CHAPTER FOUR
Assessing competition in pay TV or a wider video or communications market is not a simple task. It involves the subtle, and sometimes not so subtle, interplay between the actions of pay TV and other video and telecommunications providers in the short run, together with longer-run structural issues.
This chapter sets out the rudiments of a framework for analysing the emerging tension between competition, regulation and market structure in pay TV and related communications networks. The discussion is designed to set the scene for the more detailed discussions of competitive issues raised by the pay TV industry under Australian trade practices law in Chapters 5 and 6.
THE BASICS
There is no doubt that as an economic commodity, television is unusual. (5) A television programme is what economists call a public good in the sense that, once produced, it can be broadcast to an additional viewer at zero cost. Thus, for any given programme, the larger the audience of a channel, the lower is the cost per viewer of programming and of many other activities associated with the distribution and marketing of television.
These economies of scale explain much of the organisation of television markets. They explain why television networks exist and why there is often a threshold below which it is not commercially sensible for pay TV channels to produce their own programmes. A large network can spread the fixed costs of programming over more viewers, and has the financial capacity to produce more expensive programmes. Networks also substantially reduce the transaction costs of scheduling, transmitting and selling airtime, and enable schedules to create and capture the value of "adjacency effects" where a programme builds audiences for the next one. The fixed costs of programme production also explain the way rights are priced and "windowed" across different formats and delivery systems. (6)
There are also economies of scope. Economies of scope arise when an increase in the production of one product leads to a reduction in the production costs of another. For example, a railway or an airline may find it less costly to supply both passenger and freight transport than to specialise in only one service. The bundling of pay channels will generate economies of scope, as will the bundling of pay TV and telephone services on broadband cable. The latter is particularly significant in Australia, where the Optus fixed network has been constructed to carry both services. This means that the additional cost of supplying telephony and pay TV is significantly lower than supplying each on a stand-alone basis.
The other relevant feature of television is that audiences are relatively fixed. If anything, people are watching less television, presumably because other forms of leisure and entertainment are competing with television. Furthermore, there is little evidence that increasing the number of channels or introducing pay TV channels increases average viewing hours, although pay TV subscribers may watch more hours on average than those who watch FTA television only. Therefore, as the number of channels or stations increases, the same audience is spread or fragmented among more channels. This means that, all things being equal, the introduction of pay TV takes audience share away from FTA broadcasters, and tends to reduce average audiences per station or per channel. The net effect is to increase the average costs of attracting viewers even if the greater costs of delivery/transmission, and the impact that increased competition has on rights fees and programme production costs, are ignored.
COMPETITION AND PROGRAMME COSTS
Increased competition in television has a two-way pull on programme costs. An increase in the number of bidders increases competition for programme rights and resources. This, in turn, drives up the fees that broadcasters must pay for attractive programming. However, increased competition also fragments audiences and reduces the share of audience and revenues of each broadcaster, who will respond by reducing overall programme costs in two ways. First, broadcasters and third-party programme producers will increase the production of television programmes. Second, broadcasters will alter their schedules to incorporate cheaper programming and more repeats. For example, drama may be replaced by game shows and fly-on-the-wall documentaries.
The introduction of pay TV increases programme costs for two further reasons. First, the value of a pay TV subscriber is several times the value of the same viewer to an advertiser-supported FTA station. This is because subscribers are often willing to pay more for attractive programming than advertisers are prepared to pay for advertising airtime around those programmes. In the UK, the average revenue per subscriber hour generated by UK pay TV is twice the advertising revenue per hour for a FTA channel, although for premium programming, like movies and sport, it will be many times higher than the potential advertising revenue.
Second, the prices for pay TV rights tend to escalate rapidly where there is direct competition between two or more pay TV operators. In countries where pay TV has developed with several operators vying for programmes, a bidding war has broken out for Hollywood movies and major sports events. These are referred to as "drivers" or "killer applications" because of their importance in attracting new subscribers. Both attract mass audiences and both are highly valued by viewers. Also, the quantity of this programming is limited. Thus, competition in the programme rights market, particularly in the early phase of the development of pay TV, leads to a high-cost structure for the industry which, in retrospect, is seen to have been based on wildly optimistic projections and a "winner-takes-all" strategy premised on the assumption that the competitor will not survive. For example, the costs of the Hollywood output deals were substantially renegotiated downward when the UK's pay TV sector was rationalised in the early 1990s. (7)
The impact of pay TV on programme costs, however, should not be exaggerated. In general, programme expenditure per channel and per hour for a pay TV operator is low compared with FTA channels. This can be seen using the programme budgets of UK FTA and pay TV operators. BSkyB spends a fraction on programming per channel compared with UK terrestrial broadcast channels, although on a per-viewer-hour basis it spends the largest amount, since it has many hours of programming but relatively few viewers.
