Introduction Subscription-free Internet Service Providers 'Free' or 'unmetered'? Price comparisons with Europe

Regulation and the pace of change Barriers to unmetered telephone tariffs Conclusions References and Annexes

The Cost of Accessing the Internet In March 1999, following our approach, the Parliamentary Select Committee on Trade and Industry invited us to make a formal response to the OFTEL Memorandum to that Committee's Inquiry into Electronic Commerce. With the Committee's permission, we present here the full text of our response. It is also available as a 400KB archive in 7-Zip format including the response and both annexes in Microsoft Word, PDF/A and OpenDocument formats. © March 1999, Campaign for Unmetered Telecommunications

1. Introduction

1.1 The Campaign for Unmetered Telecommunications has followed with great interest the Department of Trade and Industry Parliamentary Select Committee Inquiry into Electronic Commerce. We are pleased to provide input to the Inquiry, even at this late stage.

1.2 This document is primarily a commentary on the Memorandum [1] Oftel submitted to the Select Committee upon which further oral evidence was taken on 2 March 1999. We do not have knowledge of the content of Oftel's oral evidence, but we find certain statements in the Memorandum to be misleading both in their general tone and in some of the facts presented. We do not consider ourselves qualified to comment upon Part I of the Memorandum (on the issue of encryption) and any comments we make will be limited to Part II, paragraphs 14 et seq, in particular the parts which affect home users and SMEs.

1.3 We would, however, like to add some general comments on other aspects of the cost of telecommunications and accessing the Internet in the UK. Although this document is not intended as an attack upon Oftel staff we are extremely critical of Oftel's approach and methodology in many instances: we invite this Committee to obtain explanations from Oftel on some points, even after initial consultation in preparation for the e-commerce Bill has been completed.

1.4 Oftel's statement in paragraphs 14 and 15 that 'recent' claims have been made about the cost of UK Internet access being too high and that, as a result of those claims, it is seriously looking into the issues raised, requires elucidation.

1.5 The Campaign for Unmetered Telecommunications was founded in February 1998. Its objectives include focussing dissatisfaction about the cost of accessing the Internet from home in the UK. That dissatisfaction had been growing steadily since at least summer 1994, when mass participation in the Internet by people in this country became a reality. That is hardly 'recent'.

1.6 Our research indicates that Oftel only started looking at the cost of Internet access to home users and SMEs with any seriousness in December 1998 when it issued a Consultation Paper on access to broadband services [2]. We enclose a copy of our reply to that Consultation Paper as Annex 1. More recently, Oftel was forced to intervene in a dispute between BT and certain other telecommunications operators (particularly Energis) and, on 10 March 1999, it issued a further Consultation Paper [3], referred to in paragraph 23 of the Memorandum. The principal conclusion of the Consultation Paper was that Oftel sees no reason for change. This is short-sighted and, we suggest, will continue to retard the use of e-commerce in the UK.

2. Subscription-free Internet Service Providers

2.1 Oftel's cost analysis in paragraphs 20-24 attempts to compare Internet access costs between the UK and other countries, particularly the US. While Oftel gives the cost of subscribing to an Internet Service Provider (ISP) as a possible barrier, too many reports to enumerate from around the world have established that ISP subscription costs are not perceived by users as an issue. Report after report indicates that, in the US, lack of interest is the primary reason given by people who do not yet use the Internet: in the UK, lack of access to the appropriate technologies (computers and infrastructure) and the cost of telecommunications are invariably the first and second reasons why individuals and companies do not use the Internet. The UK has one of the most vibrant markets in the provision of ISP services to home users and SMEs, and Internet Magazine's monthly review of the market currently lists some eight hundred national and local companies offering dialup Internet access in the UK.

2.2 The entry in autumn 1998 of Freeserve into the UK ISP market (and the subsequent launch of many similar subscription-free ISPs) was perceived by some as the death-knell to existing subscription-based ISPs, but there still exists a wide variety of subscription-based ISPs, whose charges for residential customers range from between £5 to £18 per month, and who evidently have a continued user base. We are only aware of one subscription-based ISP (Bogomip) having failed following the launch of subscription-free services. Although subscription-free ISPs have almost certainly attracted a great many new Internet users, no research appears to have been undertaken to verify this: anecdotal evidence we have collected suggests that a large number of subscription-free ISPs' customers are existing Internet users who have moved from subscription-based ISPs, often keeping the subscription-based ISP's account as a backup or because they have to maintain their previous email address. This is not surprising, as subscription-free ISPs' services differ in no great respect from those offered by subscription-based ISPs, and people would obviously prefer to obtain a service notionally free of charge rather than pay for it. We have noted that a certain proportion of subscription-free ISP customers do, however, return to their subscription-based ISPs due to dissatisfaction with the service offered: also that many people 'do the rounds' of subscription-free ISPs, often trying several in turn before finding one whose service they deem satisfactory.

2.3 We are not in a position to supply statistics on the magnitude of the shift from subscription-based to subscription-free ISPs, as membership and usage statistics are considered commercially sensitive by all ISPs we have asked, but we recommend that the Committee calls representatives of the ISP industry, both from subscription-based and subscription-free ISPs, to offer evidence on their experiences - in camera if deemed necessary.

