B2B, it becomes more valuable to others because they can deal with more trading partners in a standardised manner.

However, although one stop shopping is essential in certain types of products such as stocks, futures and other financial derivatives which are homogenous products governed by the need for liquidity, liquidity is not necessarily so important in manufacturing inputs. Buyers may prefer to deal with a particular supplier who customises a product for them. They may not really care so much about how many other suppliers are involved in a B2B and accordingly the network effects may not be as strong. Nevertheless, there may still be benefits associated with gaining access to standard electronic trading mechanisms.

Unlike traditional networks, B2Bs are not necessarily based on physical infrastructure costs. Internet infrastructure and browsers enabling connectivity are already in place. Yet these electronic platforms are not necessarily cheap or easy to build up and pull down. In talking with parties trying to establish B2Bs, the ACCC is finding that such ventures are taking months if not years to develop. The generic technology is easy to acquire. What is not so easy is customising it to meet the needs of particular industries. For example, standards may need to be developed for electronic purchasing forms, and to enable businesses to load their products onto electronic catalogues which are compatible with different buyer systems. These may seem trivial considerations, but

19

they can all add up in terms of time and cost. Thus, the infrastructure costs in such ventures may not necessarily be seen as physical assets, but intellectual property rights.

As with other virtual networks, there are significant benefits and efficiencies to be gained from B2Bs. One example of this is in the health industry. A study completed in 200047 analysed the implications of a B2B pilot, Project Electronic Commerce and Communication for HealthCare (“PeCC”), which was developed to facilitate online procurement of pharmaceutical and other supplies to hospitals and retail pharmacies. It found that the potential cost savings were in the vicinity of $340M pa, with the overall cost of an order reducing from $75 to $5. Most of the savings were found to be in very simple things like reducing basic errors in ordering products and improving inventory management. This can have significant flow on effects in terms of cash flow and payment cycles - an issue particularly relevant for small business. Another example which is illustrated in relation to Covisint, the US auto supplies B2B, was that the use of a common electronic platform could actually be used by a buyer and seller to collaborate in cyberspace on a new custom design. For buyers and sellers, there may also be significant gains to be made in identifying new trading partners or participating in spot auctioning of excess inventory.

Standard Setting

In order for e-commerce to develop, new types of standards need to be developed. These may include product numbering in electronic formats, standards for security and networks of Public Key Infrastructure Certification Authorities who need to be able to cross check the validity of digital certificates between trading partners, and standards for ordering forms and electronic catalogues.

Standards is another area where network characteristics emerge, as the more a standard is used, the more valuable it is to other users.

How will Regulators deal with network effects in B2B Electronic Marketplaces?

B2Bs and Market Power

Several recent reports have suggested that in certain circumstances, B2Bs may exhibit network effects that may result in the creation of market power and high barriers to entry in the provision of B2B procurement services.48 As stated above, it will be important to assess whether network effects are in fact likely to be strong in each case.

47 Professor Elizabeth More & Dr Michael McGrath, Health & Industry Collaboration, the PeCC Story, 2000.

48 See United States Federal Trade Commission, Entering the 21st Century: Competition Policy in the World of B2B Electronic Marketplaces, 2000; W Blumenthal, B2B Internet Exchanges: The Antitrust Basics, Antitrust Report, May 2000 p 34 - 55.

20

However, even if network effects do exist this does not necessarily mean that all B2Bs which exhibit network effects are likely to raise market power issues. This will depend on a range of factors including the following.

Market Concentration in Markets for Goods and Services Traded

If the owner or owners of a B2B have market power in the underlying old economy products and services that are traded via the B2B, then it is likely that market power in the underlying wholesale markets would also manifest in the emerging market for procurement services. Where a significant amount of throughput is controlled by the owner or owners, it will be less concerned about competition from new entrants in the B2B sphere, and therefore more likely to abuse market power than if the B2B were independent. Thus, the ACCC has stressed that in assessing B2Bs, one of the key issues is whether the marketplace is participant-owned, and what degree of combined market power those participants wield.49

Substitution Possibilities

Where there are substitutes for using a particular B2B - for example, other B2B’s, other e-procurement services, or traditional off line procurement services - then it is less likely that a B2B will be able to exercise market power. Whether or not such services will be close substitutes will depend on the level of efficiencies delivered by B2Bs, and whether other e-procurement and off line procurement procedures will be capable of providing a competitive service.