Second, the dramatic escalation in programme costs is confined to specific types of programmes. BSkyB spends the bulk of its programme budget on acquiring the rights to premium programmes. In 1995–96 expenditure on movies accounted for 34 per cent and sports 32 per cent of its total programme budget. That is, 66 per cent of the total programme expenditure was devoted to these premium channels.
Third, pay TV operators rely heavily on acquired programmes. Pay TV's general entertainment programming generally consists of acquisition of secondary rights in television programmes which have already received their first transmission on FTA television overseas. In the early phase of the development of pay TV, there is very little original production made specifically for pay channels. This is especially true of new pay TV channels where the economics of entry and the volume of programming required make it uneconomic for pay TV operators to engage in significant production of original programmes.
Pay channels also schedule their programmes in different ways which reduce costs. The number of repeats is considerably higher than on FTA. For example, six- or four-hour blocks of programming are repeated over the course of the day or week.
Table 4.1: Programming Expenditure in the UK:
Terrestrial vs Pay TV, 1996
| Channel | Programme expenditure £m | Number of channels | Expenditure per channel £m | Viewer hours | Expenditure per viewer hour (pence) |
| BBC | 1,130 | 2 | 565 | 21,910 | 5.2 |
| ITV | 810 | 1 | 810 | 17,554 | 4.6 |
| Channel 4 | 268 | 1 | 268 | 5,301 | 5.1 |
| BSkyB | 420 | 10 | 42 | 5,155 | 8.1 |
Sources: BBC, Annual Report and Accounts, 95/96, ITV, Annual Report and Accounts 1996, Channel 4, Report and Financial Statements 1996, BSkyB, Annual Report 1996, Advertising Statistics Yearbook 1997.
PROGRAMMING CHOICE
In order for pay TV to succeed it must offer viewers something significantly different from FTA television. Obviously, pay TV has greater channel capacity, thus enabling more and different programmes to be sold to subscribers. This greater channel capacity has led to the development of thematic channels devoted to one type or genre of programming. As shown in Chapter 2, pay TV led to a ninefold increase in channels devoted to specific types of programmes such as movies, documentaries, news, sport, children's, lifestyle and many others catering to specific viewer tastes and preferences.
Pay TV alters the relationship between broadcaster and viewer. It creates a genuine market for programmes, unlike commercial FTA television, which is best described as a market for audiences sold to advertisers. The direct contractual link between channel provider and viewer means that viewers' preferences backed by willingness to pay are conveyed to the pay TV operator. This, in turn, gives the operator a monetary incentive to satisfy those preferences. (8) As Professor Jora Minasian (1964, page 74) commented in an early analysis of the subject:
... a subscription system can be expected to yield a more diversified program menu than an advertising system because the former enables individuals, by concentrating their dollar votes, to overcome the "unpopularity" of their tastes.
This is because pay TV can tap the willingness to pay of those with intense preferences for particular programmes or a different mix of programmes from that offered by the FTA broadcasters. A competitive pay TV system will offer more channels or programme types to viewers.
Commercial FTA television typically consists of a handful of general programme channels providing a broad mix of programming catering to all tastes over the schedule. Moreover, advertiser demand for mass audiences has an effect on programming. Advertiser-supported FTA broadcasters usually seek a mass audience. In a competitive setting, broadcasters with one FTA channel tend to duplicate programme schedules in an effort to maximise audience share. Thus, it is claimed that advertiser-financed FTA channels result in wasteful duplication, lowest-common-denominator programming, and too little diversity and variety.
In practice, while these effects exist for FTA television financed by advertising, the interactions are more complex. First, many FTA channels are heavily regulated to encourage greater diversity and "public service" programming. Second, in Australia and many European countries state-owned (public-service) channels compete for audiences directly with commercial (advertiser-financed) FTA stations. The programming output of these channels often places a competitive constraint on the commercial FTA stations.
COMPETITIVE PAY TV SYSTEMS
A number of commentators have drawn attention to the competitive dynamics in pay TV. Indeed, a recurring theme in recent discussions of pay TV, and a central feature of the ACCC's decision to block the proposed merger between pay TV operators, is so-called "network effects" which generate positive feedback that result in a large operator becoming larger, and eventually dominating the industry. For example, the ACCC (1997) stated that the proposed merger between FOXTEL and Australis in 1997 would give FOXTEL more subscribers, and therefore enable FOXTEL to outbid Optus Vision for better programming, which in turn would attract even more subscribers. This would generate a "death spiral" for its competitors. Other commentators have gone further, contending that pay TV is a natural monopoly (Graham & Davies, 1997, page 17):
... we have a critical dilemma for public policy. High quality material can still be produced and yet cost very little per unit provided that it reaches a large number of people (exploiting economies of scale) and/or provided that it is used in a wide variety of different formats (exploiting economies of scope), but the exploitation of these economies of scale and scope imply concentration of ownership. Thus, even though the new technology has removed once [sic] source of monopoly, spectrum scarcity, it has replaced it with another, the natural monopoly of economies of scale.
