Digital Dividend Review

Statement published 13|12|2007

Foreword

This statement sets out one of the most important decisions we have ever made: how to award the spectrum freed up by digital switchover (DSO)—the digital dividend—for new uses. This has been the focus of our Digital Dividend Review (DDR) since it was launched two years ago.

This decision matters for several reasons:

• spectrum is an essential input in the modern world. Its use underpins 3% of the UK’s gross domestic product (GDP) and generates wide reaching benefits for citizens and consumers. But spectrum is a scarce resource, so how it is managed is a critical issue;
• the spectrum presently used by analogue terrestrial television is exceptional because it can readily be used to provide high bandwidth services over long distances and into buildings; and
• the opportunity to put this spectrum to new use does not arise often. Analogue television has been its main use for many decades, under a framework that dates back to 1961.

Most of the spectrum used for analogue television has already been set aside for digital terrestrial television (DTT) at DSO: 256 MHz out of a total of 368 MHz. This will allow a major expansion of the capacity and coverage of DTT.

Our main task in the DDR has been to decide how to release the remainder of this spectrum. This could be put to new uses as varied as mobile broadband, mobile television, more DTT and new “cognitive” wireless services.

Under the Communications Act 2003, our duties are to further the interests of citizens and consumers and to secure the optimal use of spectrum. Our objective for the DDR is to award the digital dividend in a way that maximises the total value to society from its future use. This includes value both to citizens and to consumers. It is expressly not our aim to raise revenue for the Government.

We published proposals for consultation in December 2006. These set out a market led approach, moving away from the command and control regime of the past. Instead of the regulator deciding who could use spectrum, how and for what, we favoured giving users the flexibility to decide how spectrum should be used and the ability to change that use over time as technologies and needs change. This is likely to generate greater value for citizens and consumers.

We also recognised that markets do not work perfectly. We set out proposals to address issues faced by smaller users in particular, like the programme making and special events (PMSE) sector and potential operators of local television.

Our proposals generated over 750 responses and a lively debate. Many stakeholders supported a market led approach, but others called on us to intervene. We had requests to reserve spectrum for many different uses and users, for a range of different reasons. There was particular concern about the prospects for DTT services in high definition (HD) and more generally about how broader social value would be delivered within a market led approach. We also received comments on our proposals for PMSE and local television. The timing of the award was a key issue for many respondents: some argued for it to be brought forward, others for delay. Separately, both the Government and the House of Commons Select Committee for Culture, Media and Sport declared their support for a service and technology neutral auction of the digital dividend.

Since consultation, we have undertaken a significant body of new work:

• we have conducted a separate, detailed consultation on spectrum access for PMSE;
• we have published detailed proposals to upgrade the DTT platform by using new technologies. These will enable a richer and more varied set of services, including HD, without needing extra spectrum;
• we have conducted more market research, responding to comments on our methodology. The new research nonetheless came to similar conclusions, with no single service commanding overwhelming support from citizens and consumers; and
• we have conducted further analysis, including economic modelling and technical analysis, of a wide range of issues.

In light of all the evidence available, we have concluded that we should take a market led approach to awarding the digital dividend. This means that, with one important exception, we will auction the spectrum in a way that allows users to decide how the spectrum should be used and creates the maximum flexibility and opportunities for different technologies and services.

We have carefully considered arguments for reserving spectrum for a variety of particular uses, but we have not found them compelling. Giving spectrum to one use will deny it to others. It will also tend to reduce flexibility and blunt incentives. We recognise that many services can provide broader social value, but we do not think that support via implicit subsidies in the form of spectrum is necessary to realise this. Explicit support through direct funding is more transparent and can achieve a better outcome.

The exception to our market led approach is for PMSE, which already uses interleaved spectrum on a large scale. This is an extremely diverse community, and we do not think it would be able to take part effectively in an auction, creating a serious risk of market failure. We will therefore hold a beauty contest to award a package of interleaved spectrum to meet PMSE users’ needs.

This statement marks the end of the first phase of the DDR. But another, equally important phase is to come: the detailed design of the award process and the licences. Promoting competition and innovation will be at the heart of our work, and we will look carefully at how to encourage entry and guard against the risk of anticompetitive behaviour. Our experience suggests that, if spectrum auctions are well designed, they can be powerful tools for creating a more competitive landscape.

