Broadband tops the agenda

Citation

Curwen, P. (2002), "Broadband tops the agenda", info, Vol. 4 No. 2. https://doi.org/10.1108/info.2002.27204bab.002

Broadband tops the agenda

Rearview

A regular column on the information industries

Broadband tops the agenda

Peter Curwen

High-speed broadband access has long been the dream of those consigned to the "world-wide wait". Most countries, as well as the European Union, have either legislated, or are trying to legislate, to bring this about. Unfortunately, a great many would-be subscribers are still waiting and, even when it arrives, broadband still mostly comes courtesy of incumbents.

Given the plethora of terminology in use over the years such as "fibre to the home", it is probably sensible to define broadband by the speed at which data are carried over a network, irrespective of its form. For this reason, the term "broadband" is usefully defined as an "always-on" connection providing a minimum of 128kbps. The current yardstick for a "respectable" broadband connection is 512kbps downstream (for data delivery) and 128kbps upstream (return channel). This can be compared with an upper limit of 56kbps for narrowband, dial-up connections. Broadband is seen as desirable because it permits high-speed access to the Internet. Broadband is almost always delivered to households either by enhancing the capacity of an existing twisted copper pair using Digital Subscriber Line (DSL) technology, or via a cable modem. However, cable networks, unlike the ubiquitous telcos' fixed-wire networks, are highly variable in their geographic coverage. One possible alternative to the above, primarily for business users, is broadband fixed-wireless access (BFWA), and there is also the increasing possibility of broadband access via satellite although this currently cannot compete in terms of price. Finally, businesses can lease dedicated lines to provide broadband connections to their premises. In general, the reality is that newer access technologies are likely to make some inroads where DSL and cable are absent for now, but most companies providing anything other than DSL or cable will struggle to survive.

The bottom line is that the local loop is invariably dominated by an incumbent, originally in almost all cases – and in certain parts of the world, often still – a monopolist owned by the state. Hence, a potential alternative provider of broadband must either buy DSL products at wholesale prices from the incumbent to sell-on or install their own equipment at a local exchange owned by the incumbent – so-called "local loop unbundling (LLU)".

The European Commission estimates that perhaps 6 per cent of European Union (EU) households have a broadband connection, although the likes of GartnerG2 put the figure at only 3-4 per cent, compared to 13 per cent in the USA. However, given the patchy coverage of cable, it is the data on DSL that tell the truest story. In the late summer of 2001 there were roughly 1.5 million DSL lines installed throughout the EU, representing less than 1 per cent of total exchange lines at the time. Belgium surprisingly came out top with 1.6 per cent, followed by Austria. Although four countries provided over 100,000 DSL lines, with Germany well in front with 630,000, only Germany among them exceeded 1 per cent and four countries provided no DSL lines at all. Importantly, all bar 80,000 of the DSL lines were installed by incumbents. In Denmark and Finland, CLECs supplied almost half of the DSL lines, and in The Netherlands the figure was 13 per cent, but everywhere else, including Germany, it was under 5 per cent. In an astonishing nine cases it was zero. The latest figures indicate that roughly four million DSL lines have now been installed, with two million in Germany alone, and the European Commission has praised countries such as Belgium for making significant progress. Nevertheless, the Commission launched infringement proceedings in December 2001 against Germany, Greece and Portugal for failing to implement LLU and did so again in relation to France, Germany, Ireland, The Netherlands and Portugal in March 2002 for failing to offer reduced prices where CLECs seek access to the local loop between the local exchange and their customers.

That incumbents have been the main beneficiaries of LLU is, to put it mildly, perverse. However, it is perfectly logical once one accepts that their obstructive behaviour in respect of matters such as permitting prospective co-locators to enter local exchanges unaccompanied, or releasing space there, has driven most CLECs to lose interest – whereupon incumbents find themselves with a major incentive to promote DSL, especially if faced by a competitive threat from cable. Even where CLECs intend simply to buy wholesale and sell-on, it is an irony that any success in driving down wholesale prices tends to result in incumbents lowering their retail prices to the point at which CLECs cannot compete profitably.

In the UK, for example, incumbent BT suddenly slashed wholesale broadband prices by roughly half to $21 a month in March 2002, implying a retail price of between $35 and $45. This could be compared with an upper limit of $35 in most other EU markets and $35 for a connection via cable in the UK if bundled with voice telephony and pay-TV. BT predicted that it would increase the number of broadband connections from the existing 145,000 – somewhat fewer than the 175,000 supplied via cable at the time – to one million by the end of 2003. However, some analysts doubted even this somewhat undemanding figure, pointing out that it would cost $300 to have a connection installed by an engineer, although self-install would be much cheaper at $125. Interestingly, the two cable companies, NTL and Telewest, responded to BT's roughly comparable monthly charge by choosing to compete on speed rather than price, dismissing BT's 512kbps as "a digital Zimmer frame" and promising 1Mbps before the end of 2002.

In the USA, the so-called Tauzin-Dingell Bill – strictly, the proposed Internet Freedom and Broadband Deployment Act – has been introduced to speed up investment in broadband access by permitting the Baby Bells to provide long-distance data services without first opening up their local markets to long-distance operators as specified in the 1996 Telecommunications Act. However, opposition from long-distance and cable network operators among others is almost certain to prevent the Bill becoming law. In this context it is worth noting that, at the end of 2001, broadband was delivered via cable modems to 7.2 million subscribers – twice as many as chose DSL access. The March 2002 decision by the FCC to classify broadband cable as an "information service", thereby removing any obligation to allow non-discriminatory inter-connection, has played somewhat into the hands of cable incumbents even if AOL-Time Warner, for example, was obliged to open up its cables to EarthLink as a condition for the merger to proceed.

Interestingly, it is South Korea that leads the world in broadband access. It has seen considerable investment in DSL and, although cable access is also important there, most other countries where broadband is progressing are reliant upon well-developed cable networks – as in The Netherlands. In the USA, as noted, the ratio is roughly 2:1 in favour of cable, in good part because cable modems were made widely available at an early stage, but despite the FCC decision the ratio is nevertheless expected to move steadily in favour of DSL. On the whole, the fact that cable is relatively easy to install provided there is a connection box in the road outside a property, and that modems can be readily bought in retail outlets, has given it an initial advantage over DSL. DSL has been difficult to roll-out on a large scale because subscribers have to be unhooked individually from the main distribution frame, has involved heavy up-front expenditures and in its most common format, asymmetric DSL (ADSL), can only operate at up to 2Mbps. Hence, the likes of NorthPoint and Covad in the USA have been driven to seek Chapter 11 bankruptcy protection. Admittedly, the finances of cable companies such as UPC, NTL and Telewest are also in tatters, but the fact that the cables have already been laid means that there will always be someone willing to take them over. Needless to say, incumbents rolling out DSL are generally much better placed financially and hence can be expected to overhaul cable provision in the medium term while retaining control over the vast majority of the market for DSL access. If nothing else, inertia among existing subscribers will work in their favour although it does have to be said that profit margins in DSL provision are less than compelling if capital is expensive and hard to come by, as is currently the case.

The future of broadband is closely tied into the services that will become available such as streaming audio and video and interactive TV, although these will generally need faster speeds than ADSL can currently provide. Hence, the fibre-optic loop will probably need to come much closer to domestic properties than it does at present for a mass market to develop, thereby reducing the cost of making the final connection. Meanwhile, a combination of incumbents' resistance, insufficiently tough regulation, over-concentration on LLU rather than competitive wholesale prices, recession and indifference among householders has set back the widespread introduction of broadband by several years.