These claims are exaggerated. The existence of scale economies does not imply monopoly provision. A high fixed-cost structure tends to limit the number of competitors but the optimal number of competitors will depend on the size of the market. Thus, there tend to be more radio stations than television stations, and more television operators than telecommunications networks. All these are more numerous in larger countries than smaller countries. The constraining influence of costs explains these differences. If these networks supply differentiated products, then more network competition may be profitable. The real question is whether, given the fixed costs, demand is able to generate sufficient revenues to permit two or more competitors to coexist profitably.
In television, competing stations and networks do profitably co-exist. The same cost conditions as exist for pay TV exist for FTA television and mobile telephony, where there is active competition between three or four networks. For example, Channels 7, 9 and 10 compete aggressively for ratings in such a way that one channel's gain is another's loss. That is, in a small-numbers setting typical of network television, there is a recognised interdependence between the stations, which leads to a competitive reaction. It could be argued that were, say, Channel 9 able to break away and establish a larger audience (as it has done), then it may be able to lock into a virtuous circle of better ratings, more advertising revenue, greater ability to buy better programming, and still better ratings. (9) Yet what we see is the viable competitive co-existence between three commercial networks in Australia, not a "death spiral". The reason is that channels confronted with the prospect of decline and loss of audience share or revenues react. Channels 7 and 10 will go into the market and spend on programming to arrest their audience decline: they invest, break the "spiral", and re-establish the equilibrium. Furthermore, the profitability of these channels clearly shows that television is not a natural monopoly, and that there is probably room for additional channels -- a fourth network such as the Fox Network in the US, and a fourth network in Australia (ABA, 1998; ABS, 1998; and Albon & Papandrea, 1998).
In addition, TV programmes are not homogeneous products. TV stations can differentiate their schedules to cater better to a specific part of the audience. The positioning of the three commercial FTA networks shows this. Channel 9 has the highest ratings with the highest revenue for those ratings largely because it dominates the early evening peak (6pm to 7pm). To achieve this, it has pursued a high-programming-cost strategy. Channel 7 attempts to follow Channel 9 with a high-cost/high-ratings strategy. On the other hand, Channel 10 has opted for lower-cost programming appealing to an audience under 35 years of age. If the costs of this strategy are significantly low it can afford to fall behind in the ratings, provided that this is not offset by a more than proportionate decline in advertising revenues. (The ACCC's arguments concerning these network effects are discussed more fully later in this chapter.)
FACILITIES-BASED COMPETITION
So far the discussion has been confined to the economic and competitive pressures between programming providers, ignoring their different ways of delivering their programming to viewers. However, competition between pay TV operators often involves competition between different delivery platforms or facilities. This is especially so in Australia, where pay TV platforms have been tied to one pay TV operator and are "closed" in the sense that third-party pay TV operators do not have access to the platform. Thus, FOXTEL is the only provider of pay TV on Telstra's broadband network and Optus Vision on Optus's broadband network; and, before it went into receivership, Australis had exclusivity of DTH satellite pay TV. Australian pay TV is not strictly vertically integrated because the TNC Heads Agreement made Galaxy core programming available on two competing delivery systems in different territories -- FOXTEL and Australis franchisees on satellite/MDS. In other countries pay TV is organised differently. In the UK, cable and satellite compete directly for viewers. The main programming provider, BSkyB, actively markets channel packages including the channels of other operators and also wholesales the Sky channels to cable networks, which compete directly with BSkyB for subscribers. Nonetheless, cable networks are still closed unless access is negotiated with the cable operator.
In order to understand fully the impact of the vertically integrated structure of Australian pay TV it is necessary to stray into some basic telecommunications economics.
The extent and economic viability of competition between facilities are determined by supply-side factors (fixed and sunk costs of networks) and demand-side network effects.
Building telecommunications facilities, either fixed or wireless, involves a significant initial investment. Optus's broadband network cost $2 billion to pass 2.1 million homes: a cost of nearly $1 000 per home passed before any are connected. If there are two wires passing the same homes, the costs may be twice as large, and escalate substantially if calculated on the basis of homes connected. These costs are fixed and so, in a more graphic way than pay TV programming costs, influence the efficient number of networks. They are also sunk costs in the sense that the network's value in alternative uses is significantly lower than the invested sum or its value outside the pay TV sector. This feature is particularly strong for Australian broadband networks where there is significant "overbuild". Optus's and Telstra's broadband networks overlap by more than 80 per cent, so that if one decided to exit, the value of its network to potential buyers would be much lower than its capital costs. Sunk costs may also be deliberately incurred to signal a strong commitment to the market, a strategy of "burning one's boat behind you" (Dixit, 1980). Both Optus and Telstra Multimedia have adopted this strategy in rolling out their networks, signalling their commitment to remain in the market. Finally, the investment is in the construction of networks which have lives of anything from ten to 25 or more years.