We thank all stakeholders for their engagement in the DDR to date and look forward to working closely with them in the months to come.

David Currie Chairman Ed Richards Chief Executive

Executive summary

Introduction

1.1 Demand for spectrum is growing quickly, fuelled by rapid innovation in wireless technologies and consumers’ appetite for mobility. Wireless services are now widely available at low cost, to the benefit of both individuals and society as a whole.

1.2 But spectrum is a scarce resource, so the way it is managed and used is an important issue for advanced economies around the world. One of our key duties is to secure the optimal use of spectrum in the interests of the UK’s citizens and consumers.

1.3 In the DDR, we have considered how we should award some of the most valuable spectrum likely to be released for new uses over the next 10 to 20 years. Consultation has been an essential element of our work, and our proposals generated over 750 responses from a wide variety of stakeholders expressing a wide range of views.

1.4 This statement sets out our answer to the fundamental question of whether we should intervene in awarding the digital dividend by favouring certain uses or users or whether we should take a market led approach, allowing the market to determine who should use the spectrum, how and for what.

The digital dividend

1.5 Analogue terrestrial television is currently the primary use of 368 MHz of spectrum in the UK. This represents just under half of the most useful spectrum available. Analogue television will cease when DSO ends in 2012.

1.6 The Government decided in 2003 that 256 MHz of this spectrum should be reserved for DTT at DSO. DTT is many times more efficient than analogue television, so many more services can be provided to viewers even while using less spectrum.

1.7 The Government’s decision means that, at DSO, there will be a major expansion of the coverage and capacity of DTT. DTT will be universally available throughout the UK for the first time. In parallel, the capacity of the six DTT multiplexes will grow by at least 20%, enabling a wider and richer range of services without extra spectrum.

1.8 The Government also decided in 2003 that the 112 MHz that will be cleared of analogue television should be released for new uses. This cleared spectrum, with a few adjacent frequencies, forms the core of the digital dividend.

1.9 The way that DTT operates means that there is also a large amount of spare capacity within the 256 MHz reserved for it. This can be made available for new uses that will interleave (i.e. share) successfully with broadcast transmitters. This capacity is known as interleaved spectrum.

1.10 We have considered the future of both cleared and interleaved spectrum as part of the DDR. Together, they constitute the digital dividend.

1.11 The fundamental reason why this spectrum is so important is its physical characteristics: an exceptionally attractive combination of capacity (bandwidth) and coverage (signals travel further and penetrate buildings more readily). This, in turn, means that it can be used for a very wide range of potential new services. These include additional television services delivered through DTT (whether in standard definition—SD—or HD), local television, new types of mobile broadband, mobile television, wireless home networks and many more.

1.12 This diversity of potential use makes the digital dividend exceptional in our spectrum awards programme.

Our objective

1.13 Our objective for the DDR is to maximise the total value to society that using the digital dividend generates over time. To this end, we have been rigorous in seeking to look at all potential sources of value. This includes the value that each of us derives as a consumer of new services. Critically, it also includes the value that communications services can create for society more widely through their potential for contributing to broad social goals like inclusion and promoting informed democracy.

1.14 This comprehensive approach meets our principal duty, which is to further the interests of citizens and consumers. It is also an approach we have applied consistently, so we have looked at citizen and consumer interests in relation to all the likely uses of the digital dividend, from mobile broadband to more DTT, mobile television and beyond.

1.15 Our work has already stretched over two years. It has included two major rounds of market research, using a variety of techniques to discover the opinions of citizens and consumers on the options for using this resource. It has also included extensive technical research and detailed economic analysis and modelling. Finally, it has included careful consideration of consultation responses and analysis of multiple policy options.

1.16 It is emphatically not our objective to award the digital dividend so as to maximise revenue for the Exchequer. Given our duties, this is not a consideration that we take into account.

1.17 In parallel with the DDR, our extensive work on public service broadcasting (PSB) has helped to identify potential threats to the delivery of public purposes in broadcasting in a timely way and contributed to a wider debate about potential solutions. These will be considered in our second PSB Review. And our ongoing work on access and inclusion is seeking to ensure citizens across the UK have access to services such as broadband Internet access and mobile telephony.

The strategic choice

1.18 The DDR consultation document, published on 19 December 2006, said that we faced a strategic choice in the approach that we took to awarding the digital dividend. We summarised this as a choice between taking a market led approach and an interventionist one.