2.4 Whilst Oftel 'welcomes the increasing competition brought on by subscription-free services', we would draw the Committee's attention to the fact that this competition has been introduced not as a result of regulation of the ISP market - in which neither Oftel nor any other regulatory body has any interest - but by that market taking advantage of, in particular, decreasing capital costs (hardware and packaged software) and improvements in technical support and marketing. The telecommunications market which Oftel is required to regulate has shown no such innovation.

2.5 In the US, subscription-free ISPs are currently unviable because of a vastly different regulatory regime and different telecommunications pricing structures. Oftel quotes America On-Line's (AOL's) unlimited access for a monthly subscription of $19.95. This is roughly in line with other US ISPs (the spread we have noted is between $15 and $25 for a monthly subscription), and is also more or less the typical rate for subscription-based ISPs in this country. The international market has quite evidently found its level. By Oftel's own figures, Internet participation in the US far outstrips that in the UK, and the fact that US residents do not have the option of subscription-free ISPs is quite obviously no barrier to Internet participation there.

2.6 In the US, the ISP charge is on top of unmetered telephone tariffs (the telephone costs for dialling in to US ISPs are paid by a monthly subscription to the local telecommunications provider, rather than metered by the minute as in this country): such unmetered tariffs change the way subscribers use the Internet. A UK market which depends upon ISPs funded from metered telephone charges can hardly be the way forward to extensive use of online services. Psychological as well as cost considerations come into play: when given the option of large downloads, or browsing complex Web sites such as online catalogues or reference sources, UK (and European) users have to make choices, quite possibly unconscious choices, at every step to save money. US users are not constrained by online minutes being clocked up on the telephone bill so can, and do, make wider use of Internet facilities.

2.7 The fact that most people using UK sites will be paying by the minute to access them has the secondary effect of many UK Web designers not availing themselves of advanced design features, thus stifling the innovation of which the British have always been proud.

2.8 An International Telecommunications Union report as long ago as 1995 [4] already recognised future affordability and effectiveness of Internet use as an issue, stating 'To develop entertainment-based services, the telecommunication industry will need to move away from its traditional usage-based pricing structures towards flat-rate tariffs'. If the UK Government wishes to increase interest in online activities, the question posed by this statement should be addressed as a matter of urgency.

2.9 Feedback to our Web site, and personal enquiries, have shown that home users and SMEs in the UK regularly face monthly telephone bills in excess of £100 per month for Internet access. The UK's liberalised telecommunications industry should by now have developed a competitive market in providing local services but, as Oftel itself admits in [2], 85% of all UK telephone traffic is still originated on the BT network and home users or SMEs have little effective choice in their supplier or choice of tariffs.

2.10 A separate but related issue is the provision of wholesale Internet connectivity. As the vast majority of content is held on servers in the US, ISPs in the UK and other European countries are obliged to purchase dedicated two-way transatlantic bandwidth, at considerable cost, in order to offer their customers a reliable service: US-based providers do not feel any need to install such bandwidth for their customers. This is a chicken-and-egg situation: until the number of active Internet users in the UK reaches a critical mass justifying production of content relevant to other parts of the world as well as the UK, there will always be a cost disadvantage. On the other hand, while this cost disadvantage exists, usage will not reach that critical mass: in fact, UK-based content providers offering services to other parts of the world are finding that they have to offer these services on US-based servers instead of, or in addition to, servers based in the UK in order to attract international custom. The only exit from this vicious circle we can see is to ensure that Internet access should be made as ubiquitous and as affordable as possible in the UK to ensure that US customers demand two-way transatlantic bandwidth.

3. 'Free' or 'unmetered'?

3.1 In paragraph 24, Oftel inadvertently establishes the crux of the matter. For some reason, Oftel continues to use the term 'free' when referring to unmetered telephone tariffs. We question its motives in doing so. Individuals and organisations campaigning for the introduction of unmetered tariffs have long petitioned for the telecommunications industry and others to use a less emotive term than 'free'. The word 'free' is so loaded ('there's no such thing as a free lunch', 'buy three get one free', 'you can't get something for nothing', etcetera) that in this context the misleading impression is given by Oftel that telecommunication companies are being invited to 'give something away'. We quote paragraph 26:

3.2

Efforts at comparing UK prices 'have tended to be based on incomplete data, and have not taken account of these recent developments [ subscription-free ISPs] in the market. For instance, an article in the FT on 19 January 1999 used figures from Datamonitor to imply that Internet access in the UK was the highest in Western Europe. However, these comparisons did not incorporate any of the subscription-free access products that are now a major part of the UK market. If these products were included they would be likely to improve the UK figures.'

3.4 We also note that, following the success of subscription-free services in the UK, a subscription-free service was introduced in Germany by Deutsche Telekom. However, it was ruled unlawful by a German court on the grounds of being anti-competitive [5]. Under European directives, telecommunications provision and other services must be separately accounted for in charging structures, so all-in charges are not admissible. We are moved to wonder why no action has been taken on the same basis against BT and Cable and Wireless Communications: in response to the entry of subscription-free ISPs to the market they introduced BTClickFree and Cable and Wireless Internet Lite respectively as their own subscription-free ISPs. No other cable operators have followed Cable and Wireless Communications' lead; again, we wonder why.

3.5 The previous paragraphs beg two questions which Oftel has not even formulated, never mind attempted to answer, in [3]:

3.5.1 If subscription-free ISPs are able to fund the cost of maintaining all necessary services to provide Internet connectivity out of the termination charge payable to them under the current regulatory regime, then surely overall metered charges for telephone usage, no matter the operator, must be enormously inflated?

