Contents

Executive Summary

Chapter 1 Introduction

Chapter 2 Price comparisons

Chapter 3 Service roll out

Annex A Exchange rates and VAT rates

Annex B DSL and cable modem prices by country (available to download as a .pdf file, for which you will need adobe Acrobat reader

Executive Summary

Scope

This International Benchmarking report covers Digital Subscriber Line (DSL) services, providing high speed telecoms connections to consumers in the UK, France, Germany and the USA over the local copper loop. These services are relatively new (particularly in Europe) but are expected to become increasingly important in giving consumers access to a range of services such as fast internet access and video on demand.

The analysis and drafting of this report have been carried out by OFTEL. The models developed by OFTEL used to generate the results contained in the report have been audited by an independent consultant, Analysys Ltd, who confirm the following:

• The models are logically sound (ie they correctly implement the methodology described in this report); and
• The tariff data has been checked and has been correctly entered and interpreted.

This study is a follow up to the study for DSL services carried out by Analysys Ltd for OFTEL, published in April 2000 (the "April 2000 study").

Data has been collected for a range of tariff packages available to residential and small/medium sized businesses for major operators and service providers in each country as at October 2000. Data collection has been carried out for OFTEL by consultants Tarifica Ltd, with some additional data provided by Analysys Ltd (from their DSL pricing service, www.dslpricing.analysys.com).

Services on offer in different countries have been compared on the basis of price and bandwidth (ie speed of connection) both downstream to the consumer and upstream from the consumer. This study only considers DSL services providing internet access which is the main application for DSL at the current time. Other applications, such as "Video on Demand", have not been considered.

In order to compare prices across countries, a price index has been constructed based on the average of the lowest priced packages offered by two different operators/service providers in each country. This forms the basis for drawing conclusions regarding price levels.

The bandwidth provided with a particular service is a key feature of the service "quality", and all price comparisons are presented in the context of some measure of bandwidth. However, bandwidth is not the only indicator of "quality" and care needs to be taken in interpreting the results.

Other key aspects which have been considered are:

• A comparison of the spread of prices available (a wide spread may imply that consumers need to be careful in their choice of operator, or that there are important non-price factors which are not picked up in the analysis); and
• A comparison of the range of services (in terms of the bandwidth offered) which are available.

Points to bear in mind

There are a large range of operators and service providers offering services to consumers (particularly in the US). It is not possible to cover, in this study, all packages on offer. While it is believed that the sample chosen is fully representative and covers the range of offers available by country, the possibility that consumers have access to additional offers outside the range calculated for the sample used in this report cannot be excluded.

The objective of this report is to provide a comparison between countries for a range of consumer types. Prices are based on a snap shot as at 12^th October 2000. Although providing a sound overall picture, the comparison of companies within a country should not be taken as representative for an individual consumer. It is particularly difficult to establish "value for money" when services differ in important respects such as bandwidth, and details of the service bundle.

The extent of service availability is also relevant in considering the cross-country comparisons: some services are only available currently in certain geographic areas. Within those areas, customers with very long lines may not be able to receive the service.

There are separate industry funded initiatives and reports which seek to help UK consumers be better informed purchasers of telecommunications services (such as individual enquiry based price comparisons - www.phonebills.org.uk; fixed and mobile quality of service indicators - www.cpi.org.uk for fixed and www.oftel.gov.uk for mobile figures; and information for small businesses about using telecoms and the internet - www.TelecomsAdvice.org.uk).

This report is not part of that body of work as it seeks to compare position between, not within, countries and it is not, therefore, intended in any way to be a buyer's guide.

Results for retail residential DSL and cable modem services

Table 1 shows the results for the price level, price index (relative to the UK) and (downstream) bandwidth provided for residential services. All these measures are constructed using the average of the two lowest priced residential offers (from different operators/service providers) in each country. Results are presented both with, and without, cable modem offers included.

Table 1: Results for price level, price index and (downstream) bandwidth for residential broadband services

From Table 1 it is seen that:

• If cable modem services are excluded, then prices in Germany and the US are around 28% and 20% cheaper respectively than in the UK, while prices in France are around 9% more expensive;
• If cable modem services are included, then prices in France are around 8% cheaper than in the UK, while prices in Germany and the US are over 20% cheaper;
• Excluding cable modem offers increases the costs in all countries, ie generally one or both of the two lowest priced services are cable modem rather than DSL services. In the UK, cable modem and DSL prices are closer than in some other countries. Accordingly, the exclusion of cable modem services improves the relative UK position;
• The bandwidth provided in the US is much higher than in Europe.

