THE IMPACT OF THE 3G MOBILE LICENCE AUCTIONS

Martin Fransman

Introduction

Towards the end of 2000 both Business Week and The Economist ran front-cover articles on, in the latter's words, 'The Wireless Gamble'. Essentially, they were pointing to the huge investment - running into hundreds of billions of dollars - required in third-generation (3G) mobile communications in the face of highly uncertain demand for 3G services and products. The necessary investment in 3G was approximately doubled by the auctioning of licences. The prices of these licences surprised everyone and added subsequently to the woes of the licencees, mainly incumbents, whose share prices were hit towards the end of the year by the combination of a number of factors.

The aim of this part of TelecomVisions.com is to initiate an analytical debate regarding the effects of the 3G mobile licence auctions. The main reason for a debate is that, as will shortly be shown, there are some fundamental disagreements among participants in the industry and analysts regarding the effects of these auctions. It is the hope that although these disagreements may not be finally resolved as a result of the discussion here, at least some of the main sources of the disagreements may be clarified.

Several important resources are provided here to facilitate this debate. These take the form of several articles that are freely downloadable in the Articles Zone of TelecomVisions.com. Participants are invited to go to Workshop 3 in the Workshop Zone to add their own analyses, information, and comments.

To begin with, TelecomVisions.com has been given three articles by Paul Klemperer, Professor of Economics at Oxford University, who was the principal auction theorist advising the U.K. government's Radiocommunications Agency, which designed and ran the recent U.K. mobile-phone licence auction. The first two papers, 'Auction Theory: A Guide to the Literature' and 'Why Every Economist Should Learn Some Auction Theory', are primarily concerned with the theory of auctions. The third paper, 'What Really Matters in Auction Design', deals more pragmatically (and less theoretically and technically) with the issues involved in designing auctions and preventing collusive, predatory, and entry deterring behaviour in the auction process itself.

The latter paper also touches on the question of the market structure that is generated by auction licences that create new industries. Regarding the German auction Klemperer concludes that "Ignoring the issue of market structure was the most serious flaw in the design of the German mobile phone licence auction - though the outcome of the auction was very good [i.e. licences were awarded to six firms], this was good luck, not good design." (p.10) The behaviour of the participants in the German auction also leads Klemperer to ponder on their rationality: "Surprisingly, the dominant incumbents first pushed the prices up to almost UK levels, but then gave up and ended the auction before pushing any of the weaker firms out --it is hard to construct beliefs about opponents for which this is rational behaviour." (p.11)

Following Professor Klemperer's papers, there are two further contributions to the debate. (These two papers were published in info, vol. 2, no. 4, August 2000 and are included here with the kind permission of its editor, Colin Blackman.) The first is by Martin Cave, Professor of Economics at Brunel University, and Tommaso Valletti, Lecturer in Economics at the London School of Economics. They support the auction of 3G mobile licences against the attacks of those who have argued that licence fees will have the effect of raising prices to consumers and delaying the roll-out of infrastructure. Of particular interest is their spelling out of the reasoning behind the argument that the licence fee (and its size) will have no impact on the price of 3G mobile services. The second of these papers is by David Harrington, Director General of The [UK] Telecommunications Managers Association (TMA). He argues, on the contrary, that the UK's 3G spectrum auction "imposes unnecessarily high barriers to entry for bidders and brings with it the inevitability that customers will pay higher prices".

SOME OF THE KEY QUESTIONS

What are the objectives of Government?

Since governments have been in control of the allocation of 3G mobile licences it seems sensible to ask what their objectives are. In general it would seem that governments have been motivated by at least two objectives. First, to create a dynamic, competitive mobile industry that will serve the interests of consumers well by selecting the 'fittest' available competitors. The second objective is to appropriately price a public asset, namely spectrum.

Governments appear to have considered two alternative means to achieve these ends. The first is an auction while the second has become known as a 'beauty contest'. The latter involves some government-controlled process of selection of licencees from amongst applicants who must meet specified criteria. (Klemperer's papers show that there are many different alternatives under the auction option, but these will not be considered here.)