3.5.2 Conversely, BT is able to provide national and non-geographical telephone connections, taking only its slice of the Number Translation Services payment in return, without adversely affecting its profits (apart from BTClickFree, subscription-free ISPs usually use network providers other than BT). Surely this reflects badly upon BT's remaining metered charges?

3.6 These are fundamental questions which affect all telephone users, not just those who access the Internet through their telephone line, and should be addressed as a matter of the greatest urgency.

3.7 We consider Oftel's statement in paragraph 28 that metered charges have dropped substantially in the UK over the last few years, and the implication that price drops are due to its own influence, particularly disingenuous. The UK telecommunications industry has had the benefit of over a decade of liberalisation, whilst Germany and France undertook telecommunications de-nationalisation as recently as 1996 and 1997 respectively. UK charges not having dropped substantially would have indicated complete market failure, not merely the failure of Oftel's regulatory powers.

3.8 A separate issue, but one which should be noted, is that BT does not have the tremendous financial burden its German competitor has of bringing the outdated infrastructure it inherited in Eastern Germany up to the standards of the rest of its network. Deutsche Telekom appears to be able to fund such extensive upgrading whilst decreasing its existing prices. Furthermore, BT is rightly proud of its achievement in being the first national telecommunications operator in the world to have upgraded its network to be all-digital, among the avowed benefits from which are reduced costs and increased capacity and flexibility.

4. Price comparisons with Europe

4.1 On obtaining tariffs from the Web sites of various telecommunications operators we note that, although Deutsche Telekom and France Télécom tariff information is immediately available, to find BT tariff information one needs to delve deep into a complicated site structure. This is suggestive of BT misunderstanding the possibilities of e-commerce - instead of being an example to other companies, it proves itself not to understand the way its customers use its site.

4.2 Whilst a direct comparison of UK charges to those prevalent in the US is, in our opinion, not possible as charging structures between the two countries are completely different, such direct comparisons with our principal European competitors are reasonably straightforward.

4.3 With regard to the US situation we attach as Annex 2 a summary prepared for us by Thomas Long.

4.4 We now examine the principal elements of telephone access costs charged by the dominant operators in the UK, France and Germany. We limit ourselves to these operators, as the vast majority of home and SME customers in these countries have little choice, and while a certain amount of 'shopping around' may result in slightly lower prices, such choices are limited.

4.5 Table 1 gives the publicly-quoted tariffs by British Telecom, France Télécom and Deutsche Telekom as on 23 March 1999, applying inter-bank currency conversion rates as at that date:

	Line installation	Monthly line rental	Low-user line rental	Min charge	Local peak	Local off-peak	National off-peak	Per min to Italy	Per min to US
British Telecom	£116.33	£8.92	£4.34	5p	4p	1.5/1p	3p	25.97p	20.86p
France Télécom	305.66FF (£31.16)	78FF (£7.95)	39FF (£4.01)	74c (7.5p)	28c (2.9p)	14c (1.42p)	50c (5.1p)	18.5c (18.86p)	200c (20.39p)
Deutsche Telekom	100.87DM (£34.49)	24.82DM (£8.49)	9.09DM (£2.72)	12Pf (4.1p)	8Pf (2.7p)	4.8/3Pf (1.35/1p)	6Pf (2p)	72Pf (24.6p)	72Pf (24.6p)

Table 1 - Basic telephone tariffs

4.6 As Oftel points out, these charges are generally comparable. Two unduly large discrepancies are immediately obvious, however:

4.6.1 The installation of a new BT telephone line costs almost four times that of its French and German counterparts, despite many years' advance on a liberalised market and the existence of direct local loop competition from cable companies, this not being the case in France or Germany where the incumbent national operators have effective monopolies.

4.6.2 Calls at peak times cost over 25% more from BT, completely negating the Memorandum which claims that BT's prices are the lowest in the market!

4.7 All operators offer roughly similar discounts on the tariffs in Table 1 to residential customers, amounting to approximately 10% to 30% off metered charges depending on the customer's own preferences and the option to pay a further subscription fee. France Télécom also provides a subscription-based (50FF monthly) discount scheme which offers 50% discount on charges to Internet dialup numbers. The three companies offer slightly different definitions of 'peak' and 'off-peak' which we will not enumerate as this would only serve to lengthen this document.

4.8 Differences become quite obvious when we compare the cost of digital (ISDN) connections as per Table 2:

	Line installation	Monthly line rental	Peak (per min)	Off-peak (per min)
British Telecom	£283.83	£27	4p	1.5p
France Télécom	597FF (£61.83)	169FF (£17.50)	28c (2.9p)	14c (1.42p)
Deutsche Telekom	100.87DM (£34.49)	51.44DM (£17.59)	8Pf (2.7p)	4.8Pf (1.35p)

Table 2 - ISDN tariffs

4.9 All three operators offer various charging packages for digital lines. The prices quoted in Table 2 are for residential users; business users can choose various packages dependent on usage, including higher installation charges and lower monthly rental, or lower installation charges and higher line rental, together with various levels of call allowances (a predefined number of call minutes included in the line rental before per-minute charges are levied). Unlike its equivalents in France and Germany, BT charges to convert an existing analogue line to digital; for home users this charge is £138.33 (the same charge as installing a new analogue line) and, for business users, it is £149 although, just before this report was submitted, BT announced reductions to these charges.

