The short future of public broadcasting: Replacing digital terrestrial television with internet protocol?

Capacity and coverage: Cellular/mobile networks

Although the ‘mobile’ telephone has existed for decades, its capabilities were limited and it remained a niche product for much of that history. From the 1980s, with the development of cellular networks and the reuse of frequencies (a technique long used in terrestrial broadcasting) it became possible to extend usage of cell phones to large numbers of people, such that the vast majority of the world’s population are now regarded as cell phone users (Manninen, 2002: 298).

The mobile phone networks use extensive arrays of low power, short range radio transmitters, each creating a geographically defined ‘cell’ in the network. Frequencies may be reused in non-overlapping cells and so the capacity of the network depends upon the size of the cells: building more transmitters means, effectively, the smaller the cell and the greater the capacity of the network overall. However, increasing capacity in this manner inevitably costs more, and so a balance between capacity and price is struck, based on presumed demand and return on investment. Decisions must also allow for the peak in traffic at times of high demand, when services may be slowed by the number of simultaneous users, and the cost of unused capacity at times of low demand. As examples, in countries like Finland and the UK, such considerations result in tens of thousands of base stations and cell transmitters.

The capacity of the cellular network also depends on the allocation of frequencies for the networks to use: in general, more frequencies offer greater capacity. However, the particular range of frequencies used is significant: lower frequencies travel further than others at low power and also penetrate buildings better, providing better indoor reception. The higher frequencies travel less far and so the resulting large number of transmitters needed means sparsely populated areas are often judged uneconomic and may experience slow rollout of cell phone services or may not receive coverage at all. Many countries assigned the high UHF band for cellular use some years ago: they are now beginning to similarly allocate the lower UHF frequencies recently vacated by the transition to DTT, with implications for coverage. For example, the high UHF 3G services of some UK networks cover just one-quarter to one-third of rural homes some 15 years after the licences were awarded; in contrast, the allocation of lower frequencies is anticipated to enable faster rollout of 4G services (Ofcom, 2014d: 83–90).

Capacity and coverage: Broadcasting

In contrast with cell phone architecture, television and radio transmission has traditionally relied upon a small number of high power transmitters supplemented by a larger number of lower power ‘relay’ transmitters filling in the gaps left by the main transmitters. For example, in the UK 80 main TV transmitters serve around 90% of households, but it requires more than 1,000 to reach a further 9%. Frequency reuse is employed but the larger transmission areas mean efficiency gains are limited and so DTT (and radio) services can accommodate around 50–60 channels. The DTT network is efficient in one sense – the small number of main transmitters, once installed, has only limited oncosts while serving large populations in densely populated areas and can cope with unlimited levels of demand at peak times. The large number of additional relay stations, however, represents a significant cost for operation and maintenance. So, while all broadcasters aim for cost efficiency, PSBs are obliged to provide universal coverage whereas commercial broadcasters are not. In the UK, for example, only the PSBs transmit from the denser array of relay stations (Ofcom, 2014d: 113).

As alternatives to the terrestrial broadcasting platform, satellite and cable delivery have limitations. Satellite reception requires either an individual dish or a relay infrastructure, and not all homes are suitable for either; furthermore, satellite transmission represents a large investment for the platform operator and so in most cases satellite television requires viewers paying by subscription. Cable TV is also usually organised as a paid-for service, the operator charging consumers in order to repay the significant investment required to install the infrastructure. Since cable must be laid to every home that it serves, cable TV services are normally restricted to urban centres – in the UK, cable is available to less than half of all households, for example. Thus, terrestrial transmission, once initial investment in transmitters has been made, provides services to those who either may not have other means of reception available to them, or who otherwise would have to pay an often costly subscription to the television provider. Any consideration of alternatives to terrestrial transmission therefore raises questions about reach and universal access.

Convergence and the ‘black box fallacy’

The proposition of a universal transmission technology – IP based, say – delivering everything, text, sound, video, is superficially an attractive one, an ultimate form of ‘harmonisation’ perhaps. This implies a new era of converged devices – the universal ‘black box’ – able to receive content in all forms over a single, universal standard. Such ‘convergence’ is presumed to be a consequence of digitisation of TV coupled with the emergence of high-speed data standards such as 4G. This focus on technological convergence emphasises the relative novelty of these developments and, it is argued, this novelty should drive changes in infrastructure arrangements. Rather than thinking in terms of convergent technologies, however, the current position is more the result of continual change, both of technology and of the industries that develop and shape it. Goggin, for example, describes the uneven progress in mobile television as being ‘as much forged at the interface of Internet and mobile media cultures as it has at the joining of broadcasting and telecommunications infrastructures’ (emphasis added) (Goggin, 2012b: 120).