1.19 Under a market led approach, we would release the spectrum in a way that would allow the widest possible range of technologies and services to be deployed. We would leave it to the market to decide how the spectrum should be used and create flexibility for users to change the use of spectrum over time, reflecting changes in technology and the preferences of citizens and consumers.

1.20 Under an interventionist approach, we would limit the way that the spectrum could be used through regulation. We would select particular uses or users, reserving spectrum for them and excluding others.

1.21 We noted that, historically, spectrum has been managed in a very interventionist way, with detailed regulation controlling who could use spectrum, for what and how. Spectrum policy has been used to pursue many different objectives, including the delivery of social policy goals, by gifting spectrum as an alternative to funding. But we observed that, as scarcity increased, spectrum management was changing around the world, with more emphasis on market mechanisms and flexibility for users and less resort to regulation.

1.22 We said that we had a clear presumption in favour of a market led approach and that this reflected our wider strategy for spectrum management: reducing regulation and making more use of market mechanisms. But we also recognised clearly the risk of market failure: the risk that markets might not deliver the best outcome for citizens and consumers in all circumstances.

1.23 We said that the key question for this phase of the DDR was whether intervening to control the use of the digital dividend was the best way of maximising total value to society. We set out a clear analytical framework for addressing this question, according to which we would consider both the potential benefits and the potential costs of intervening. And we identified two major types of cost: the opportunity cost created by displacing other uses and users and the long term effects of reducing flexibility in spectrum use and blunting incentives for efficiency.

1.24 We invited views on whether this was the right analytical approach, and we included an initial analysis of many arguments that had been put to us about the need to intervene.

1.25 In preparing this statement, we have again examined thoroughly the choices that we face and our analytical framework for assessing them. We conclude that the strategic choice we have identified remains valid and that the analytical framework we have used is robust.

1.26 We also think it is right to retain our presumption against intervening to limit the use of spectrum and in favour of a market led approach. In rapidly changing and converging markets, we think that the market is better placed than the regulator to determine the best uses of spectrum, including the digital dividend.

1.27 We have applied our analytical framework in detail to a wide range of potential uses and looked at many different options for intervening. We now turn to these.

The case for particular uses

1.28 Many stakeholders have suggested to us that we should ensure the digital dividend is used for particular uses or by particular users. Their proposals have included PSB in HD, local television, mobile television, wireless broadband in rural areas, public safety services, healthcare, education, community development, and providing new services for people with disabilities. We have also had many representations that we should protect the existing use of interleaved spectrum for PMSE, particularly wireless microphones.

1.29 The breadth of these proposals reflects the status of spectrum as a basic input into the economy, akin to capital or labour, which can be used in a vast array of ways in downstream products and services.

1.30 We have examined the case for intervening by looking at arguments common to several cases and considering in detail those potential uses that we think are most plausible.

1.31 Our analysis has identified two broad types of concern that are potential sources of market failure. These are:

• the costs of coordination. Some types of spectrum use involve large numbers of small users who act largely or wholly independently of each other. It can be very costly for these users to coordinate their demand for spectrum so that its value can be tested through market mechanisms against other uses and users; and
• the potential existence of broader social value. Some uses of spectrum can bring value to society (e.g. encouraging greater participation in civil society) that goes beyond what individual consumers might be willing to pay for a service.

1.32 We have also identified two points that are particularly relevant to the case for intervening on the grounds of broader social value:

• the need to look at all the possible ways of intervening and to consider whether allocating spectrum is the best mechanism; and
• the need to think about all the possible ways of delivering this value, including not just the digital dividend but other spectrum bands and wireless services and the whole range of wired communications.

1.33 Our conclusions are set out below. We deal first with the scope for authorising use of the digital dividend on a licence exempt basis, which avoids the need for individual licences, and then with possible licensed uses.

1.34 We propose to allow licence exempt use of interleaved spectrum for cognitive devices. Some licence exempt uses are able to coexist successfully with higher power licensed uses. Cognitive radio is a new technology that can detect spectrum that is otherwise unused and transmit without causing harmful interference. It has the potential to support a wide range of uses, including high speed always on broadband. It is particularly suited to operating in interleaved spectrum, where significant capacity is often unused at any one location at least some of the time.