The cost structure of telecommunications networks has two other influences on competition and market structure. The first is the so-called "integer problem". This occurs when demand is sufficient for more than one network to operate at the minimum efficient scale (MES) but not all networks. For example, suppose MES is 150 units of output, but the competitive output is 220 units. This would support one network operating at MES with the other having average costs well above the first. It would therefore be hard for the second network to compete effectively. The result will either be monopoly, or duopoly of one strong and one weak operator (or two weak operators) which could easily be unstable. Second, the cost structure of networks exhibits high fixed but low and often negligible marginal costs. This is especially the case for broadband networks, where the addition of more traffic on a high-capacity network which has already been built is virtually costless. This cost structure contains the recipe for destructive competition in which an emphasis on price may lead to severe financial difficulties as competition drives price down to zero marginal costs, and revenues do not cover costs.
These cost considerations differ considerably as between different delivery systems. Satellite has the lowest cost per subscriber, followed by MDS and then cable. The costs of cable will vary depending on the type of wires (coaxial or the more expensive optic fibre) and the civil engineering used (ducted cable is more expensive than aerial cable). On the other hand, these delivery systems have different functionality and capacity. Broadband cable has greatest bandwidth (that is, it can carry more traffic and channels), and is interactive like a telephone. Satellite and MDS are not fully interactive and often have more limited capacity. As a result, broadband cable has the ability to offer telephone and data services, which give rise to significant economies of scope in the provision of both pay TV and local telephony (joint provision is cheaper than separate provision).
ENTRY ASSISTANCE
Telecommunications policy and regulation in Australia have been designed to overcome one major problem: the monopoly of one entrenched public network operator with access to all customers, namely Telstra. Telstra was therefore in a position to block others using this network or to extract onerous terms. Moreover, the competition issues become more acute when the owner of the facility also provides services (that is, when it is vertically integrated). It then has a conflict of interest since it provides an "essential" input to enable other service providers to compete directly with its services. A policy of laissez faire was therefore regarded by policymakers as inadequate to generate sustainable competition within a reasonable timeframe. This led to regulatory affirmative action designed to assist entrants to overcome the ubiquity and entrenched position of Telstra.
Governments have used two approaches to open the telecommunications market to greater competition: service competition and facilities-based competition. This distinction is crucial to understanding regulatory policy in Australia. Service competition takes place between the content providers such as different pay TV operators, or between those providing telephone and value-added services on telecommunications networks. Facilities-based competition relates to the provision of carriage or delivery services by different operators (in other countries it is called "network competition", "alternative infrastructure" or "infrastructure competition"). It can involve head-to-head competition using the same technology such as between broadband networks (called in the US "overbuild"), or between networks using different technologies such as fixed and satellite networks.
Most countries have liberalised their telecommunications sector. In doing this they have focused on these two areas of competition, often giving one greater emphasis during the early phase of liberalisation. Some regulatory regimes encourage service-based competition through open access and interconnection arrangements on fair, reasonable and non-discriminatory terms. Under this "interconnection model", the difficulties involved in building a new fixed network to compete with an entrenched facilities operator are treated as sufficiently great to require perhaps a decade to remedy, and so competition is not introduced rapidly. Moreover, in the absence of an effective access regime, an entrant will be disadvantaged because it cannot offer potential customers universal connectivity (the ability to call anyone connected to a network). If there are no interconnection arrangements with other networks, then the value of any newly constructed network will be substantially reduced because (as discussed in greater detail below) the value of a network service increases with the number of other individuals using or interconnected to that network.
Australian telecommunications policy has gone down a different route. It is based on the premise that direct competition between facilities providers is essential for a fully competitive telecommunications market. It therefore encouraged entrants to build new networks to compete directly with Telstra's fixed network. By giving consumers a choice of network, large benefits may arise in the form of lower prices, better service, more innovation and generally a better deal for consumers. (10)
A facilities-based policy often involves an initial distortion of competition known as entry assistance or asymmetric regulation. Entry assistance is based on a simple proposition: special measures are needed to overcome the entrenched position of the incumbent network operator and to give the entrant the incentive to make the investment needed to build a network, much of which will be sunk and so irretrievable. The best way to foster effective competition is to allow only one new entrant, and to give it protection for a limited period while at the same time bearing down on any anti-competitive abuse by the incumbent: the so-called "duopoly policy". Australia operated such a policy, giving Optus five years of exclusivity as the only public telecommunications operator (PTO) in addition to Telstra.
Entry assistance can take a variety of forms, some of which appear contradictory. The incumbent PTOs can be prohibited from certain lines of new business, or entry into related activities such as pay TV. Access arrangements have often been "rigged" to achieve entry assistance objectives by setting interconnection charges either too high (thereby encouraging entrants to build their own facilities) or too low (thus fostering service competition based on cheap carriage).