Results for retail business DSL and cable modem services

Table 2 below shows the results for the price level, price index and bandwidth provided for business services. Note all these measures are constructed using the average of the two lowest priced business offers (from different operators/service providers) in each country. Results are presented with cable modem offers included.

Table 2: Results for price level, price index and bandwidth for business broadband services (cable modem services included)

Note: DS = downstream, GM = geometric mean, ie squareroot(downstream bandwidth x upstream bandwidth)

From Table 2 it is seen that:

• For the most basic services available to business (ie where the business has no minimum bandwidth requirement):
• prices in Germany and the US are around 34% and 20% cheaper respectively than in the UK; and
• prices in France are considerably more expensive than in the other countries - but provide significantly higher bandwidth.
• For business services where a minimum geometric mean bandwidth of 500kbit/s is required:
• Prices are widely spread between countries;
• UK prices are close to prices in France, are about twice as much as those in the US, but less than half of those in Germany; and
• The bandwidth in the UK is about the same as in France, but less than in Germany and the US.
• For business services where a minimum geometric mean bandwidth of 1000kbit/s is required:
• Services are only available in Germany and the US; ( While both France and the UK offer services with a downstream bandwidth of at least 1000kbit/s, these services are asymetric and the geometric mean bandwidth is less than 1000kbit/s) and
• Prices in these two countries are very different, with Germany much more expensive than the US.(Comparisons are based on BT's IPStream products as these appear to be the most closely comparable to the Covad {USA} and QSC {Germany} services)

Results for wholesale DSL services

Comparisons of wholesale prices can only be made on a limited set of data. Based on the information available for one company from three countries (BT, QSC and Covad), UK wholesale prices are similar to those elsewhere for services with similar bandwidth, though the range of bandwidths offered is wider in Germany and the US.

Roll out

Collecting detailed information on current and future availability of services is difficult. The data that has been collected suggests that, for DSL services:

• by the end of 2000, the availability of DSL services in the UK will be behind that in Germany, but ahead of that in France;
• availability and take up in the US is greater than in the European countries covered in this study; and
• information regarding future roll out plans is too uncertain to enable conclusions to be drawn.

Chapter 1

Introduction

OFTEL’s aim is to provide the best possible deal for telecommunications customers in terms of quality, choice and value for money through effective competition. International comparisons of telecoms services are one way of assessing the achievement of this aim.

This International Benchmarking report covers Digital Subscriber Line (DSL) services, providing high speed telecoms connections to consumers in the UK, France, Germany and the USA over the local loop. These services are relatively new (particularly in Europe) but are expected to become increasingly important in providing consumers with access to a range of services, such as fast internet access and video on demand.

OFTEL intends to carry out a market review for access to broadband services (and hence the DSL market) in 2005. It is, therefore, appropriate to establish a basis for assessment well in advance of that date.

This study focuses on comparing the cost for different types of users of DSL services providing internet access. Cable modems represent an alternative broadband service to DSL. Cable modems provide the user with similar functionality to DSL and, hence, cable modem service prices are also included in the analysis.

Services differ in the amount of bandwidth (both downstream to the consumer and upstream from the consumer) that is available. The price comparisons presented are, therefore, viewed in the context of the bandwidth provided.

Results are presented for 3 market segments: residential, business and wholesale. For business and wholesale markets, results are further split into low, medium and high bandwidth services. However, it is noted that there is not necessarily a clear dividing line between residential and business services. A significant number of home workers and small businesses may well find an entry level service, aimed primarily at residential consumers, is adequate for their needs.

As well as cost, it is also important to consider issues such as current availability, take-up and future roll out plans. Detailed information for these areas is, however, difficult to obtain.

There are three subsequent chapters:

1. Comparison of price (Chapter 2);
2. Availability/roll out (Chapter 3); and
3. Conclusions (Chapter 4).

A detailed list of all packages considered are set out in Annex B ( only available as a .pdf file to download) to this report.

The following countries, and number of service providers/operators, are included in the analysis:

• France (4 DSL and 3 cable modem operators/service providers);
• Germany (6 DSL and 1 cable modem operators/service providers)
• UK (7 DSL and 2 cable modem operators/service providers); and
• US (9 DSL, of which 3 are incumbent local operators, and 3 cable modem operators/service providers).