1.35 We see significant scope for cognitive equipment using interleaved spectrum to emerge and to benefit from international economies of scale. But use of equipment in the UK will need to protect licensed users of this spectrum, including DTT and PMSE, against harmful interference. We will not allow cognitive equipment to use interleaved spectrum until we are satisfied on this point.

1.36 We think that allowing licence exempt cognitive use of interleaved spectrum is likely to be justified. Allowing access in this way will overcome the coordination problem they would otherwise face while imposing limited costs on other potential uses. We also think it is likely to encourage more innovation and competition in the provision of services, promoting the interests of citizens and consumers.

1.37 We have decided not to set aside any of the digital dividend exclusively for licence exempt use. The opportunity cost of setting aside spectrum just for licence exempt use would be high, and the additional benefits would be limited given the prospects for cognitive access to interleaved spectrum and the fact that large amounts of spectrum are already available for licence exempt use in other bands.

1.38 We have decided not to hold back spectrum as an innovation reserve. We have considered this idea carefully since the consultation but concluded that the arguments against it are strong.

1.39 Holding spectrum back would impose large costs on citizens and consumers now by reducing the availability of new services. It would also not be a good way of promoting innovation, not least as there would be significant uncertainty over when we should make the spectrum in the reserve available. It would only be beneficial if a high value use of the digital dividend emerged after the award and could not access the spectrum. This is not impossible, but we think it unlikely, not least given that our proposals for cognitive access to interleaved spectrum and allowing licensees to trade spectrum and change its use minimise this risk.

1.40 We see technological change in this area as continuous and indefinite. We do not think there will be a “eureka” moment at which the right future use of spectrum will become clear. We think the best way of promoting innovation is to award spectrum, not reserve it, and to pay particular attention to doing so in a way that will encourage new entry and new uses.

1.41 We have decided to reserve most of the available interleaved spectrum to meet the needs of PMSE users. PMSE is an existing use of interleaved spectrum. It comprises a large and diverse community of businesses, community organisations and individuals. We think that PMSE users would find it difficult to coordinate a bid for access to spectrum, and we think there is a high risk of market failure as result. However, with a careful transition, they can move to accessing spectrum via market mechanisms in the future.

1.42 We will award a single package of interleaved spectrum to a licensee that will act as a band manager. To help PMSE users with the transition to market mechanisms, we will use criteria designed to ensure that the band manager’s interests are aligned with those of PMSE users. The band manager will pay a charge for the spectrum based on Administered Incentive Pricing (AIP) and will be able to earn revenue by charging its customers for access. But regulation will ensure that it has to meet reasonable demand from PMSE users on fair, reasonable and non-discriminatory terms. So long as these obligations are met, the band manager will be able to allow others to make use of its spectrum.

1.43 We have decided that channel 69 should continue to be available for PMSE use throughout the UK on a licensed basis. We will also promote greater licence exempt use of channel 70 for PMSE, in the interests of community users.

1.44 These decisions close the separate PMSE consultation that we launched in June 2007. Shortly, we will also publish detailed information on the availability of interleaved spectrum for PMSE after DSO.

1.45 We have decided to award geographic packages of interleaved spectrum suitable for local television, but we will not restrict their use to this service. We have found some interest in the use of the digital dividend for local television among viewers (through our market research ) and potential operators. But we think that if we release only UK wide packages of interleaved spectrum, local television operators would find it difficult to coordinate a bid. We will respond to the demand that we have identified by packaging some spectrum in geographic lots, based on main transmitter sites serving major towns and cities and areas where local television operators are already providing an analogue service.

1.46 We have identified around 25 possible locations across the UK where there is enough evidence of demand to justify offering such packages. We will be willing to consider other locations if there is persuasive evidence of demand. We will offer one or two packages in each location. Each package should allow, for example, the operation of a low power DTT multiplex carrying several SD channels.

1.47 We have considered arguments made by some local television operators and community media organisations for intervening more extensively, to reserve spectrum exclusively for local television and to award it by beauty contest. We have concluded that further intervention is not justified. It would have a high opportunity cost, displacing other potential users who could also generate high value for society. It would not ensure that local television is economically viable, and it would reduce incentives to use spectrum efficiently. We also think that other delivery mechanisms, such as broadband, might be an attractive alternative for delivering social goals.