The fact that there are two objectives raises the first questions. Is there a trade-off between the objectives? And, if there is, how is this trade-off to be resolved? For example, if government deems that an 'appropriate price' is the maximum price that it can get for the sale of spectrum, will it be jeopardising its objective of establishing the most dynamic mobile industry possible? If so, how?

It would appear that some governments have come to the conclusion that maximising the price of the licence may indeed inhibit their simultaneous attempt to create a highly dynamic mobile industry. This, for example, seems to be the reason why Japan and Finland, both with highly dynamic second-generation mobile industries, have opted for the 'beauty contest' route and why other governments have chosen hybrid forms. The problem, however, is that if the choice of these government was motivated by the belief in a trade-off between the two objectives, they have not, so far as can be seen, made explicit their understanding of the trade-off, how it works, and what its quantitative effects would be. This is unfortunate since, in the absence of explicit argument and details, it is not possible to sensibly debate the issue. One hope is that TelecomVisions.com might be used to analyse and discuss this question further. (If you have any further information or analysis on this issue, please go to Workshop 3 in the Workshop Zone.)

A further set of questions arise in connection with the relative efficiency of auctions and governments as governance mechanisms, that is as ways of governing the processes that are needed to produce the outcome of a dynamic, competitive mobile industry. For example, those who support auctions frequently stress the advantages of explicitness and transparency that auctions lend themselves to. On the other hand, they tend to be suspicious of government processes and the ability of allegedly knowledgeable government officials to make the right decisions. Once again, however, these judgments can be analysed further. You can make your contribution in Workshop 3.

How will auctions affect the price consumers have to pay for 3G services?

This is a key question on which many of the other important questions hinge. If, for example, the higher the price of auctioned licences, the higher the price that consumers will have to pay for 3G services, then there may be an important trade-off between the two objectives. For if maximising the price of licences leads to higher than otherwise prices of 3G services, this may slow the rate and speed of diffusion of these services and have other negative knock-on effects, for instance on the suppliers of complementary inputs, such as equipment, content and applications. Furthermore, there may be global distributional consequences. Companies and countries with greater reliance on beauty contests may gain at the expense of those that are more dependent on the outcome of auctions.

For this reason it is important to understand how the auction price will affect the price that consumers will have to pay. The problem here, however, is that there appears to be a fundamental disagreement between two camps. It is this fundamental disagreement that needs to be further analysed. Even if it is not possible to bring about an eventual consensus, the very least that should be aimed for is an analytical elaboration of the causes of the disagreement.

The 'Water Under The Bridge Hypothesis'

In the one camp are those that adhere to what may be called the 'water under the bridge hypothesis'. According to this hypothesis, the price that the mobile operator must pay for the 3G licence is a fixed or sunk cost. This fixed/sunk cost, however, cannot be recouped from customers. Mobile operators, therefore, although they may not like it, will have no option but to ignore the fixed/sunk cost in their decisions about providing for, and selling, mobile services. In effect, they will collectively regard the additional fixed/sunk cost imposed by the licence fee as 'water under the bridge'. As Cave and Valletti put it (see their article in the Articles Zone),

"As far as the firm and its competitors are concerned, the licence fee is an irrevocable sunk cost. When deciding how to set prices, the firm rationally only takes account of its own forward looking costs and revenues and the likely behaviour of other firms. Since the licence fee is a sunk cost for all firms, it falls out of the pricing equation for all of them. Hence the size of the licence fee does not affect prices." (p.349 in the original info article.)

The problem, however, is that, clear though this statement is, those in the opposing camp do not appear to agree either with the logic of the 'water under the bridge hypothesis' or the implied characterisation of the way firms actually behave under the assumed circumstances. And those in the opposing camp include some very important players in the global 3G mobile industry. One example is Dr Keiji Tachikawa, President of NTT DoCoMo, Japan's largest mobile operator and the world's largest mobile Internet access company with its highly successful i-mode service. According to Dr Tachikawa,

"Operators will have to pass on the added cost [i.e. the cost of the 3G mobile licence] to consumers. This could be a hurdle for the spread of 3G." (Business Week, October 16, 2000, p.25)

Similarly, David Harrington, Director General of The [UK] Telecommunications Managers Association (TMA) argues that the UK's 3G spectrum auction "brings with it the inevitability that customers will pay higher prices". (See Articles Zone.)