The study covers the services offered by a range of different operators and service providers in each country. For each operator/service provider, the range of packages on offer (including residential, business and wholesale products) have been identified, and data on pricing and service details has been collected.

The objective of this study is to provide a cross country comparison, rather than to compare individual telecommunications providers. However, in order to compare prices between countries, it is necessary to analyse the prices of individual services from a range of providers within each country. While the results may relate to specific services or providers, these should be seen as representative of the country, rather than being viewed as a consumers’ guide.

This study is a follow up to the study ‘International Benchmarking of DSL Services’ carried out for OFTEL by Analysys and published in April 2000. This study has been expanded to include a comparison of cable modem prices.

The April 2000 report included a section providing an overview of those aspects of the regulatory framework relevant to the provision of DSL services, and how those regulatory factors affected the market for service provision. This information has not been fully updated for this report. However, in order to set the context for the remainder of the report, the following section provides some brief comments on the status of local loop unbundling in each country and how this relates to the DSL services currently on offer.

Local loop unbundling

DSL services are provided over the local loop. The owner of the local loop can provide DSL services by installing additional equipment both at the consumer site, and within the local exchange to which the consumer is connected. Competitors to the incumbent can provide DSL services, by one of the following means:

• Building a new local loop plus DSL investment;
• Purchasing wholesale DSL services from the incumbent or wholesale service provider; or
• Leasing the local loop from the incumbent, through local loop unbundling, plus DSL investment.

In general, the first of these options is unattractive due to the high cost and the advantage enjoyed by the incumbent in terms of economies of scale. Service providers may wish to use a combination of the second and third options to provide widespread coverage. However, the third option will provide greater scope to provide a wide range of services and the flexibility to respond to customer needs. It is expected that, in the future, competition will develop on the basis of the third option (ie it is dependent on the process of local loop unbundling).

Table 1.1 (below) provides a summary of the current position in the countries considered in this report in relation to local loop unbundling and the competitive provision of DSL services.

Table 1.1: Local loop unbundling and competitive provision of DSL services

*A "resale" product can be sold directly to an end user, while a wholesale product can only be sold to a service provider.
** Based on the data collected for the incumbent operators covered in this study.
Source: data collected by Tarifica and Analysys (www.dslpricing.analysys.com).

The detailed results presented in the remainder of the report need to be viewed in the context of Table 1.1. From this table it is to be expected that the range of services available might currently be greater in Germany and the US (where LLU is underway) than in the UK and France (where LLU is being introduced).

Future benchmarking reports on DSL services will reflect the effect of local loop unbundling in the UK and France on price and range of services.

Chapter 2

Price comparisons

Introduction

This chapter focuses on price and bandwidth comparisons for DSL and cable modem services.

Services have been defined in terms of their target market. Services have been split into three categories:

• Services aimed primarily at residential consumers;
• Services aimed primarily at business consumers; and
• Wholesale services which are provided, and sold, to residential and business retail customers by a third party service provider.

A difficulty in comparing DSL and cable modem prices is that services differ in respect of key parameters which define "service quality". The most obvious factors which differentiate services are the down and up-stream bandwidth which define the rate at which information can, in principle, be passed to and from the consumer site respectively. The price comparisons, therefore, need to be presented in the context of the available bandwidth (or "speed").

There are, however, a number of other parameters which, in practice, define the speed actually experienced by the consumer and, hence, which may be of equal importance as bandwidth. A key parameter is the contention ratio, a measure of the extent to which bandwidth is potentially shared by consumers. This will be important in determining the bandwidth that a consumer in practice has access to, especially at popular times. The results, therefore, need to be viewed and interpreted with caution.

DSL services are typically provided in conjunction with internet access and may include a range of "bundled" products such as e-mail and web space, a number of static and/or dynamic IP addresses, the ability to connect more than one user etc. These other features included in the "service bundle" will also affect the price. This is seen explicitly for some operators/service providers who offer a selection of services, all with the same down and up-stream bandwidth but with different prices reflecting optional "add-on" features. The approach taken in this report has been to focus on the lowest priced packages which meet given bandwidth requirements. However, packages from different operators/service providers will vary in the details of what is provided. It has not been possible to explicitly adjust for this, but it is not expected that the impact on the overall country comparisons is significant.

The remainder of this Chapter considers:

• The methodology used to compare prices;
• The results obtained for residential, business and wholesale services.