Despite assumptions about convergence, we actually witness an increasing proliferation of devices being suited, paradoxically perhaps, to greater diversity of platforms, as Mueller anticipated a decade ago (2004: 314–315). The TV and the smartphone are certainly able to receive the same content but each is used in different ways: TVs tend to be used more for long form, linear watching, usually delivered over broadcast platforms; small devices are suited for short-term attention, where portability means wireless connection is most appropriate, this via a home-based Wi-Fi extension of a wired broadband network, a public Wi-Fi hotspot or a cellular data network. This multiplicity of platforms reflects the diversity of the media and communications sector, an industry which itself is not fully converged. For example, few consumers depend entirely on 3G/4G/LTE for all their communications needs. Thus, three-quarters of UK homes have fixed broadband connections, with the consequence that most ‘wireless’ data traffic takes place using Wi-Fi links to fixed broadband connections at home, work, school or college rather than via a mobile operator (Ofcom, 2014c: 317).

The vision of ever greater convergence and the universal black box has a long and recurrent history. Yet evidence that convergence has taken place is absent. For example, the expanding cable television systems in the 1970s were expected to be universal data networks as much as television pipes, supported by government policies seeking to place themselves in the vanguard of the information revolution (Carey and Elton, 2009). The TV set was no longer to be passively watched but now it would become an information resource. Despite expectations of transformations in information consumption, however, the use of these technologies was slow to change, the networks’ primary purpose (and source of revenue) being TV watching. The changes in consumer behaviour and cultures of consumption that have occurred have not primarily been due to technological change, but rather the restructuring of businesses and related shifts – sometimes dramatic – in government policy in favour of economic liberalisation, privatisation and deregulation. Current debates about changing consumer behaviour, individualised consumption and, thus, a more commercial approach to allocation of radio spectrum are in fact less novel than often assumed and less technological in their explanation.

Questioning growth

There are reasons, then, to question the presumption of an ever-increasing demand for data capacity requiring the allocation of more spectrum from broadcasting to a ‘converged IP’. Beutler and Ratkaj (2014) suggest that insufficient research has been carried out into the equation between more data and increased spectrum allocation. For example, they suggest that reconfiguration of IMT networks to operate smaller cells, as described earlier, could increase capacity without necessarily requiring more spectrum. Further, as noted above, much wireless traffic is consumed via Wi-Fi networks, often at home, rather than cellular networks. This ‘Wi-Fi offload’ – the proportion of traffic generated by cellular devices actually carried through Wi-Fi links to fixed broadband – is a significant factor in forecasting mobile data traffic. Current evidence suggests more than 70% of all smartphone traffic travels via home Wi-Fi as consumers tend to reserve higher bandwidth activities for the home or workplace, engaging only in lower bandwidth activities while outside (Analysys Mason, 2015; Ofcom, 2014c: 320). Uncertainty over offload compounds the problem of producing traffic forecasts, which, Ofcom notes, ‘vary widely’. For example, its own commissioned study suggested traffic growth between 2012 and 2030 could be anything between 23 times and 297 times (Ofcom, 2014b: 3; Real Wireless, 2012: 62). Such variations in forecasts are understandable in a new market. Thus uncertainty nevertheless suggests that it is premature to claim that the broadcast platform is losing ‘relevance’ or becoming ‘antiquated’.

Capacity crunch

Whatever the uncertainty over the precise nature of data traffic growth, it will of course increase substantially. So far, much of this article has highlighted debates about the most appropriate means of distributing audiovisual content as part of that growth. However, in some senses, this is just a small part of a much larger question: can any network, however configured, accommodate the levels of growth some predict? Concern is growing that the capacity limits of all networks, wired and wireless, may soon be reached. Even the optical fibre cables that have replaced copper in much of the communications infrastructure – including that supporting the mobile networks by linking base stations for example – have capacity limits. While one response is to deploy more optical fibre more quickly, any derived capacity increase would be ‘linear’ in nature while data traffic is presumed to increase exponentially. Thus, this may simply delay a little the reaching of the limit, dubbed ‘capacity crunch’ (e.g., Ellis, 2014). The superficially simple solution of deploying ever greater volumes of physical infrastructure also raises concerns about consumption of resources and energy, and so instead emphasis should be on using existing capacity more efficiently (e.g., Wang and Manner, 2013). In this respect alone, it is by no means obvious that replacing broadcasting with IP networks offers any advantages.