What is happening here? Why is there apparently such fundamental disagreement on such a key question? Assuming that both camps genuinely believe what they are saying, why have they come to such contradictory beliefs? Logically, they must have different assumptions, theories, or models in mind - even if these are implicit rather than explicit - about the way in which firms behave under the conditions postulated. But what are these assumptions, theories or models? This question must be answered if we are to get to the bottom of the puzzle of why such fundamental disagreement exists.

A possible starting point, it may be suggested, is for those supporting the 'water under the bridge hypothesis' to communicate in a more effective way to their opponents precisely why they have come to the conclusion they have. The following argument, one suspects, will not have a great deal of meaning for mobile operators grappling with their business decisions: "When deciding how to set prices, the firm rationally only takes account of its own forward looking costs and revenues and the likely behaviour of other firms. Since the licence fee is a sunk cost for all firms, it falls out of the pricing equation for all of them."

Is this decision process, and the outcome that it leads to, true of the real firms in the current mobile industry? Or is it only true of rational firms? And if it is only true of rational firms, how do we explain the behaviour of supposedly irrational or non-rational firms, like the major mobile companies bidding in the German auction, the behaviour of which, as we saw earlier, Paul Klemperer found difficult to reoconcile with rationality? If the argument is meant to apply to real firms in the current mobile industry then it is surely necessary to explain why they only take account of "forward looking costs and revenues", ignoring the additional fixed/sunk cost imposed by the licence fee. What factors or forces will lead them to ignore these significant costs which they will have to pay? Similarly, it is necessary to explain why the fixed/sunk costs should "fall out of the pricing equation for all of [the firms]".

Clearly, the answers to these questions are dependent on some rather complex theories. Their very complexity makes it difficult to communicate logically but verbally to practitioners lacking the same theoretical and analytical background. However, there is no valid reason in principle why such an explanation cannot be provided. In view of the significance of the underlying question - namely the impact of the licence fee on the price consumers will pay for 3G services - it seems important to try and elicit a more satisfactory explanation.

For if it is true that whatever the licence price there will be no effect on the price of 3G services - that is that this conclusion is consistent with the actual behaviour of the mobile operators doing business in today's mobile markets - then there are extremely important consequences. To begin with, it will eliminate one important possible source of trade-off between the two assumed objectives of government. By maximising the price of the public resource, spectrum, the government will not be sacrificing the development of a dynamic competitive mobile industry through the mechanism of a lower and slower rate of diffusion of mobile services caused by higher prices for mobile services. And if this is the case, the 'auction countries' will not find themselves at such a disadvantage compared to the 'beauty contest countries' (at least as far as this mechanism is concerned).

For these kinds of reasons it is important that this fundamental disagreement between the two camps is further analysed. (Your own contributions can be made by going to Workshop 3 in the Workshop Zone.)

What will be the impact of the auctioned licence fee on mobile operators' cost of borrowing?

Even if the 'water under the bridge hypothesis' is correct and the additional fixed/sunk cost imposed by the auctioned licence fee is ignored and therefore does not directly affect the price of 3G services, it does not follow that there is no possible indirect impact on price. One possible line of reasoning proceeds as follows: higher licence fees force the mobile operator to borrow more; greater borrowing (on top of already high borrowing to finance capital investment and mergers) increase gearing ratios; higher gearing ratios increase the 'perceived risk' of lending; this leads to lower credit ratings by agencies such as Standard & Poor's and Moody's Investors Service; this increases the coupon payments that mobile operators have to make to investors holding bonds; in turn, their borrowings fall; there may also be sale of other assets by the mobile operator in an attempt to improve credit ratings; these sales, however, may negatively impact on the operator's returns from economies of scale and scope; furthermore, many large mobile operators, going simultaneously to the markets for funding, together with the desire of lenders to spread their risks across sectors, may lead to higher costs of borrowing for all telecoms companies. The result may be higher prices for mobile services.