Methodology

The discussion of methodology covers:

• Construction of a consistent measure of price;
• Consideration of bandwidth;
• Consideration of other factors;
• Inclusion of cable modems; and
• A summary of the basis on which prices are compared.

A consistent measure of price

The price of a DSL service is composed of a number of features which may include:

1. Monthly rental;
2. One off installation charge;
3. Cost of modem - this may be a separate monthly rental or a one off cost; and
4. For a small number of services there may be usage charges based on either:
• a cost per minute usage
• a cost per GigaByte (GB).

In order to construct a single price measure, it is necessary to convert any one off charges into an effective monthly charge. This is done by assuming a "write off period", which is taken to be 3 years, as was used in the April 2000 study (ie the one off charges are divided by 36 to obtain an effective monthly rate). The results are relatively insensitive to this assumption. Where modem costs can be covered by either a one off charge or a monthly rental, the monthly rental figure has been taken. Prices between countries are presented on a consistent basis:

• By using "Purchasing Power Parity" (PPP) exchange rates to convert all prices to UK£; and
• VAT is included for residential prices but excluded from business and wholesale offers.

See Annex A for details of the rates used.

Most services on offer are "flat rate" (ie do not include any charges which vary with usage). However, in Germany it is relatively common for services to include charges on a per GB basis.

For residential consumers all operators/service providers in Germany covered in this report provide at least one flat rate service. To ensure comparability, only flat rate services are compared for residential consumers.

For business consumers, however, usage charges are more common and some operators/service providers in Germany offer only variable rate services. In order to include these in the price comparisons, it is necessary to make assumptions about consumer usage. In the April 2000 report the annual traffic volume is taken to be 200 GB for an SME business. However, this level of usage appears relatively high for small business use. Some operators/service providers in Germany offer businesses the choice of variable or flat rate service. For these operators/service providers, it is possible to calculate the "break even point" (ie the level of usage at which the price of the variable rate service is equal to that of the flat rate service). Levels of usage for different consumer types have been estimated using this approach. The figure derived in this way for higher end business use is very close to the assumption used in the April 2000 report. See Table 2.1 (below at paragraph 2.25) for the usage levels used. The sensitivity of the results to variations in these figures has been tested and is commented on (see paragraph 2.46).

Consideration of bandwidth

In comparing price it is also relevant to consider the bandwidth provided for a particular service. In principle, price would be expected to increase with increasing bandwidth. This is both because there are some additional costs in providing higher bandwidth and because it might be expected that consumers would be prepared to pay more for a higher bandwidth service. This price/quality relationship (higher price for higher bandwidth) is not, in practice, clearly seen either between, or within, countries. This is probably due both to the immaturity of the market, and the fact that bandwidth is only one of several measures of quality. It is not, therefore, possible to "normalise" prices for differences in bandwidth.

In order to present information in a meaningful way, price information is presented in the context of a measure of bandwidth. This enables unambiguously "Good value" offers to be identified (ie low price and high bandwidth). Other results will be more ambiguous in terms of the value for money they present (eg a service with a high price, but high bandwidth, may or may not be better value than a service with low price and low bandwidth, depending on the requirements of the consumer).

For residential retail services, the results focus on the cheapest deals available from operators/service providers, irrespective of the bandwidth offered.

For business and wholesale offers, information has been presented for different user types, based on the "minimum bandwidth required". For these market segments, services are provided over a wide range of bandwidth. Users will need to determine which service best meets their requirements (eg an asymmetric service, offering 500kbit/s downstream bandwidth, may not be an effective substitute for a symmetric service offering in excess of 2Mbit/s in both directions).

Both up and down stream bandwidth are relevant in defining the service. In order to compare services on the basis of bandwidth, it is convenient to construct a single measure of bandwidth. This report follows the approach used in the April 2000 study:

• For residential, low bandwidth business and low bandwidth wholesale services, the measure used is the downstream bandwidth. This reflects the expectation that these consumers will primarily be interested in downstream bandwidth.
• For higher bandwidth business and wholesale services, the geometric mean of down and up-stream bandwidth is used. It is assumed that these consumers will be equally interested in bandwidth in both directions. ( The geometric mean is defined as Squareroot (downstream bandwidth x upstream bandwidth)

This study covers both asymmetric (ADSL) services, in which the downstream bandwidth is greater than the upstream bandwidth, and symmetric (SDSL) services, in which the bandwidth is the same in both directions. For two services with the same arithmetic mean bandwidth, the more symmetric service will "score more highly" on the basis of the "geometric mean" bandwidth. The choice of geometric mean as the measure of bandwidth reflects the assumption that higher end business consumers will place more value on symmetric DSL services.