PSB coverage and universal access

Most European PSBs have occupied a very specific role in developing and implementing DTT, respectively, in their home countries, and the BBC and Yle are no exception. As mentioned earlier, national governments required both companies to sell their broadcast distribution networks in order to finance the transition to digital television while both continue to have particular obligations concerning DTT provision. The current BBC Agreement (2006) requires the company to maintain its principal television services on DTT, while, more recently – in 2012 – in Finland Yle became required to provide nationwide terrestrial digital television services on both DVB-T and DVB-T2 standards up to 2026. This is in addition to the law obliging Yle to provide ‘versatile and comprehensive television and radio programming with the related additional and extra services for all [citizens] under equal conditions’ (Act on Yleisradio, 1380/1993). However, none of this means that PSB obligations can never be changed, but if the universality principle was abandoned or severely compromised, it is questionable whether the result could then be defined as public service. On the other hand, if national governments continue to require universality, meeting that requirement in an economically rational and cost-efficient way without broadcasting and DTT might be impossible.

PSM universality online and net neutrality

Unlike in the broadcast environment, there is no assured access to public service content in the internet. Until recently, most Europeans did not even have any legal protection for their right to use the full open internet. National laws protected net neutrality – equal treatment of network traffic – only in the Netherlands (2012), Slovenia (2013) and Finland (2015), while all the largest member states relied on internet service providers (ISPs) to comply with either self-regulatory codes of practice or just regulatory guidelines (EC, 2014). Now, these will be replaced by the new EU telecoms law agreed in October 2015. However, some argue that the law is a result of political compromise, and ‘fails to establish the principle of net neutrality and instead it centres on the regulation of traffic management by network operators’ (Horten, 2015). At the time of writing it is still unclear whether certain traffic management practices are permitted or whether these would be in contradiction with the EU legislation on the freedom of expression. Another mechanism guaranteeing universal access to PSB services would be a must-carry type positive discrimination, but given trends in EU policy decisions, it is hard to imagine this being supported.

Broadband is not a Universal Service Obligation (USO) in the EU

It would be impossible to successfully stream video content like live TV without a fast enough broadband connection – and even downloading large digital video files may take ages when the internet connection is slow. So considering the importance of fast broadband for EU Digital Agenda 2020, one would imagine that access to broadband services would also be an integral part of the EU ‘Citizens Rights Directive’. However, the official EU definition of broadband is just 144 kbps while the European USO (2009/136/EC) does not have any reference to broadband at any speed. The European Commission has decided not once but on two occasions not to include broadband in USO (Nuciarelli et al., 2014; Venturelli, 1998). In Finland, the availability of 1 Mbps internet access was made part of USO in 2010 and the minimum speed was raised to 2 Mbps in 2015, while in the UK, the idea of making 2 Mbps broadband internet part of the USO has been discussed for years, but with limited results. Even then, such speeds are rather low for a good quality video.

Public broadcasting is legally prioritised activity

Terrestrial broadcasting and public service broadcasting, in particular, still has a special position in both the UK and Finnish legislation as the broadcast licences are not granted through a spectrum auction. Although auctions are claimed to be a more transparent, clear and efficient system for allocation of frequencies compared with granting licences through a political process (i.e., a ‘beauty contest’) the UK and Finland have both been considering another kind of market system – administered incentive pricing (AIP) – to encourage broadcasters to use spectrum more efficiently. However, broadcasters are not yet charged market-based fees reflecting the opportunity cost of using the spectrum but rather the cost of spectrum management. In the UK, the BBC was also to be subject to the AIP system, but this plan is now on hold at least until 2020 (Sims et al., 2015: 85–86). In Finland, Yle was excluded from AIP as the Act on Yleisradio Oy (1380/1993) obliges the Ministry of Transport and Communications to ‘take into account the operating requirements of public service’. However, the Ministry does not have much power to protect the PSB interests outside the broadcast domain and national borders.