4.10 We do not present a comparison of leased line costs. Firstly, these are so expensive as to be of minor interest to SMEs and home users: secondly, packages are very complex so direct comparison is difficult. However, we draw the Committee's attention to the conclusions of the recent, and DTI-sponsored, International Benchmarking Study [6], which placed the cost for basic leased line installation at UK $616, France $133 and Germany $58. We do not feel that these figures require any further comment as they are so eloquent of themselves.

5. Regulation and the pace of change

5.1 We would submit that the primary cause of the UK's disfavour compared to the US (and indeed, France and Germany) lies in the regulatory structure of the UK telecommunications industry. A similar comment can be made about the regulation of all UK utilities, and we welcome the Government's plans on reform currently under review. UK regulatory bodies see their main function to be the regulation of abstract markets rather than the defence of consumer interests. In the US, for instance, any interested party can gain access to full cost and provision charges for any telecommunications company on signing a non-disclosure agreement. Debate on the true cost (and therefore charging structure) of telecommunications in the UK is not possible: all interested parties are obliged to accept Oftel's determinations, which in turn depend on cost analyses presented by BT without the benefit of any independent audit.

5.2 As noted above, Oftel has recently launched a debate on the provision of broadband services in the UK. Although we welcome this debate, as does the whole industry and many other interested parties judging by the volume of responses to their Consultation Paper Oftel have made public, we note that widespread rollout of xDSL services remains in BT's future, and no cable company is yet offering cable modem services in the UK although the first rollout is close at hand. Both xDSL and cable modems have been available in North America for some considerable time, for home users and SMEs, for payment of a monthly subscription fee and a one-off installation charge.

5.3 Deutsche Telekom announced in a press release on 16 March 1999 [7] that '… by the end of the year it plans to have 100,000 customers using T-DSL in 43 local networks around the country. […] Starting in July, the new ISDN/ADSL package will be available to residential customers for a basic charge of DM 98 [£33.50 per month]. […] Even as it now introduces ADSL, Deutsche Telekom is already working on the next generation of xDSL technologies. Symmetrical 2 megabit/s services based on HDSL (high bit-rate digital subscriber line) will even become available this year.'

5.4 On 26 March 1999 Telefónica SA, the incumbent telephone operator in Spain - a country not usually considered to be as technologically advanced as the United Kingdom and with a considerably lower GDP - announced rollout of ADSL this year [8]: 'Telefónica will reach 30 percent of all Internet users with its ADSL network this year, regulators said. The rest of the country will have access to the network by 2001'. We draw the Committee's attention to the fact that this announcement was made by the Spanish regulator, and that the service will be available for a monthly subscription fee of about £20 plus a one-off installation charge.

5.5 We also point out that Digital Powerline, which uses the electrical mains network for the transmission of broadband services, is now in the process of being deployed in other countries, specifically Sweden and Italy [9]. To the best of our knowledge no UK utility is even near to deploying what is both a wholly British invention, developed by Nortel, and a remarkable technical achievement.

5.6 Innovation in tariffing in the UK is rare. Videotron UK, a subsidiary of the Canadian cable operator Groupe Vidéotron Limitée, was the first cable operator to introduce unmetered local off-peak voice and data calls in this country (several operators, including Telewest and Eurobell, offer similar tariffs for voice calls only). When Videotron UK was merged with three other cable operators to form Cable and Wireless Communications (a UK-owned company) the 'Videotron tariff' was immediately put under threat although vigorous campaigning by those who enjoy the tariff has, for the moment, ensured its survival. 'Videotron-style' tariffs for local data calls (for example, those previously offered by Nynex and Telewest) have been successfully taken away from users in the past.

5.7 Furthermore, Kingston Communications in Hull has recently introduced an unmetered tariff, Karoo Xtra [10], for users of its own ISP (Kingston Internet): unmetered voice calls over Kingston Communications' own network have always been part of its service. We note that Kingston Communications are a monopoly in the area in which they are licensed to operate.

5.8 Finally, we note that the first cable modem rollout in the UK is being performed by NTL, a US-owned company, within the next few months.

5.9 Paragraph 28 appears to be astonishingly defensive of the state of the market Oftel is supposed to be regulating. It makes value judgements verging on an advertisement of BT's prices! There is also a factual error, namely 'with BT's discount schemes such as Friends and Family, users can get nearly 3 hours of connection at the weekend for just £1'. This sentence has caused a certain amount of consternation among Internet users: applying all possible BT discounts (some of which require a subscription fee to be paid before the discount on metered charges is available), the lowest rate available is 0.65p per minute or 39p per hour, which means that three hours of Internet use do not cost what Oftel states. We find it telling that one of the few 'facts' Oftel presents in its document is actually incorrect.

5.10 In any event, this 'lowest possible BT rate' is only available at weekends (midnight Friday to midnight Sunday). How many home users or SMEs can be, or want to be, online only between those times? During working hours, the cost of access remains 4p a minute (or slightly above 3p if one applies discounts).

5.11 In later parts of the Memorandum Oftel relies very heavily on an OECD survey which is still unpublished and is thus unavailable for scrutiny. The basic methodological assumption the survey makes is that users will spend 20 hours online per month. This represents a very low level of Internet use - how much useful work can be done in 40 minutes a day? An assumption of at least twice that usage level would be far more realistic: we would suggest 60-80 hours as the level above which SMEs could make any meaningful use of e-commerce. However, the ITU report referred to above states unequivocally that the eventual aim should be for businesses to be able to afford a permanent connection.