Other factors

While the most obvious measure of service quality is bandwidth, there are a number of other quality aspects which may, in practice, be important:

• In particular, the extent to which the required bandwidth can actually be obtained in busy periods (bandwidth reservation) could be the main contributor to the final service standard. The key parameter is the contention ratio (ie the ratio of the potential maximum demand to the actual bandwidth). The higher the contention ratio, the greater the number of users that may be trying to use the actual bandwidth at any one time and, therefore, the lower the effective speed offered, especially at peak times.
• The extent of service availability:
• some services may be available only in certain geographic areas (eg in the USA, service from an individual supplier may only be in the coverage area of one of the "baby Bells");
• even within areas where the service has been rolled out (ie the local exchange has been upgraded), customers with very long lines may not be able to receive the service;
• there are likely to be areas where services are not offered as roll out of DSL services is not economically viable.
• The delay in transmission between successive packets of data, or latency, which affects how continuous or disjointed a communication appears to the user.

Limited data on aspects such as contention ratio is available. This information is not, therefore, presented in this report.

Services also differ in respect of what is included in the service "bundle" (eg the number of e-mail addresses, the amount of web-space, the number of dynamic and static IP addresses).

It should be noted that these other factors can be significant in terms of the price offered by operators/service providers. As an example of this, the "BT Business 500" product is priced at £40 per month - this has the same bandwidth (512/256) as the "BT Business 500 plus" product which is priced at £100/month. Key differences between the services are contention ratio of 50:1 for the former and 20:1 for the latter and the ability to connect only 1 user for the former but up to 4 users for the latter. The price/bandwidth information can, therefore, be misleading when considered in isolation. The comparisons should, therefore, be treated with a degree of caution. It is expected that the general conclusions regarding cross country comparisons will be valid. The report is not a guide to the relative "value for money" offered by individual operators/service providers.

Cable modems

In addition to collecting data for DSL services, data has also been collected for cable modem services. Cable modem services are expected to provide a close substitute for DSL services, particularly for residential consumers. In presenting the results, cable modem service prices are displayed alongside the DSL prices. Country price comparisons are presented both for DSL and cable modem services taken together, and for DSL services only.

Factors such as contention ratio also play an important role in determining the level of service received by cable modem subscribers. Care is, therefore, needed in interpreting any price/quality relationship when comparing cable modem and DSL services.

Summary

Table 2.1 provides a summary of the comparisons undertaken.

Table 2.1: Summary of the comparisons undertaken

Note: GM = geometric mean

Other issues

There are a number of additional issues that arise in presenting the results. The approach taken to these is outlined below:

• Some providers provide a range of bandwidth (rather than a single figure) potentially available to a consumer for a particular service. The bandwidth actually received may, for example, reflect the distance of the consumer from the local exchange. These offers are presented using the mid-point of the range quoted. However, the impact of using the upper limit of the bandwidth range is also commented on;
• The vast majority of retail packages include internet access. Where internet access is offered on an unbundled basis, the costs of this are added in to ensure comparability;
• VAT rates vary by US State - an average figure is used for California and New York (the impact on the results of this simplification is very small);
• The standard installation costs are used where these are available (ie short term special offers are not taken into account). Prices for technician install (rather than self install) are used;
• Where an operator offers a lower rate for payment upfront (eg paying the annual amount upfront), the monthly rate has been used; and
• Where DSL is provided over ISDN (ie in Germany), the difference between the ISDN and PSTN monthly subscription charge is included in the DSL cost (ie a PSTN line is taken as the standard "base"). However, it is noted that ISDN use is relatively widespread in Germany, especially for business use, and the impact of removing the ISDN costs is commented on.

Regarding the last of these points, it is noted that there is no clear answer as to whether additional ISDN costs should or should not be included. For a consumer who does not already have an ISDN connection, the requirement to have an ISDN connection is an additional cost in taking up a DSL service. However, for consumers who already have an ISDN connection there is no additional cost.

Presentation of results

As in the April 2000 study, results are summarised by presenting charts, ranked by price, with the relevant measure of bandwidth (ie either downstream or geometric mean) superimposed. This provides one way of viewing and commenting on the results.