1.48 These geographic packages of interleaved spectrum will be auctioned. We plan to begin with the first regions where DSO will take place: Border, Granada, West Country and Wales. We aim to hold the first awards by the end of 2008.

1.49 We have decided not to reserve spectrum to provide more DTT services in SD. The Government reserved almost 70% of the spectrum released by ceasing analogue television for DTT in 2003, before Ofcom took on its responsibilities. A number of stakeholders have pressed us to allocate more spectrum to DTT so that additional capacity will be available for SD services. Some service providers with public policy goals (e.g. Teachers TV, the Community Channel and NHS Direct) have linked this to the difficulty they face in accessing the capacity they would like on DTT multiplexes.

1.50 We recognise that a number of broadcast services, not just PSB delivered by organisations like the BBC, can deliver important public goals. We do not, however, think that allocating additional spectrum is an appropriate response to the issue identified. This is because it would displace other uses and users of the spectrum, including new uses that might be more innovative than additional SD services. This could impose a high cost on society. This is particularly so given that new technologies are likely to increase the capacity of the existing DTT multiplexes and hence allow more services to be provided without using additional spectrum that is in high demand for other uses. We also think that intervening in spectrum allocation is not the right way to address difficulties organisations like these may face in accessing multiplex capacity. This is first and foremost a matter of funding and choices about the best way to reach audiences.

1.51 We will, however, consider the role that services like these can play in delivering broader social value in our second PSB Review.

1.52 We have decided not to reserve spectrum for DTT services in HD. Many organisations and individuals pressed us during the consultation to set aside much or all of the digital dividend to provide DTT services in HD. They argued that, unless we intervened by reserving spectrum, PSB services would not be available in HD on DTT. They said that this would lead to a decline in the value and competitiveness of DTT and that this, in turn, would undermine the future of PSB.

1.53 The DDR consultation document observed that enough capacity would be available on the DTT platform at DSO to offer HD services without needing extra spectrum and that this could be done without having to reduce the number of other services available. Our confidence in this view has grown over the past year. We published detailed proposals on 21 November 2007 for upgrading the DTT platform to introduce new technologies on one of the six multiplexes. This will allow HD services to be offered as DSO is rolled out, starting in the Granada region in late 2009 or early 2010—substantially earlier than would have been possible using the digital dividend.

1.54 Given that there are several options for offering HD services using the 256 MHz of spectrum already reserved for DTT, we do not consider that there is a case for allocating more spectrum to achieve the same goal.

1.55 We have decided not to reserve any of the digital dividend for other services, such as mobile broadband and mobile television. But we will package the spectrum in a way that enables the widest possible range of uses, including additional DTT multiplexes as well as new mobile services.

1.56 We think that mobile broadband could generate significant value for citizens and consumers, but we do not see a case for reserving any spectrum exclusively for this use or for other applications like mobile television.

1.57 We will, however, package cleared spectrum to enable the widest possible range of uses, in particular to meet potential demand for spectrum for additional DTT multiplexes, for mobile broadband and for mobile television. We regard these as likely uses of cleared spectrum. We have seen evidence of interest from both citizens and consumers (through our market research) and from potential service providers.

1.58 We also propose to offer a package of interleaved spectrum in a way that could enable a wide range of new uses, including mobile broadband or extra DTT. This would comprise channels 61 and 62, which are adjacent to the upper cleared spectrum.

Award process and timing

1.59 We have decided to auction cleared spectrum. This reflects our view that an auction is the fairest and most transparent way to award rights to use spectrum and that it is superior to a beauty contest. We think that market mechanisms are the most effective tool available to encourage efficient use of spectrum and should be used unless there is a compelling case to the contrary.

1.60 We have decided to auction channel 36 alongside cleared spectrum. Channel 36 is different because it is currently used for ground based aeronautical radar, a use that will cease at the end of March 2009. Channel 36 will therefore be available UK wide before other cleared spectrum, which will only be fully vacated when DSO ends in 2012. We will allow early use of channel 36 that does not materially degrade analogue television services operating in adjacent channels. We have also considered whether to award channel 36 before the rest of the digital dividend, but there are many options for using this spectrum that would involve combining channel 36 with other frequencies. Auctioning channel 36 separately could impose very large costs by making it difficult to realise these options. We also see few advantages to a separate auction, which would, at best, advance the award by a few months.