A facilities-based policy is frequently based on closed or proprietary networks. In order to encourage entrants to invest and take the risks to build a new network, they are often permitted to determine its uses and users. Thus, C&W Optus was allowed to build a new broadband network on which it exclusively provided pay TV. This approach is also usual for mobile telephone networks across the world with the result that upward of three networks compete. The facilities-based model employed in Australia has therefore created (quasi-) vertically integrated pay TV systems offering exclusive programming on incompatible networks. It is this structure that some see as the source of the pay TV industry's problems.
CLOSED NETWORKS AND COMPETITION
To understand fully the dynamics of competition between closed networks it is necessary to discuss another economic concept: network effects or demand-side economies of scale. Network effects exist when the number of other users affects the value of a product or service to a user. (11) It is therefore closely related to the economist's concept of an externality or third-party effect; however, it arises not from technological or cost factors but because the demand of consumers is interdependent. (12) Professor Jean Tirole (1989, page 405) offers a more detailed definition:
Positive network externalities arise when a good is more valuable to a user the more users adopt the same good or compatible ones. The externality can be direct (a telephone user benefits from others being connected to the same network; computer software, if compatible, can be shared). It can also be indirect; because of increasing returns to scale in production, a greater number of complementary products can be supplied -- and at a lower price -- when the network grows (more programs are written for a popular computer; there are more video-cassettes compatible with a dominant video system; a popular automobile is serviced by more dealers).
Thus the economic benefit or value of an additional subscriber exceeds the value of the transaction to individual subscribers. His or her connection confers an external benefit. The literature also refers (approvingly) to Metcalfe's law which states that the "value" of a network increases geometrically with the number of people who use it. In telephone networks, positive network externalities provide the motivation for networks to interconnect.
It is argued that this demand interrelationship leads to reinforcing feedback that generates growth and economies for larger networks. In the computer industry, for example, users will pay more for a popular computer system, other things equal, or opt for a system with a larger installed base if the prices and other features of two competing systems are equivalent. This apparent advantage, it is argued, enables firms with a high market share to get larger, leading to monopolistic market outcomes. The implication is that a small network is at a disadvantage to a large network, and that there may be a critical size for a network to be viable.
How does this relate to Australian pay TV? We have seen that the ACCC attempted to invoke a type of network effect arising from the effect that more subscribers would have on the ability of FOXTEL to pay for more and better programming. This is, however, not in the same class as the network effects discussed above. The demand of individual subscribers for pay TV services is not interdependent. The impact and existence of an additional subscriber do not directly affect the value of pay TV to other subscribers. It is true that there are economies of density in cable networks as the average infrastructure costs and the costs of settop boxes decrease as total take-up increases. But this is a supply-side effect unique not to a specific cable operation but to the take-up of cable generally when unit costs fall with volume production.
However, where network effects do come into operation is when subscribers are required to choose between incompatible pay TV platforms. (13) In such cases, consumers will often be reluctant to commit to one platform, fearing that their up-front investment in receiving equipment will lock them into one pay TV operator.
This phenomenon leads to a variety of effects. First, the wrong technical standard may be selected. The literature refers to excess inertia arising from the fear of early adopters of a new technology that they risk losing their initial investment because the technology is not adopted by a sufficiently large portion of the market. In this case, the parties' fear of being "stranded" with a low-value technology (14) may result in deferring its purchase. If many in the market take this stance, there is "excess inertia" resulting in an efficient standard not being adopted, even though buyers would be better off if it were. Others have questioned this conclusion, arguing that this literature offers a set of hypothetical market failures and misinterprets the limited evidence, and that market forces do in practice lead to better choices than governments. (15)
Nonetheless, the existence of technical incompatibilities has implications for competition in the pay TV industry. There are two effects: consumer switching costs and lock-in on the one hand, and reduced value of individual pay TV offerings on the other.
The costs to consumers of switching between products can limit competition and affect take-up. Prospective customers will recognise that if they purchase a satellite dish and settop decoder to receive the service from a pay TV operator, this cost may be sunk (unrecoverable) and they will be locked into the service. This is particularly the case with satellite or pay TV, where satellite dishes cannot receive the signals of two different satellites because of the line-of-sight requirement, or when different technical standards are used. In these cases, the ability of subscribers to switch between the services is limited by cost factors or simple technical incompatibility. In order to avoid being stranded, potential subscribers may delay their purchase decision. (Often this is accompanied by strategic manoeuvres by the competing operators to increase customer confusion and denigrate one another's service. Platform incompatibility also may reduce take-up because attractive programming is usually split between two exclusive systems.