These effects may also have distributional implications amongst mobile operators as those who are more dependent on 'auction country markets' lose relative to those that rely more on 'beauty contest country markets'. Against this potential negative set of effects, however, account would have to be taken of the possible positive effects of state expenditure of the revenues raised by the auction.

Cave and Valletti, however, argue that there is a "genuine but small risk" of this happening (see Articles Zone; p.349 in info). They examine the argument that "large licence payments erode the finances of operators, requiring them to assume debts thus reducing their credit rating. The effect is to increase the cost of borrowing for all purposes, and hence to raise the forward looking costs."

The reasoning on the basis of which they conclude that although the risk is genuine, it is small, is the following. Firstly, "High licence payments do not in any way influence the systematic risk associated with a mobile phone company (the degree to which its returns are linked returns to the stock market as a whole)." Secondly, they acknowledge that there is "potentially a small increase in bankruptcy risk" as a result of higher borrowing "which may have the effect of modestly increasing the cost of capital for smaller companies". However, "the effect on telecommunications giants would be tiny."

Once again, these arguments are extremely important and merit further analysis and discussion, both theoretically and empirically. For example, what additional impact has the high auctioned licence fee had on the credit rating and cost of borrowing of mobile operators who have won European 3G licences (i.e. in addition to the borrowing they would in any event have made to finance network and other related investments and mergers)? These operators are, for the most part, telecomunications giants. Is it correct that higher borrowings have no impact on "systemic risk"? If it is true, why have credit rating agencies downgraded the rating of many large mobile operators (to the extent that this has been due to borrowing to finance licence fees)? etc. (Your contributions on these kinds of issues can be made by going to Workshop 3 in the Workshop Zone.)

How quickly will new 3G services be rolled out?

Will new 3G services be rolled out more rapidly in 'auction country markets' than in 'beauty contest country markets'? One possible hypothesis is that to the extent that there is a similar degree of competition in both kinds of countries, 3G services will be developed at the same speed. (In part the speed will be determined by the ability of 2.5 generation technologies and networks to offer substitute services to those of the third generation. While the main advantage of third generation networks is the greater bandwidth, some mobile services are not improved with higher bandwidth availability.) However, an alternative hypothesis may suggest that 'auction market countries' will perform better in terms of roll-out. The reasoning is that in addition to the pressures of competition, mobile operators in these countries face extra pressure to repay their high borrowings with the revenues brought in from selling 3G services.

Which type of country will perform better and why? What is the evidence, theoretical and empirical? You can share your views by going to Workshop 3 in the Workshop Zone.

Will dynamic 'Scandinavia Effects' operate?

One of the most interesting characteristics of the 'mobile revolution' is the extent to which Scandinavian companies, such as Ericsson and Nokia, have managed to use to their advantage the early lead in mobile communications established by the Scandinavian region. (For a documentation of the origins of this lead see M.Fransman, Evolution of the Telecommunications Industry into the Internet Age, Articles Zone, and the articles by Maureen McKelvey quoted therein.)

This phenomenon raises a number of important questions. To begin with, what precisely are the mechanisms that have allowed Scandinavian companies to benefit from the effects of early leadership that may be generically summarised as 'dynamic increasing returns'? How sustainable is this leadership? If other countries, apart from Scandianvia, are able to get ahead in 3G mobile, are they also likely to be able to benefit from dynamic increasing returns? Specifically, will Japan as the first large country that will roll out 3G mobile services, and Japanese companies, be able to assume global leadership? If so, will it be in all areas of this industry, that is mobile network operation, mobile phones and equipment, mobile content and applications development, and dynamic mobile consumption, or only in some?

Please make your contributions on these important questions by going to Workshop 3 in the Workshop Zone.

Martin Fransman