5.12 The UK is supposed to have one of the most modern networks in the world, operating in the most advanced and liberalised market in Europe. That the UK is 'more expensive than a number of other OECD countries' in any situation is telling. As for services from cable operators, many people do not have that option and, we suggest, many will never have it. Oftel imply that 40% of the UK population will have no access to cable networks by 2001: many operators have halted expansion of their telephony services due partly, in some cases, to underestimated demand, and partly to their current priority of implementing digital television.

5.13 As Oftel itself admits, despite a dozen years of liberalisation with their own regulatory framework in place, over 85% of all UK calls are initiated from the BT network.

6. Barriers to unmetered telephone tariffs

6.1 In paragraphs 32-40, Oftel lists the barriers it sees to the introduction of telephone tariffs including unmetered calls and its views thereon. We would like to examine each of these in turn.

6.2 Who pays the costs incurred?

6.2.1 Before that question may be answered or even discussed, the true value of these costs must be established. This cannot be done until full and reliable real costs and pricing data are made available by telecommunications operators to interested parties. Oftel base their knowledge of costs and pricing on figures submitted to them by BT. These costs are not independently audited and are not available to any other interested parties. As it would seem that BT is able to sustain its network, even with the recent growth in Internet use, on its retention from NTS-rated calls, we suggest that the cost estimates Oftel have accepted from BT are vastly inflated. Some independent analysts have, using as much information as they have available and some guesswork, calculated that the marginal cost of connecting and maintaining a local call is one fifty-thousandth of a penny, and remarks from the US (see Annex 2) are suggestive of very low costs. BT claims a much larger figure, of course, but provides no independently validated figures to substantiate its claim.

6.3 In the US, the costs of free local calls tend to be met either through a higher fixed rate charge (sometimes accompanied by a limit on usage) or by cross subsidy from other tariffs, which are therefore higher than they would otherwise be.

6.3.1 Our research has indicated that the 'higher fixed rate charge' is not much higher. In fact, in many cases, the fixed rate charge is lower than that levied by BT or cable operators. Annex 2 explains the structure of US 'cross-subsidy' in sufficient detail for us not to need to reiterate those arguments. More importantly, Oftel neglects a concept which it has been established to champion: choice. Unlike the UK, where almost everyone is forced to use wholly-metered tariffs, most US operators offer a range of bases of charge reflecting different usage levels and uses of the telephone. Light users may subscribe to wholly-metered tariffs with a relatively low line rental. Medium users might use a half-and-half scheme, with unmetered calls available in a restricted local area and the rest metered. Finally, for heavier users, there would be the option of a relatively high line rental with unmetered calls available over a wider geographical area. Our US contacts indicate that only a small proportion of users, generally those on low incomes, take up wholly-metered tariffs.

6.3.2 As for 'cross subsidy from other tariffs', we note that Oftel is again making generalised statements without substantiating them with any facts. Prices for long-distance and international calls from most US operators are roughly comparable to BT's. We also note that UK organisations who resell bandwidth from competitive trunk network and international operators are able to offer long-distance and international calls at rates typically a fraction of BT's. For example, for calls to the US, BT's published charges are 20.86p per minute whereas most resellers offer such calls for 6p or 7p per minute. Presumably such resellers' charges are indicative of the costs involved, yet BT is not obliged to explain why its charges are so much higher. The only implication one can draw is that the retail price of such calls is subsidising other services offered by BT.

6.3.3 In any event, we note that US long-distance and local providers are not generally the same companies, so direct cross-subsidy is improbable if not impossible. Furthermore, there is fierce competition in the US at both local and long-distance level. At local level the dominant operators, Incumbent Local Exchange Carriers or ILECs, compete with Competitive Local Exchange Carriers or CLECs encouraged to set up in business as a result of the 1996 Telecommunications Act: we note that some CLECs, such as Covad, only provide broadband services over the ILECs' local loops. At long-distance level there is a plethora of deals, as a look at any Web page carrying US advertising banners demonstrates. We have had contacts from people in the US who have changed local or long-distance carriers several times in a year to take advantage of the best offers.

6.3.4 Furthermore, BT claims that its BTClickFree service will be paid for by advertising. BT's revenues from this source will not make inroads, for some time, into the capital expenditure involved in setting up BTClickFree, and we suggest that advertising revenues may never recoup that expenditure - ISPs supported solely by advertising have been launched, and failed, in other countries. Surely BT is cross-subsiding BTClickFree from call charges or other sources? Our understanding of EU directives would suggest that such cross-subsidy is in fact unlawful.

6.3.5 On the issue of advertising, we observe that no Internet company depending solely on online advertising revenue has yet posted an operating profit: all profits are paper profits predicated on share price movements. Also, if BT considers advertising to be the main source of revenue for its BTClickFree service, we find it strange that it requires users to dial in by use of NTS-rated numbers rather than true local access numbers. By allowing true local access, BT would, we presume, reduce its own costs and maximise subscribers' exposure to advertising. As it does not do so, we can only assume that online advertising is not the cash cow it is perceived to be.