A further way of summarising the results is also presented in this report. A price index has been constructed based on the average price for the two lowest priced offers (for each consumer type) from two different operators/service providers in each country. This is based on the approach taken in OFTEL's other benchmarking work. ( for example International Benchmarking study of mobile services and dial up PSTN internte access). As bandwidth is also a relevant consideration, the average bandwidth is shown alongside the price indices. This type of comparison was not undertaken in the April 2000 study. This was partly because prices in the UK at the time were mainly based on trial prices and hence it was appropriate to comment on relative prices in a more qualitative way. It is expected that the analysis of price in this report will form the basis for constructing a price trend in subsequent reports.

These price indices do not take account of availability (which is discussed in Chapter 3). However, where it is known that availability is, and will continue to be, on a very limited basis, the operator has been excluded from the index.

Results are also presented in terms of:

• Charts which show the spread of prices available in each country based on the cheapest service by operator (which meet the minimum bandwidth requirements listed in Table 2.1); and
• Charts which show the spread of bandwidth available in each country based on the minimum and maximum bandwidth offered by each operator.

Results for residential

Price index

Table 2.2 below shows the results for the price level, price index and (downstream) bandwidth provided for residential services. All these measures are constructed using the average of the two lowest priced residential offers (from different operators/service providers) in each country. Results are presented both with, and without, cable modem offers included.

Table 2.2: Results for price level, price index and (downstream) bandwidth for residential broadband services

Include cable modems				Exclude cable modems
	Price level £/month	Price index	Bandwidth kbit/s	Price level £/month	Price index	Bandwidth kbit/s
France	37	92	512	48	109	512
Germany	31	78	576	32	72	448
UK	40	100	506	44	100	500
US	32	79	2,000	36	80	695

From Table 2.2 it is seen that:

• If cable modem services are excluded, then prices in Germany and the US are around 28% and 20% cheaper respectively than in the UK, while prices in France are around 9% more expensive;
• If cable modem services are included, then prices in France are around 8% cheaper than in the UK, while prices in Germany and the US are over 20% cheaper;
• Excluding cable modem offers increases the costs in all countries, ie generally one or both of the two lowest priced services are cable modem rather than DSL services. In the UK, cable modem and DSL prices are closer than in some other countries. Excluding cable modem services, therefore, improves the relative UK position;
• Compared to the results for business (see section below), the price levels for residential consumers are relatively close together; and
• The bandwidth provided in the US is much higher than in Europe.

It is also noted that virtually all residential services are asymmetric services. Only a small number of services in the US are symmetric.

The results have also been run with some of the assumptions varied:

• Excluding the cost of ISDN improves the German position and makes German prices almost half of those in the UK - this comparison may be more relevant for those consumers who already have an ISDN connection; and
• Including the upper limit of bandwidth ranges offered by US operators/service providers, increases the average bandwidth of the two lowest priced offers to 2,500.

It is also noted that Primacom offer a variable rate cable modem service in Germany, which has not been included in the comparisons. At low levels of usage, the variable rate service is cheaper than the flat rate service, although it also offers much lower bandwidth. Including this service at a usage rate of 0.5GB/month would reduce German prices to around 40% cheaper than UK prices, but would also reduce the bandwidth to around 40% of that offered in the UK.

The figures in Table 2.2 (above) exclude the Kingston service on the basis of its limited availability. (Prices quoted are for Kingston's local access network area; prices of services which may be offered in future outside this area are not available). The inclusion of this service would make the UK the cheapest of the countries considered. The Kingston offer is based on an "internet over TV" package and, hence, is different to the other DSL services presented which offer internet over a PC.

The results can also be presented by considering a chart which lists the cheapest offers of the operators/service providers considered in this study (arranged in price order), with downstream bandwidth superimposed. See Figure 2.1 below (note UK operators/service providers are highlighted using darker coloured bars).

Figure 2.1: Residential DSL and cable modem services in ascending order of price, with "speed" of service (downstream bandwidth) shown separately.

From Figure 2.1 it is seen that:

• The UK Kingston offer (for internet over TV) is the cheapest service of the operators/service providers considered, though it provides a relatively low bandwidth;
• Cable modem services in the US such as Roadrunner and AT&T appear to provide good value for money in terms of offering a low priced service plus high bandwidth;
• The overall spread of prices is not that great (when compared with business offers) - the bandwidth offered does vary considerably;
• All countries appear to have some offers available spread throughout the price range.

The results can be compared with those in the April 2000 report. Generally it appears that there has been relatively little movement in terms of prices with a number of packages being offered in the range of £30-£40 both in March 2000 and October 2000.