PSB has free access to spectrum but not anywhere else

As noted above, the current system guarantees PSB a free access to an amount of publicly owned spectrum which is necessary for fulfilling the public service remit. In a vertically integrated system, where the PSB also owned and operated the broadcast network, it was able to run the distribution services in house at cost price. But if the distribution networks are owned and operated for profit by listed companies the PSB cannot control the price of its broadcast distribution even though it has free and secured access to spectrum. In a situation where a PSB would abandon the system built for utilising the publicly owned spectrum and instead use only telecom networks, the distribution of public service content would be totally dependent on auctioned and privatised spectrum. PSBs would then either be subject to full market rates or governments would have to introduce some complex subsidy system for existing broadcasters. In a worst case scenario, the network operators would become new gatekeepers to PSM content, while people would be paying for the content twice: first for the production (public funding from citizens) and then for the delivery over the network (operator payments from consumers).

EU allows state aid for public broadcasting

One of the aftermaths of the failure of the European analogue HDTV project was the adoption of the principle of technological neutrality in the EU as one of the five basic policy principles for a new regulatory framework (Kamecke and Körber, 2008; Lembke, 2002). The idea of regulation which does not impose a specific technology but lets the market rather than the state decide on the success or failure of a system may sound simple, but in practice the European Commission has violated its own principles. As Brevini (2013a) has pointed out, the EU Communication on State Aid to Public Service Broadcasters (EC, 2009) is not technologically neutral. It has much more restrictive approach towards PSB online activities than broadcasting. In its earlier decisions about PSB online, the Commission was reasoning that the new online activities should be ‘closely associated’ with radio and TV, which form the foundation of the PSB remit. Similarly, the Communication of 2009 does not question the existing broadcast services but requires an ‘ex ante test’ for significant new services (Brevini, 2013a; Donders and Moe, 2011). According to a strict interpretation, a PSM corporation no longer offering broadcast radio and TV would not be eligible for state aid. One can only speculate what the outcome of any subsequent review of the EU Communication might be.

Increased vulnerability in time of emergency

One of the first lessons learned by American journalists while reporting the events of the September 11, 2001 was that mobile cell phone networks were of no use because of congestion. So many people were trying to use their devices at the same time that network cells became overloaded. Even though the capacity of the mobile networks has improved in the last decade or so, the number of mobile devices and users has also increased. As the basic structure of the networks remains the same, congestion is still a problem in times of emergency – as seen after the bombings at the 2013 Boston Marathon (Stone, 2013). Dedicated broadcast networks can never become congested by the sheer number of users. But if linear terrestrial television services were delivered in the same commercial networks as mobile voice and data, there would be a risk that, at times of technical disruption, sabotage or major emergency, all three communications services could be lost simultaneously (NKOM, 2015). Snowfall in November is by no means any kind of emergency in Central Finland, but in 2015 a heavy snowfall cut down over 400 mobile network base stations for several days after their emergency batteries ran out of power during a long break in electricity supply (Nieminen, 2015).

User security and surveillance

It is to be expected that both commercial and governmental organisations are using communication systems for surveillance purposes. Any form of effective organisation requires surveillance of some kind, which means that organisation and observation are actually ‘conjoined twins’ (Webster, 2014: 280). In traditional mass marketing, distributing identical messages to the largest possible audience was effective perhaps, but as soon as flexible customised production and marketing emerged, there was also a need for consumer surveillance through transaction-generated information. In broadcasting, companies have very few direct transactions with their audiences, but in telecommunications the systems are inevitably producing consumer information for surveillance purposes (Samarajiva, 1996). Every networked device needs an IP address and, in mobile use, location information. While a regular broadcast receiver without any in-built feedback channel lacks the capacity to provide any information of its use, new ‘smart’ devices with internet connection are able to record user behaviour and deliver it to third parties often without the user’s knowledge. In addition to commercial surveillance, governments are also using the interactive networks to monitor their citizens. The UK government alone has paid the major British telecoms companies more than £37 million for data on customers and their activities in the past five years (Thomas, 2016). The use of IP in all communication therefore broadens the scope for comprehensive and continuous surveillance.