6.4 Recent price comparisons carried out by Eurodata showed that long distance call prices in the US tended to be higher than in the UK. This suggests that those customers who are significant users of long distance calls are subsidising those who use a high proportion of 'free' calls.

6.4.1 Although we dispute the first sentence, as explained earlier, and note the unproved assertion in the second, we fail to see the relevance of these statements. All US telephone subscribers have an array of long-distance carriers from which to choose and, in most states, incumbent operators are legally obliged to inform their customers that such offers exist. With carrier pre-selection, use of such services becomes transparent. In any event, the reverse could be said to happen in the UK, as discussed previously. Furthermore, long-distance telephone calls across the US cover distances that UK telecommunications operators would class as international and charge accordingly.

6.5 Another model is where higher fixed rate charges compensate for 'free' local calls. This model, if no other tariff options are available, may cause problems for lower income families for whom the line rental charge is typically a significant element of the phone bill.

6.5.1 In the UK in general, and certainly for BT, there is only one tariff - based on paying by the minute, although at different rates at different times of day. It is clear that lower-income families who wish to make significant use of the telephone must be disadvantaged, this being one of the issues the current Inquiry is designed to address. As mentioned above most US operators offer a variety of tariff options, including relatively low line rentals, and we again invite Oftel to substantiate their claims. It would have been helpful if Oftel could at least have stated which report they were quoting.

6.6 Some proponents of 'free' local calls argue that the marginal cost of a call over the network is very low.

6.6.1 These arguments are based upon recent advances in technology and, in particular, on the UK network being up-to-date and all-digital. How low marginal costs, or indeed any costs, really are is not known. An analyst from Magenta Systems claimed in a Mail on Sunday interview [11] in mid-1998 that '... no one outside the companies knows the real costs'. The same article paraphrased Malcolm Matson, founder of COLT Telecom Group plc: 'The true cost of connecting a call remains a mystery' and 'hidden in this information vacuum there is evidence that domestic customers should have free calls and pay through flat-rate line rental charges'.

6.7 Additional usage during peak times causes congestion unless additional capacity is installed.

6.7.1 Concerning this, we suggest that a relevant feature of home Internet access is that its peak times are different from those experienced by business. From personal observation the 'home peaks' appear to be 2pm and 10pm on Sundays and 11pm on weekdays; these cannot be the same as 'business peaks'. Given that, surely home Internet access exploits network capacity installed to cope with 'business peaks' that would otherwise be underused?

6.7.2 Furthermore, congestion is not necessarily a bad thing; overcoming it sparks off technical advances in telecommunications hardware and software. Many of the remarkable inventions of the last few years, such as photonics, dense wave data multiplexing (which allows enormous capacity increases in transmitting data over existing fibre-optic cables) and IP telephony, which benefit or will benefit everyone, would not have been developed without existing or projected pressure on networks.

6.7.3 In any case, installing 'additional capacity' on a digital network should not be a costly enterprise: compare analogue or partially analogue networks where Strowger electromechanical exchanges, which contain thousands of moving parts and require much effort in maintenance, were the norm in the UK and are still found in other countries.

6.7.4 Local telecommunications operators in the US, with older and busier networks than those the UK, are able to report profits while offering unmetered local calls 24 hours a day, 7 days a week, and the Federal Communications Commission and other authorities consider such calls an integral part of the US pricing structure. The only reasons for UK telecommunications operators to be unable to offer such unmetered calls must be unwillingness to invest in infrastructure (if investment is indeed required; as explained earlier, we do not know whether it is or not), inefficiency, or greed.

6.7.5 As for inefficiency, we reference a study by McKinsey [12] which suggests that UK telecommunications operators are half as efficient as those in the US.

6.7.6 The issues raised by the last few paragraphs lie squarely within Oftel's remit, yet it appears unable to shed any light whatsoever on them.

6.8 In the US, there are examples of network difficulties arising where 'free' calls are offered with no limit on usage and there is thus no incentive to 'log off'. Communications International have estimated that, whereas the average Internet call in the UK is about 25 minutes, in the US it is about 2.5 hours. Such long connection times have created congestion and prevented other calls from being carried. Quality of service in some areas of the US has declined quite significantly and it has been known, in some instances, for calls to emergency services to not get through.

6.8.1 It is disappointing, but to be expected, that telecommunications operators would make such unsubstantiated statements. When Oftel joins in we must question its motives. We have searched the Internet widely over the past year and have written (without any replies or even acknowledgements) to every UK telecommunications operator. We have found only one verifiable instance (with times, places and effects readily available) of such problems. It occurred in the US in 1997 on the launch day of AOL's flat-rate Internet subscription service. These problems were occasioned, furthermore, by AOL underestimating demand and were nothing to do with telecommunications operators; AOL quickly overcame these teething problems, now have 14 million US subscribers and are teaming up with US telecommunications operators to bring broadband services to their subscribers. It should be demanded of Oftel to provide details of any 'instances of emergency calls not getting through'.

6.8.2 As highlighted by various consumer-oriented television programmes in the last few months (including BBC Watchdog and Channel Four's The Mark Thomas Comedy Product; also Private Eye and BBC News Online [13]), a real problem with the UK emergency services has been Cable and Wireless Communications' handling of its emergency services database which has led to fifty 999 calls each month proving untraceable. We are not aware of Internet use having any bearing on these problems.