Price and bandwidth spread

Charts showing the spread of prices and bandwidths are shown in Figures 2.2 and 2.3 respectively. The price spread chart (Figure 2.2) shows the price for the cheapest package offered by each operator as a percentage increase over the cheapest price across all countries. The bandwidth spread chart (Figure 2.3) shows the range of bandwidths on offer (based on the lowest and highest bandwidth offered by each operator).

Figure 2.2: Price spread chart

Figure 2.3: Bandwidth spread chart

Figures 2.2 and 2.3 show:

• The range of prices in all countries is similar both in terms of level and spread;
• The range of bandwidth on offer in the UK and France is very limited. This is in part due to the fact that DSL products are all based on the provision of service by the incumbent operator who offers only one residential service; and
• Both Germany and the US offer a similar range of bandwidth which is much broader than in the UK and France.

Results for business

Price index

Table 2.3 below shows the results for the price level, price index and bandwidth provided for business services. All these measures are constructed using the average of the two lowest priced business offers (from different operators/service providers) in each country. Results are presented with cable modem offers included.

Table 2.3: Results for price level, price index and bandwidth for business broadband services (cable modem services included)

	No minimum bandwidth			Minimum GM bandwidth of 500 kbit/s			Minimum GM bandwidth of 1000 kbit/s
	Price level £/month	Price index	Bandwidth - DS kbit/s	Price level £/month	Price index	Bandwidth - GM kbit/s	Price level £/month	Price index	Bandwidth - GM kbit/s
France	126	302	1,000	126	103	506
Germany	28	66	448	264	217	615	533		1,262
UK	42	100	506	122	100	500
US	33	80	608	60	49	857	71		1,127

Note: DS = downstream, GM = geometric mean

From Table 2.3 it is seen that:

• For the most basic services available to business (ie where the business has no minimum bandwidth requirement):

• prices in Germany and the US are around 34% and 20% cheaper respectively than in the UK; and
• prices in France are considerably more expensive than in the other countries - but provide significantly higher bandwidth.

• For business services where a minimum geometric mean bandwidth of 500kbit/s is required:

• Prices are widely spread between countries;
• UK prices are close to prices in France, are about double those in the US but less than half of those in Germany; and
• The bandwidth in the UK is about the same as in France, but less than in Germany and the US.

• For business services where a minimum geometric mean bandwidth of 1000kbit/s is required:

• Services are only available in Germany and the US; ( While both France and the UK offer services with a downstream bandwidth of at least 1000kbit/s, these services are asymmetric and the geometric mean bandwidth is less than 1000kbit/s.)and
• Prices in these two countries are very different, with Germany much more expensive than the US.

It is also noted that, in both Germany and the US, a range of symmetric services are offered, while in the UK and France, all services are asymmetric.

The results have also been run with some of the assumptions varied as follows:

• Excluding the cost of ISDN improves the German position for the most basic business services making German prices less than half of those in the UK. This comparison may be more relevant for business consumers who already have an ISDN connection. There is no impact for business consumers requiring higher bandwidth services;
• Including the upper limit of bandwidth ranges offered by US operators/service providers results in relatively small increases in the average bandwidth;
• Halving the usage rate (in GB) has no impact for "low" business use. It improves the German position in respect of price for "medium" and "high" business use, but German prices for the former remain 50% higher than those in the UK; and
• Doubling the usage rate again impacts on the price for medium and high business use, resulting in the gap between Germany and other countries being further increased.

Excluding cable modem services has relatively little impact on the results as these are largely offered to residential, rather than business, consumers as at 12 October 2000.

The results can also be presented by considering a number of charts which list the cheapest offers of the operators/service providers considered in this study (arranged in price order), with the relevant measure of bandwidth superimposed. See Figures 2.4 to 2.6 below (note UK operators/service providers are highlighted using darker coloured bars).

Figure 2.4: Business DSL and cable modem services in ascending order of price, with "speed" of service (downstream bandwidth) shown separately - no minimum bandwidth required.

Note: cable modem services have (CM) after the operator's name. UK services are shown in bold.

Figure 2.5: Business DSL and cable modem services in ascending order of price, with "speed" of service (geometric mean bandwidth) shown separately - minimum bandwidth 500kbit/s (GM).

Note: cable modem services have (CM) after the operator's name. UK services are shown in bold.