1.61 We have decided to auction the packages of interleaved spectrum, except for the package with PMSE obligations, which we will award by beauty contest. We have concluded that, in this latter case, an auction would not be appropriate. Qualitative criteria—notably the technical and managerial ability to meet the needs of PMSE users—will be required to select the band manager, and the intervention required to allow these users to move to a market led approach to spectrum access reduces the benefit of an auction. We recognise that this level of intervention in the use of interleaved spectrum creates risks to the long term efficiency with which it is used. We will mitigate these by charging the band manager AIP, creating an incentive to use the spectrum more efficiently.

1.62 We have decided to award the digital dividend as soon as possible. This will allow citizens and consumers to benefit from new services with minimum delay. We consider that the auction of cleared spectrum could be held in the first half of 2009 but that some of the geographic packages of interleaved spectrum suitable for local television and the package with PMSE obligations could be awarded by the end of 2008.

Issues for further consultation

1.63 The approach set out above will create new opportunities for the UK communications sector to bring an array of new products and services to citizens and consumers, to deploy new technologies and to create new ways of generating and delivering value.

1.64 The significance of this opportunity is immense. Our analysis suggests that the value to citizens and consumers generated by using the digital dividend could be £5-10 billion (net present value—NPV—over the next 20 years). This is not an estimate of auction proceeds.

1.65 But there is a great deal more to be done to turn these opportunities into reality. This statement concludes the first phase of the DDR by making decisions on the strategic choice that we identified last year. There is another, equally important phase to come: the detailed design of the awards and the terms of the licences awarded.

1.66 During this phase, we will consider in detail how awarding this spectrum can best promote competition and innovation in downstream markets and guard against the possibility of anticompetitive behaviour such as hoarding.

1.67 Our experience of spectrum awards suggests that a well designed auction can be a powerful tool for encouraging new entry, new services and new technologies. We will harness experience of spectrum markets around the world to address issues such as the packaging of the spectrum, the options for auction design, the contents of licence conditions and the need for specific measures to protect and promote competition.

International developments

1.68 Spectrum management in the UK takes place within international frameworks set both globally and in the European Union (EU). Two recent developments are particularly relevant to the DDR.

1.69 First, the World Radiocommunication Conference 2007 (WRC-07) agreed to change the international Radiocommunication Regulations to make spectrum currently used for analogue television more flexible, in particular enabling mobile use.

1.70 This has limited direct effect on the UK because agreements with other European countries already give us substantial flexibility. But the indirect benefits of the agreement could be large, opening up the prospect that many more countries will make a digital dividend available for new wireless services. This will help to create global economies of scale for equipment, so reducing prices for UK consumers.

1.71 Second, the European Commission has published a Communication on a common approach to the digital dividend in the EU. This recommends identifying common bands that can be optimised by enabling “clusters” of services using a similar type of communications network: broadcasting, mobile multimedia and mobile broadband. These bands would be planned and harmonised in some form at EU level.

1.72 We recognise that the UK will benefit if other EU Member States release a digital dividend, particularly if they do so in a broadly similar way. But requiring them to replan the frequencies they are using to deliver DSO would significantly delay this process. Discussions in the EU also need to be concluded quickly so that citizens and consumers can benefit from new services as soon as possible.

1.73 We therefore strongly support a non-mandatory approach to harmonising the digital dividend in the EU. This will allow Member States to participate to the extent that they wish while not favouring one use of the digital dividend over others.

1.74 We will contribute fully to EU discussions in the months to come. In the meantime, we believe it right to press ahead with the DDR in the interests of bringing benefits to UK citizens and consumers at the earliest possible date.

Next steps

1.75 We will consult on the detailed design of the digital dividend awards in spring 2008.

1.76 In relation to cleared spectrum and interleaved channels 61 and 62, we hope to publish the information memorandum and auction rules by the end of 2008. The auction itself could take place in the first half of 2009.

1.77 In relation to interleaved spectrum, we expect to auction geographic packages in Border, Granada, West Country and Wales by the end of 2008. We expect to award the package with PMSE obligations by beauty contest to the same timetable. Auctions of the remaining geographic packages will follow in 2009.

1.78 This timetable is subject to factors outside our control, particularly developments in the EU, and may change during the course of the rest of the DDR.