6.8.3 There are endless statistics available on average Internet call lengths, almost invariably with no indication of the methodology used to derive them: we suggest that AOL, being an ISP themselves (the largest in the US and until Freeserve's birth, also the largest in the UK) and thus in a position to make direct measurements, are probably more authoritative than most sources, and they quote average Internet call lengths as being 16 minutes in the UK (with metered access) and 55 minutes in the US (with unmetered access) [14].

6.8.4 Given the multiplicity of sources and lack of information as to the reliability thereof, we believe that all that can safely be said is that those with unmetered access spend a few times as long per Internet session as those with metered access.

6.8.5 Although US Internet users indeed have 'no incentive to log off', as Oftel states, average online sessions there amount to 55 minutes, or 2½ hours, or whatever survey figure is to hand; they are not infinitely long. Despite the lack of incentive, US users have proved themselves to be responsible and appear to remain online only for as long as they require. In any case, sessions can be managed by the ISP; for example, the maximum call length to Kingston Internet appears to be seven hours.

6.8.6 As a side issue, perhaps the OECD basket-based figures Oftel uses should be recalculated with respect to Oftel's quoted level of Internet usage for the US. With unmetered tariffs, the cost of access in the US and some other countries remains constant whether 20 hours (the OECD assumption) or 75 hours (on the basis of 2½ hours access per day) - or indeed any number of hours access per month - is assumed, whereas in the UK metered tariffs result in the cost of access rising monotonically with time.

6.8.7 Furthermore, a report by Economics and Technology Incorporated in January 1997 [15], prepared as input to a dispute between local and long-distance telecommunications operators in the US, states unequivocally that, firstly, 'Internet use is not responsible for disproportionately increased PSTN costs' and, secondly, that 'Internet and other data communications users are major sources of increased [telecommunications operator] revenues'. Also, 'Use of the public telephone network for Internet and on-line service access is not out of proportion to the subscriber access lines that have been installed to support such use'. In its Introduction, the report states that: 'The increase in data communications traffic has produced additional revenues for the local exchange carriers that far exceed their costs in accommodating that traffic. The long-term solution for accommodating increased data traffic lies in the stimulation of competition and in the deployment of appropriate data-friendly network technologies, and not in the imposition of per-minute access charges for use of the current voice-oriented circuit-switched network.'

6.9 For those companies offering primarily local service the provision of 'free' local calls could act as a disincentive to further investment in the local phone network. This is because these operators would never be able to recover the cost of their investment through revenues from local calls. Thus, if they are unable to recover the costs of 'free' calls from other services these operators are put at an overall commercial disadvantage.

6.9.1 Empirical evidence shows that investment by US local telecommunications operators is not only stable but increasing and the range of services being offered is constantly being expanded. Yet again, we do not know the true costs of provision in the UK, and doubt Oftel's grounds for certainly in this respect. Again, Malcolm Matson of COLT Telecom Group plc sums up the situation in another Mail on Sunday article [16]: 'The phone companies have take the world to the cleaners by pretending that network capacity is a scarce resource that needs to be rationed by pricing'. Being Managing Director of a telecommunications operator himself, he is perhaps in a very good position to appreciate the argument.

6.10 OFTEL believes the consumer is best protected by competition, and the UK is now seeing growing competition from a variety of carriers.

6.10.1 Competition is being felt in the business and long distance market. The majority of the population making the majority of the telephone calls (by Oftel's figures, three out of four calls are local) does not benefit, as the effect of a dozen years of a competitive telecommunications market has been to shrink BT's market share by as many percentage points. This is not competition working: rather, it indicates the existence of an effective cartel and regulation more concerned with an abstract market rather than with services.

6.11 If BT were to offer 'free' local calls, OFTEL would want to be satisfied that this was not designed to damage such competitors. It is worth recording that in the US there is little competition to local carriers offering service to residential customers. In the UK, by contrast, cable operators will pass 60% of households by 2001.

6.11.1 We disagree with 'in the US there is little competition to local carriers': see the description of ILECs and CLECs above.

6.11.2 The implication of this paragraph is that 40% of the population will not have the benefits of limited competition (BT, as the incumbent operator, plus a single cable operator) by 2001. We suggest that such competition will never be seen in areas where cable companies do not project immediate returns on investment. And surely much of the purpose of the Government's various projects to increase Internet use and awareness is to help those who are not currently in a position to make full use of technological advances?

6.11.3 With interconnect charges metered by the minute, competition and innovation in tariffing is severely limited in scope. Metering is passed all the way down the chain from operators to carriers to subscribers, narrowing the range of tariffs that can be offered to subscribers. When Cable and Wireless Communications was formed, it petitioned Oftel to allow for flat-rate arbitrage of interconnect charges (in other words, bandwidth, rather than being paid for in arrears by the minute, paid for in advance by negotiations between operators to buy fixed amounts). Oftel refused to budge after representations made by BT.

6.11.4 Oftel should ask whether or not, in general, US local telecommunications companies and Kingston Telecommunications, which is an anomalous regulatory situation in the UK, offer effective and valued solutions to their customers. In general, they do. The effect of competition on pricing has been that cable operators' tariffs are almost invariably only slightly lower than BT's, colloquially 'BT minus 10%'.

6.11.5 As an aside, we believe that, if a company cannot compete, it should not be propped up by use of price controls or deliberate distortions to the market. We are less worried than OFTEL about 'damage' to BT's competitors.