Dynamic Characteristics

Dynamic characteristics such as the potential for new technology to leap frog over the old incumbent is a relevant consideration in assessing the extent to which an incumbent may have market power.

In assessing dynamic characteristics, consideration should be given to whether the incumbent believes that it is about to be overtaken and whether this constrains its ability to exercise market power. In each case it will be necessary to look closely at the history of an industry to discover how long it does take for a new entrant to replace an incumbent - either with a new product or a new variation - and whether the incumbent does appear to respond to the threat by keeping prices down.

Again, the underlying market structure is relevant. If owner-participants of a B2B marketplace controls the majority of throughput, the threat of independent innovation is much less threatening.

Also, much innovation stems from building on existing knowledge. In each case, access to knowledge depends on intellectual property rights and the willingness of an incumbent to provide access to a product or standard for the purposes of creating compatible or competing services and products. In some cases, incumbents will make

49 ACCC, B2B e-commerce and the Trade Practices Act, ACCC Update, Issue 8, February 2001.

21

that knowledge available, which may enhance the possibility of dynamic change in the industry. However, if they believe that they have sufficient market power, they may close or restrict access to knowledge in order to retard the development of new technology - or rather, new technology that is not controlled by them.50 In many cases, such activities may have the protection of intellectual property laws.

Switching Costs and Compatibility

Also, there is the question of switching costs. Whilst the Internet itself is based on open standards, more complex interfaces between businesses and online trading mechanisms arising in B2Bs may involve significant internal investment. It is possible that networks will do deals to reduce the cost of access in order to promote the development of an application. However, although entry costs may be lowered, this does not necessary mean that switching costs will be low.51 Users may be locked into a network, which increases the ability of the incumbent to exercise market power.

Some may argue that in order to attract customers, B2Bs will ensure that switching costs are low and that their systems are compatible with others. However, this will depend on the degree of underlying market power the B2B may have at the outset and the degree of sophistication of users in identifying whether they are being “locked in” or not. In some cases, it may be that compatibility can occur - but only if the user purchases an expensive conversion software program.

Other Barriers to Entry - Supplyside Economies of Scale

It may be that certain types of B2B also exhibit supply side economies of scale, and may require significant investment - particularly in relation to research and development - which should also be taken into consideration in assessing whether new entry is likely to occur.

Accordingly, there are a wide range of factors that need to be considered before determining whether a B2B may, because it exhibits network characteristics, have the ability to exercise market power.

Assessing the conduct of a B2B under the Trade Practices Act

Access Regimes

At this time, B2Bs are not covered by a specific access regime under the Act. As outlined above, Part IIIA is only likely to apply in relation to services to provide facilities which are uneconomic to duplicate, or of national significance. Whilst this has not yet been tested by Australian courts, it is unlikely that many B2Bs will fall within the scope of Part IIIA. However, if a particular B2B grows to a size where it is

50 Shapiro & Varian, Information Rules, 1999.

51 FTC, Entering the 21st Century: Competition Policy in the World of B2B Electronic Marketplaces, 2000 Part 3 p 23.

22

considered to be of national significance or uneconomic to duplicate then it is conceivable that in the future a small number of B2Bs may be subject to an access regime.

Competitor Collaborations to Establish a B2B

Arrangements between competitors in wholesale markets to establish a B2B or mergers between B2Bs may raise issues under s 45 or s 50 of the Act, where it is likely to result in a substantial lessening of competition in a market.

Where the arrangement enables competitors to combine their existing market power in wholesale markets to create a combined position of substantial market power, this may provide those competitors with an opportunity to exercise that market power to harm their competitors and to engage in collusive behaviour. This may occur in a number of ways including the exercise of monopsony power (where the B2B is owned by buyers), tacit collusion on price, quantity or other strategic information sharing, and refusal of access to, or discrimination against third party competitors. In each case, this may depend on the nature of the B2B which is proposed.

The analysis may also take into consideration the effect of the development of market power on the competition for B2Bs, and in particular whether this is likely to stifle the development of innovative and efficient competitors.