There is evidence of these effects. For example, in the UK, during the late 1980s, two pay TV operators -- Sky Television and British Satellite Broadcasting (BSB) -- rushed to establish two different satellite pay TV platforms. These used different satellites, required the subscriber to purchase different settop boxes, and offered different programming. The presence of two incompatible systems was not at the time commercially viable, and the companies merged to form BSkyB, which went on to become the success story of pay TV. (16) Australis was also affected by this concern given the reports of its impending collapse.
In Australia, this particular problem has been minimised because the pay TV operators do not require the subscribers to purchase the settop boxes or satellite dishes. Ownership of these is retained by the operator and they are retrieved by the operator on disconnection.
BROADBAND ENTRY STRATEGIES
Australia's duopoly policy led to the construction of two broadband networks which have heavily influenced the provision of pay TV. As noted, the speed with which facilities-based competition developed was not expected, even though government policy actively encouraged it. It was the more surprising because the UK's duopoly policy gave stronger protection to new entrant Mercury (also owned by CWC, the majority shareholder in C&W Optus) by banning British Telecom (BT) from providing pay TV on its network. This, however, had the opposite result, with poor cable rollout and weak competition. Viewed with British policymaker's eyes, the decision of the Australian government to permit the then state-owned Telstra to build a broadband cable network would have seemed the death knell for Optus. The EC Commission has also sought to divest national telecommunications operators from ownership of cable networks in order to foster more facilities-based competition. (17)
A number of explanations have been advanced for why duopolistic competition between Telstra and Optus led to such ferocious broadband construction. The government encouraged Optus to build its own network by protecting it from competition under the fixed duopoly policy, giving it relief from environmental laws so that it could use much aerial cabling, and blocking it from using Telstra's ducting. It has also been claimed that Optus found it difficult to negotiate commercially "realistic" interconnection charges with Telstra and found Telstra's interconnection charges hard to calculate, that the interconnection charge set by the then telecommunications regulator was "too high", (18) and that reliance on interconnection would have placed 30–40 per cent of Optus's revenues under the control of Telstra, its main rival. Further, it was federal government policy to continue untimed local calls using Telstra's network so that interconnection would have not been financially attractive to Optus.
There were also less parochial reasons for the strategy. Optus's entry into the Australian telecommunications market clearly drew on lessons from the experience of the UK cable industry. In the UK, the failure of broadband cable to become the dominant delivery platform for pay TV is widely regarded to have been the result of its fragmented structure (made up originally of over 130 regional franchises), slow and fitful construction, and absence of exclusive programming. Optus's strategy appears to have addressed each of these difficulties. It took advantage of the legal shield from competition by investing heavily in a rapid and ambitious broadband cable rollout, and pre-emptively signed up exclusive programming, particularly movies. This gave Optus the coverage and exclusive programming not available to later entrants. If the absence of these factors was the reason for cable's failure in the UK, then the architects of Optus's strategy must have thought they were delivering a "killer blow" to other entrants and pay TV operators in Australia. That is, a classic "winner-takes-all" strategy was adopted.
Optus Vision's decision to enter pay TV using a broadband network with huge sunk costs and massive financial exposure was a daring one. If other potential entrants, including Telstra, adopted the same strategy or did not regard Optus's strategy as credible and/or viable, the outcome would be an escalation of infrastructure costs and excess broadband capacity with the consequence that two broadband competitors may not be commercially viable. The inefficient cost structure of the industry may nonetheless be justified if the resulting competition generates consumer benefits in terms of lower prices and/or better services in the medium to long run.
Looked at from Telstra's point of view, the picture is similar. Telstra was permitted to enter the video market and to build a network overlapping Optus's. This led to direct network competition in broadband services, including pay TV. Telstra's action did not prevent entry of Optus as an alternative telephony/pay TV operator, a concern which has led several countries to ban PTO entry into pay TV and cable networks. The result is a unique level of network competition. In assessing the competitiveness of the Australian pay TV sector and the alleged anti-competitive nature of Telstra's strategy, this central fact cannot be ignored.
Once pay TV operators have made huge initial investments, what might be their competitive actions? The situation is potentially unstable since the investment is for the most part sunk and the costs fixed. They can therefore price at any number of levels in a desire to build up subscribers or compete. Potential entrants, recognising this outcome, will seek one of two solutions: avoid competition because it is unsustainable, or engage in "cut-throat" competition designed to deliver a "killer blow" to their competitors. Such intense competition can arise in what economists have modelled as a "war of attrition", where intense competitive rivalry is followed by the exit of one of the two firms (Tirole, 1989, pages 311–14).