6.12 Oftel is keen to see the introduction of innovative tariffs by BT and other carriers to meet the needs of customers. Oftel would welcome new tariff arrangements that meet the needs of customers. These might include, for example - packages in which 'free' calls were provided with a limit on usage and upfront fixed fee, or special cheaper rates for data traffic or prices based on use of the network (capacity charging). But it is for the operators, not Oftel, to come forward with such offerings.

6.12.1 Oftel's keenness appears directly proportional to the self-interested representations made to them by those they are supposedly regulating. Indeed, some operators offer innovative packages which involve some element of 'free' calls - for example, Cable and Wireless Communications offer a limited number of 'free' minutes, based on the amount spent on metered calls, per month and Kingston Communications offer charge-capped calls to all numbers within their own network, but such packages are often inadequate (being almost invariably off-peak only) or hedged with restrictions.

6.12.2 We believe that a lack of innovative tariffs is because regulation is keeping pace neither with the Government's desire to make Internet access more broadly available nor with changing patterns of telecommunications usage. Refereeing negotiations and disputes between operators is crucial, but a regulatory regime which takes a direct interest in services as well is needed. From Consultations it has conducted in the past we can determine no sense of urgency in Oftel's approach to any regulatory matter. We sincerely hope that the situation never arises that a company fails entirely due to Oftel not dealing with some issue quickly enough - if that situation has not in fact arisen already.

6.12.3 Regulation should encourage the setting up of true local ISPs (such as exist with great success in the US) where high-bandwidth data traffic is kept within high-capacity local telephone networks. At the moment, with ISPs usually accessed through non-geographic numbers (in effect, national calls charged at a local rate), data calls compete with voice calls for capacity on trunk lines. Surely this is technologically, logically and financially absurd?

7. Conclusions

7.1 In paragraph 41, Oftel claims that it 'would in general welcome the introduction of tariff packages that might make access even more cost-effective for Internet users'. If there was a truly competitive market, a variety of choices would already be available to fulfil the wishes of a varied customer base. In the half-decade since the Internet came of age in the UK, the only true innovation which the telecommunications and ISP industries have introduced has been the subscription-free ISP. The market has not adapted to reflect the Internet's arrival or to take full advantage of the possibilities it offers, but has closed in on itself to protect vested interests.

7.2 Oftel's role must be altered so that its main task is to stimulate innovation in services, something it has singularly failed to do until now. The lack of urgency in Oftel's regulation of ISDN and leased lines and its wait-and-see attitude towards techniques such as xDSL and cable modems bodes ill.

7.3 In many branches of telecommunications the UK is a world leader in innovation. Having been one of the first countries to develop competitive cellular and digital services (and a deregulated infrastructure), the UK should be able to bring similar innovations in land telephony, cable services and Internet technologies, not to mention Digital Powerline, to its own markets sooner than anyone else can to theirs. Yet it is not doing so and, as the recent announcements in Spain and Germany illustrate, it is lagging even within the EU.

7.4 The reasons for this lag must be determined as a matter of extreme urgency and solutions found just as urgently. Our own observations suggest that the current regulatory regime has much to answer for. Each and every regulatory action should have one aim: to do the best for UK plc.

References and Annexes

[1] Trade and Industry Select Committee Inquiry into Electronic Commerce: Memorandum by OFTEL
http://www.oftel.gov.uk/publications/1999/info_super/tisc0399.htm

[2] Access to Bandwidth: Bringing Higher Bandwidth Services to the Consumer
http://www.oftel.gov.uk/publications/1995_98/competition/llu1298.htm

[3] OFTEL Consultation Paper on the Relationship between Retail Prices and Interconnection Charges for Number Translation Services
http://www.oftel.gov.uk/publications/1999/pricing/nts0399.htm

[4] Building information infrastructures
http://www.itu.int/ti/wtdr95/c5.htm

[5] German State Court Rejects Deutsche Telekom's New Pricing Scheme for T-Online in Response to AOL Europe Complaint
http://www.aol.co.uk/press/releases/1999/pr990312.html

[6] Moving into the Information Age: an International Benchmarking Study, 1998
http://www.isi.gov.uk/isi/bench/mitia/index.html

[7] Dawn of a new online age
http://www.telekom.de/english/news/journal/031699.htm

[8] Spain to get flat-rate Net access
http://www.news.com/News/Item/0%2C4%2C34304%2C00.html?dd.ne.txt.032%206.09

[9] inter alia Sydkraft and Tele2 sign agreement
http://www.nor.webdpl.com/press/990219sydkraft.htm

[10] How will I be charged for my Karoo account?
http://www.kingston-internet.net/charges-main.html

[11] Mail on Sunday, 2 August 1998

[12] Driving Productivity and Growth in the UK Economy
http://www.mckinsey.com/features/uk/index.html

[13] Phone company 'endangered lives'
http://news.bbc.co.uk/hi/english/uk/newsid_222000/222917.stm

[14] Daily Mirror, 4 March 1999

[15] Mail on Sunday, 28 February 1999

Annexes

[1] CUT Response to the OFTEL Consultation Document 'Access to Bandwidth: Bringing Higher Bandwidth Services to the Consumer': Option Zero, the Path to Divestiture
http://www.unmetered.org.uk/reference/resp_080399.htm

[2] Comments on the US Situation by Thomas Long
http://www.unmetered.org.uk/mythbusters/mythbustersus.htm