Figure 2.6: Business DSL and cable modem services in ascending order of price, with "speed" of service (geometric mean bandwidth) shown separately - minimum bandwidth 1000kbit/s (GM).

Note: cable modem services have (CM) after the operator's name. UK services are shown in bold.

The results can be compared with those in the April 2000 report. Generally, it appears that prices for basic business use are now lower than in April 2000. In the April 2000 report, prices for basic business use typically started at close to £100. However, in figure 2.4 above, prices start at around £30-£40.

Price and bandwidth spread

Charts showing the spread of prices and bandwidths are shown in Figures 2.7 and 2.8 respectively. The price spread chart (Figure 2.7) shows the price for the cheapest package offered by each operator as a percentage increase over the cheapest price across all countries for each business type. The bandwidth spread chart (Figure 2.8) shows the range of bandwidths on offer (based on the lowest and highest downstream bandwidth offered by each operator).

Figure 2.7: Price spread chart

Figure 2.8: Bandwidth spread chart

Figures 2.7 and 2.8 show that:

• For the most basic services available to business (ie where the business has no minimum bandwidth requirement), the price spread in the UK and US is similar, both in terms of price level and range of prices, while prices are much more narrowly spread in France and much more widely spread in Germany;
• For business services where a minimum bandwidth of 500kbit/s (geometric mean) is required, the spread of prices in the UK is relatively narrow (only France has a narrower range);
• For business services where a minimum bandwidth of 1000kbit/s (geometric mean) is required, only the US and Germany offer services, and prices in Germany are both higher and more widely spread; and
• Taking all services on offer, the range of capacities on offer in Germany and the US is much wider than in the UK while, in France, only one bandwidth level is available (the narrow spread of prices in France is likely to be a reflection of the limited range of services available).

Results for wholesale DSL

Information has been collected for the following three wholesale providers of DSL services:

• BT in the UK;
• QSC in Germany; and
• Covad in the US.

BT provides DSL services over its own network. QSC and Covad are examples of wholesale providers in Germany and the US who lease the local loop and then install DSL equipment to allow them to offer a wholesale DSL product.

A "wholesale" product is defined in this report as a product which is not sold directly to end users, rather, it is sold to a service provider who then provides the service to the consumer (and maintains the relationship with the consumer). The France Telecom Netissimo product (on which all DSL services in France are based) is considered to be a resale product, as it can be sold directly to the consumer. This product is not included directly in the retail comparisons (as it does not include internet access), but is included indirectly as part of the French ISPs DSL services. Similarly Deutsche Telekom offers certain resale products which have not been directly included.

It has been relatively difficult to establish the price level of wholesale offers outside the UK (Note data on wholesale tariffs collected by Tarifica has been supplemented by data collected by Analysys. The data provided by Analysys has a tariff valid date of 1^st Nov 2000).

Both QSC and Covad appear to offer a range of prices which may depend on the particular ISP they are selling to and the detailed contact terms. The results are presented on the basis of the lower limit of the ranges, but with the upper limit superimposed.

Given that only three companies are considered, results are presented in chart form only. Two charts are presented:

• The first (Figure 2.9 below) shows all services for a "basic" service - defined as having a minimum bandwidth (geometric mean) below 500kbit/s. However, where a higher bandwidth service is lower in price than some of these low bandwidth services, these are also included; and
• The second (Figure 2.10 below) shows all services for higher bandwidths ie with a minimum bandwidth (geometric mean) above 500 kbit/s.

This way of presenting the results is somewhat different to that for the retail services. This is because with only three operators/service providers, it is useful to include all of their offers.

It is noted that the comparison of wholesale services is not easy. It is difficult to be sure that the comparison is of "like with like". The BT IPStream wholesale services have been chosen for the UK comparison in this study, as these appear to be the most closely comparable to the Covad and QSC services.

Two BT products with the same bandwidth are included in the comparisons. The BT IPStream 500 product is aimed at residential and basic business use and differs in certain key respects (such as contention ratio) from the more expensive BT IPStream 500s product which is specifically aimed at business use.

Figure 2.9: Wholesale DSL services in ascending order of price, with "speed" of service (downstream bandwidth) shown separately - lower bandwidth services (bandwidth < 500kbit/s GM).

Figure 2.10: Wholesale DSL services in ascending order of price, with "speed" of service (downstream bandwidth) shown separately - higher bandwidth services (>500 kbit/s GM).

Figures 2.9 and 2.10 show that, based on the information available:

• BT's most basic wholesale service (IPStream 500) i