OVERBUILD AND ECONOMIC EFFICIENCY
The discussion has arrived at one of the critical issues surrounding pay TV regulation: has past government policy and regulation, combined with the commercial strategies of Telstra and Optus, led to uneconomic duplication of facilities? The answer is "yes". As noted above, no other country has so extensively invested in one broadband network, let alone two overlapping networks. (19) Although the costs of technology are constantly falling and demand expanding, there is scepticism about the viability of interactive local broadband networks. For example, after detailed and exhaustive studies of the economics of multimedia, Bruce Egan (1996, page 160) concluded that:
Based on cost data ... even under heroic assumptions of quick mass market deployment, the additional per household monthly revenues required to pay for the original investment is staggering. ... Overall, the current demand and revenue data from the telecommunications sector indicate that a competitive service provider of two-way residential broadband network services faces an uphill battle. New revenue growth is always going to be subject to the ability of households to afford to pay for fancy new services and the terminal devices that support them. ... Even the telco's own financial simulations for public broadband networks are pessimistic.
To the extent that broadband overbuild has directly resulted from entry assistance policy, the government created an inefficiency, albeit a competitive one. This cost of regulation is well recognised. Mark Fowler (Fowler et al., 1986, pp 193–4), past Chairman of the Federal Communications Commission (FCC), wrote on leaving office:
It can be argued that some of the Commission's regulatory actions ... in fact encouraged entry by uneconomic providers and uneconomic construction of excess capacity. If this is true, the gradualist approach to deregulation of interexchange markets will have resulted in substantial, unnecessary costs for society that never would have been incurred in a truly competitive marketplace. Moreover, this approach will have directly increased consumers' costs by requiring regulated firms to charge higher prices to protect competitors during the transition.
The subsidy inherent in entry assistance can only be justified in competitive terms if its effect is to generate more competition in the medium term than would otherwise have occurred and this competition yields benefits to consumers which outweigh the short- and long-term costs and inefficiencies associated with entry assistance. Even if the cost penalty arising from the existence of two overlapping local networks were ignored, direct competition would have profound implications for their financial viability. This is especially so if they rely, as has been the case in Australia, on pay TV. Two pay TV operators dedicated to different broadband networks based on exclusive programming would fragment the audience, drive rates down and split take-up.
THE REGULATORY GAME
There is another consequence of entry assistance, which is becoming more prominent. In today's political and regulatory environment, regulators are increasingly co-opted by entrants, often going beyond the mix of regulation and intervention required to foster genuine competition. Regulation is not imposed from above but is the outcome of the competitive strategy of the main participants, albeit in the political marketplace or courtroom. Getting a favourable regulation or decision can be worth hundreds of millions of dollars and can significantly hamper one's competitors. As Baumol and Sidak (1994, page 128) observe, it takes the form of a Greek tragedy where each actor plays his or her pre-ordained part: the incumbent resists entry; the regulator moves to favour the weaker entrant; the entrant seeks to compete by using regulation to handicap the rival. The result is that the entrant co-opts the regulator who fears the accusation that it has failed to carry out its duties (Sappington & Weisman, 1996, Chapter 8). This problem is succinctly put by John Haring (1985), a former FCC staff member:
A firm does not have to possess a large market share to exercise economic power. The OCCs [other common carriers] do not possess large market shares, but they can certainly exercise power by threatening to make government officials who inflict huge costs on consumers to promote competition look bad. They can do this by threatening to fail. A small market share and low profits can be assets in such an extortion campaign. They make the threat of failure more compelling and thus make it more likely that government officials will yield to extortionate demands and as is always the case with extortionists, giving in merely encourages additional blackmail attempts.
Favourable regulation substitutes for competition, and generates an ethos among entrants of complaining to regulators rather than vigorously competing with the established facilities provider.
Entry assistance gives rise to a type of "moral hazard" problem induced by regulation. In the US, it has been argued that mandatory disclosure requirements imposed on Regional Bell Operating Companies regarding network or service plans reduced innovation. These requirements enabled entrants to quickly copy and oppose proposals of the incumbent, with the result that the incumbent was inhibited from developing its business and the entrant was reduced to imitation. As one commentator stated: "Asymmetric regulation gives rise to an inferior breed of competition -- more adept at imitation than innovation and more prone to battle in the hearing rooms than the marketplace" (Egan, 1996, page 195). This, it has been claimed, leads competitors to "adopt a strategy of 'optimal mediocrity' " (Egan, 1996, page 197). In the UK, there have been recurring concerns that the government's policy created a "cosy duopoly" between BT and Mercury (then partly owned and now fully owned by Cable & Wireless), which made Mercury complacent in the market and overly reliant on regulatory assistance. It was seen as a weak competitor, targeting new business rather than competing head-on with BT. In return, BT adopted a live-and-let-live strategy designed to give Mercury sufficient slack so as to ensure its survival and/or to avoid more unfavourable regulation.
CONCLUSIONS
This chapter has sought to set out the background and basic economics relevant to assessing the efficiency of Australia's pay TV and telecommunications industries. It does not pretend to do more than scratch the surface. Nonetheless, a number of conclusions emerge.
- Economies of scale place a limit on the number of competitors, but their existence does not necessarily imply that the sector is a natural monopoly.