Conduct of a B2B - Membership and Operating Rules

The ongoing conduct of a B2B - whether it is formed as a competitor collaboration, a single competitor, or an independent B2B - may also raise issues under s 45, 46 and 47 of the Act.

Some issues may arise particularly in relation to membership rules which exclude access to some businesses where this is likely to result in a substantial lessening of competition in the relevant wholesale markets. In each case it will be necessary to consider carefully whether exclusions restrict competition, or actually enhance it. There may be cases where exclusions may appear restrictive, but in fact are put in place to protect other users (for example, prudential controls). Also it is important to consider whether rules are clear and objective, or whether they contain subjective elements which may enable owners to use them for competitive purposes.52

If the rules of the B2B enable competitors to obtain commercially sensitive information about each other’s trading activities, or to combine their trading activities, this may raise issues of price fixing or tacit collusion under s 45 of the Act.

Also, careful consideration will need to be given to exclusivity clauses, or conduct which promotes exclusivity as this may be used to increase barriers to entry and preserve market power.

52 Forthcoming paper, Gans & King, Competitive Issues Associated with B2B E-Commerce, 2001.

23

Role of Network Effects in Competitive Analysis

Proponents of the school of thought which argue that network effects do not raise market power issues may suggest that such issues are unlikely to arise, and even if they do arise, would be counter-balanced by the pro-competitive benefits of B2B marketplaces.

It may be argued that operators of a B2B would be unlikely to unreasonably refuse access as it is in the interests of the network owner to encourage participation in order to increase the value of the B2B itself. In fact, if it were to do so then it would risk losing customers to a competing network. Also, it could be argued that participants would not collude on price or share information because if they were to do so, customers on the other side of the market would move to a competing network.

However, as discussed above, it is important to note that whilst network effects may assist in the development and maintenance of market power, issues do not arise purely due to network effects. There are other factors in play which, combined with network effects, lead to the development of market power which could enable a B2B operator to engage in such conduct.

Assessment of Pro-Competitive and Public Benefit Elements

That is not to say that the Act does not provide for the consideration of pro-competitive elements in assessing B2Bs. In assessing the potential effects on competition pursuant to s 45 and s 50, or whether a person has substantial market power for the purposes of s 46, pro-competitive elements may be taken into account where it can be shown that as a result of such elements, anti-competitive conduct is not likely to occur.

Alternately, the parties may seek authorisation to enable a wider consideration of public benefits in determining whether to allow a B2B to proceed.

Addressing Concerns about the Application of Competition Laws to Networks in the New Economy

At the beginning of this paper, three common issues or concerns regarding the application of competition laws to industries exhibiting network effects in the New Economy were identified. The following paragraphs then provided a brief summary of the competitive implications of network effects, competition laws in Australia which currently deal with such issues, some recent domestic and international case examples and a brief overview of the potential application of the law to an important development in e-commerce - B2B electronic marketplaces.

On the basis of this analysis, I would make the following comments in relation to those concerns:

1. It is sometimes argued that competition regulators over exaggerate the potential market power issues arising in relation to network effects. This, it is submitted by some commentators may be detrimental and stifle pro-competitive ventures.

24

Under the Act, network effects are one of a number of factors which are taken into account in determining if a firm or firms have a substantial degree of market power in a market. Products or services which attract network effects do not necessarily fall within the scope of specific access regulation or pricing controls, and in fact, in many cases will be unlikely to do so. Also, as discussed in relation to B2Bs, not all networks will have sufficient market power to raise issues under the competitive conduct rules. However, where access regulation is applied, such as in the telecommunications area, or where competitive conduct rules have been applied, as in the credit card payment system, the objective of regulatory intervention was not to stifle competition, but in fact to assist the development of pro-competitive innovations which were threatened by the conduct of incumbent network operators.

Also, it is notable that current competition laws would rarely prevent the development of a network. Most regulation, with the exception of s 50 and s 45 which could restrict collaborations to form a network, is designed to address the conduct of a network operator, not prevent the development of a network.

2. It is sometimes argued that in New Economy markets, the potential threat of new entry is particularly strong because technology is changing rapidly, therefore, any market power that may arise due to network effects will be short lived.

As seen in the Microsoft case, one of the key issues is not simply whether network effects exist, but whether the parties are able to use that position not only to do competitive damage in another market, but to preserve their position in their primary market. Accordingly, while in some cases it may be that new inventions will lead to the rapid rise and fall of some networks, it is important to consider under competition laws whether the owner of the network is behaving in a manner that is likely to prevent such competition from occurring.

3. It is also argued that competition regulators confuse network owners competitive conduct in offering lower prices and otherwise encourage access to, and participation in their networks with predatory behaviour.

As seen in relation to the Microsoft case and credit card schemes, networks sometimes do restrict access and low prices may be subsidised by the setting of high prices to other market participants. Accordingly, it is prudent not to assume that all networks will act in a pro-competitive manner at all times. This will be a question of fact in each case.

Conclusion and Further Issues

The above discussion indicates that the existence of network effects in e-commerce products and services does not necessarily show, by itself the existence or the absence of market power. The above analysis of the Act also indicates that the legislative framework is relatively neutral in its treatment of network effects, and whilst not discounting the potential market power issues, is capable of balancing such concerns with potential pro-competitive arguments.

25

However, there are still some key policy issues arising from the application of the Act to networks that may need to be addressed.

These include:

If pro-competitive issues will be assessed under competition rules or under the authorisation process

Under the current legislative framework, pro-competitive elements of networks may be taken into account in assessing whether conduct is likely to substantially harm competition. However, in order to have wider public benefits such as employment opportunities and industry growth taken into consideration, parties may need to submit their conduct for authorisation by the ACCC.

Where conduct involves a per se offence (for example price fixing, as was the case in relation to credit card interchange fees), pro-competitive arguments may not be taken into account unless the parties seek authorisation.

In assessing efficiency benefits and pro-competitive elements under either an authorisation or competition test, difficult issues will still arise in assessing the potential strength efficiency gains in relation to new technology markets. For example, it may be argued that allowing participant-ownership in a B2B is important in order to guarantee throughput - otherwise a B2B may be too risky to launch. However, it will be difficult to test this proposition in an immature market environment.

Treatment of Intellectual Property Rights and s 51(3) exemption

Intellectual property rights will be highly valuable in e-commerce, and as seen in the Microsoft case, how they are used may have important implications for the competitiveness of markets. However, currently the application of the Trade Practices Act to the use of intellectual property rights is uncertain, and may be limited to s 46 situations. This may not be adequate to address all issues, as this would only cover issues which constitute a misuse of market power and fall under a purpose test. Other provisions of the Trade Practices Act also involve an effects test which bring into consideration the wider competitive implications of the conduct in question.

Duration of Market Power

This paper suggests that a cautious approach should be taken in assuming that in new technology industries market power associated with network effects will be fleeting as new technology overtakes old. However, regulators will still need to consider carefully what constitutes a significant period of time for the purposes of identifying whether intervention should be taken to prevent the use of market power. This may need to take into account not only the absolute length of time that an incumbent may be expected to be in a position to exert substantial market power, but what degree of consumer detriment may be caused, even in a relatively short time period.

Section 46 - Purpose v Effects Test

26

In some cases, conduct of a network operator will be assessed under s 46 of the Act. Because s 46 relates to the purpose of the conduct, rather than the effect on competition, it may be difficult to apply, particularly where the motivation behind the conduct appears to be ambiguous. For example, when a network operator decides to offer low connection prices to its network, this may imply either a pro or anti-competitive purpose, depending on the facts. In fact, it may be more appropriate to apply an effects test which goes to the heart of assessing the competitive implications of particular conduct, and is more suited to balancing pro- and anti- competitive elements.

Administrative Issues - When to Assess B2Bs

As discussed above, B2B issues may arise during the formation stage and on an ongoing basis. Whilst a B2B may seek ACCC clearance for the start-up service, it may be that over time the membership and operating rules change, the nature of the service may alter as new business opportunities emerge and the B2B’s market position changes. This raises the issue of whether a B2B collaboration which has initial clearance will need to continue to seek informal clearance from the ACCC when its rules change. In fact, where a particular venture has been authorised, parties will need to consider whether a particular rule change will in fact invalidate the terms of authorisation and require further or separate authorisation.