- Competition in the pay TV industry leads to excessive programming and other costs which may be unsustainable. This is because, during the initial phase with two or more pay TV operators, the entry strategy is "winner takes all".
- In Australia this has been exacerbated by government regulation which has induced an excessive and unsustainable level of facilities-based competition in Australia.
The relevance of these factors to the ensuing discussion of the ACCC's intervention is straightforward. The ACCC ignored the above efficiency concerns which suggested that three-operator or three-platform competition was not economically viable. Rather, it addressed a range of narrow demand-side competitive concerns while implicitly accepting that the present structure of the industry required reform. It also adopted a highly bifurcated approach to the treatment of costs: these were ignored where they pointed to industry rationalisation but highlighted where they suggested a disadvantage to C&W Optus as a result of the merger. It is also noteworthy that, unlike some other jurisdictions, the ACCC has acknowledged in its merger guidelines (ACCC, 1996a, paras 5.19–5.20 and 5.159–5.162) that it will take economic efficiency into account when assessing a merger. This it did not do!
ENDNOTES
5. For economic analyses of broadcasting and television see Owen, Beebe & Manning (1979), Home Office (1986), Veljanovski (1989), Hughes & Vines (1989), Owen & Wildman (1992), Congdon (1992), and Beesley (1996).
6. "Windowing" is the temporal and territorial release of licences or rights in video programming to maximise the value to the intellectual property owners. For Hollywood movies a new movie will be released first in the cinema, then on videocassettes, maybe on PPV, then on pay TV, and then on FTA. The release pattern will also vary geographically. The licences will usually be sold at prices which reflect the different market conditions and age of the movie. Windowing is a form of price discrimination.
7. An "output" deal is an arrangement whereby the broadcaster gains exclusive rights or the right of first refusal to the entire catalogue or future output of a studio.
8. For Australian literature focusing on demand-side issues see Parish (1968), Brown (1986), BTCE (1991 and 1993), Brown & Cave (1992) and Albon & Papandrea (1998).
9. There are reasons to believe that this effect may be relatively more pronounced for commercial FTA channels because of non-linearities in pricing advertising airtime. Often, the cost per 1000 audience rises the larger the audience share of a channel such that the price for the same 1000 audience is greater than for a smaller channel providing a 1000 audience but with a smaller commercial share. This has been attributed to the reduced wastage when commercial share increases.
10. The OECD (1995) in a recent review of liberalised telecommunications markets (UK, USA, Sweden, Japan, Australia) found that network competition brings substantial benefits in the form of increased choice, greater innovation, better services, and greater investment in and modernisation of the telecommunication infrastructure.
11. Shapiro & Varian (1998) and Case Associates (1997b).
12. An externality is said to exist if a transaction imposes a cost or a benefit on others not taken into account by the transacting parties. The classic case of an externality is pollution where the production of a good (for example, paint) gives rise to a third-party effect (polluted rivers) not priced in the market. As a result, the activity in question is over-expanded because society at large is effectively subsidising its production.
13. Tirole (1989, Chapter 10), Owen & Wildman (1992, Chapter 7), and Besen & Saloner (1989).
14. Farrell & Saloner (1985) and Besen & Johnson (1986).
15. Liebowitz & Margolis (1990).
16. This destructive competition can occur quite easily in network industries. It occurs often in the newspaper industry when "circulation wars" break out. Australia has had a recent experience with the FTA networks. During the late 1980s all but one of these channels changed hands at least once. These were sold at high prices to Alan Bond, Frank Lowy and Christopher Skase in 1987. To acquire the stations, all these incurred substantial debt that they could not repay; and by 1990 all three were effectively in receivership. Channels 10 and 7 were in the hands of the banks and Channel 9 reverted to Kerry Packer, the original owner.
17. EC Commission (1998) and Draft Commission Directive amending Directive 90.388/EEC with regard to its effective application in a multimedia environment, by legally separating the provision of telecommunications and cable TV networks owned by a single operator (1998). Also Veljanovski (1996).
18. King & Maddock (1996, page 142) suggest that AUSTEL's decision to increase the interconnection charge by more than 10 per cent was one of the motivating factors in Optus's decision to develop its own network.
19. Hazlett & Spitzer (1997) suggest, on the basis of a simple spreadsheet model, that an off-the-shelf cable entrant in the US could earn an 18 per cent rate of return at current cable subscription rates, assuming that the current penetration of 66 per cent is split between the two. The problem is that rates are unlikely to stay the same. The evidence indicates that, in the US, duopolistic competition would lead to rates for the full cable package falling by 20 per cent. Even with this adjustment, Hazlett suggests that entry of a second cable network would be financially viable. However, apart from difference in geography and demand (pay TV penetration of homes with access to cable in Australia is 19.8 per cent, split between two operators), the capital cost of Australia's cable networks will be higher because they are interactive.
No comments:
Post a Comment