Chapter 1. The Nordic telecommunication markets
Face - Contents - Bottom - Previous/Next "Telecompetition" Chapter 1. The Nordic telecommunication marketsIntroductionIn this chapter, general trends and developments on the markets for telecommunication services in Denmark, Iceland, Norway, Sweden and Finland are described. First, markets for fixed line telephony are described. Then, focus turns to markets for mobile telephony and, at last, Internet services. Telecommunication markets have generally been very dynamic and any comparison of prices, usage, penetration and market shares will thus only provide snapshots of the market situation at a given time. The data used in this analysis have the following features: — Wholesale and retail prices are from June 2004. Only the incumbent operators' prices are reported. Actual prices experienced by customers in general depend on the price development of entrants as well. But competition from entrants will induce incumbents to respond by lowering their prices as well. A comparison of incumbents' prices alone will therefore to a certain extent reflect overall market conditions. All prices are exclusive VAT1 to best reflect market conditions. All national Nordic telecommunication markets have been liberalised in order to introduce competition. In all the countries, mobile telephony was the first market exposed to competition followed by data, fixed line minutes and eventually fixed line access. Empirical studies by Boylard and Nicoletti2 indicate that liberalisation in itself, even before competition actually emerges, has positive effects for consumers and society as such. An interesting parameter for market development is penetration rates for different services measured as customer relationships per 100 inhabitants. Table 1.1: Penetration overview (2003)
Source: ITU at http://www.itu.int/ITU-D/ict/statistics/ (Main telephone lines, Mobile and Internet) and OECD at http://www.oecd.org/document/31/0,2340,en_2649_34223_32248351_1_1_1_1,00.html Mobile penetration and penetration of main telephone lines have apparently reached saturation levels in all the five countries. In some countries, penetration of main telephone lines has even stagnated lately. In all the Nordic countries3, mobile penetration soared during the 1990es and is at present slightly higher than in the rest of Europe. As to broadband, all the countries have penetration rates above OECD and EU15 averages. During the last six months of 2003 the number of broadband users rose between 18 percent (in Denmark and Sweden) and 56 percent (in Finland). High growth rates in Finland and Norway reflect that these countries have been catching up with the other countries. Iceland experienced an increase in broadband penetration of almost 30 percent during the same period and has thus been increasing its lead among the Nordic countries. To better explain the Nordic countries' different starting points for fixed line and mobile penetration, the countries' individual demographic and geographic characteristics are taken into account. Higher population density and degree of urbanisation decrease network construction costs reflecting the economies of scale inherent in telecommunication networks. Network construction costs depend to a greater degree of geographical roll-out of the network than of number of customers served. This in turn influences the possibilities to reap the benefits of economies of scope as well because it can prove difficult to achieve a sufficient customer basis for add-on products in scarcely populated areas to profitably exploit economies of scope. As illustrated in table 1.2, there are large differences in population density and degree of urbanisation across the Nordic countries. Table 1.2: Population density and urbanisation (2002)
Source: ITU at http://www.itu.int/ITU-D/ict/statistics/ (inhabitants per km) and Development data Group, World Bank http://www.worldbank.org/data/countrydata/ictglance.htm (urbanisation), CIA Denmark has by far the highest population density and a relatively high degree of urbanisation. Theoretically, Denmark should therefore have the best preconditions for roll-out of competing networks. GDP per capita varied approx. +/- 8 percent around an average of 29.000 $ pro annum in 2003 for all countries but Norway, whose GDP per capita exceeds average GDP for the other countries by 30 percent. The importance of population density and urbanisation depends of the relevant service and the level of penetration. National coverage is e.g. often a crucial feature for buyers of mobile services. Since population density affects the potential profitability of covering a particular geographic area, it affects the economic basis for competing networks and thereby entry barriers. The degree of urbanisation influences primarily how fast and at what costs entrants can cover a significant part of the population with their own network, thereby gaining independence of the incumbent operator's network. The degree of urbanisation also provides an indication of the proportion of customers operators can cover without having to invest in less densely populated areas, where investments may be more difficult to recoup. Price comparisons are made in Euros, the national currency in Finland. All the other countries have different national currencies. All price comparisons carried out in chapter one are based on average exchange rates from the period 1 July 2003 to 30 June 2004. The maximum deviation from the average exchange rate was in Norway, where the highest price for a Euro has deviated slightly more than 6 percent from the average exchange rate. All in all, exchange rates have been fairly stable and cross-country comparisons can be carried out with only a minimum of caution on this account4. The markets for fixed line telephonyFixed line markets (for subscriptions and traffic) were the last markets to be liberalised in the Nordic countries. In Sweden, there has never been legislation preventing potential competitors from entering the market. But no competitors actually did enter the market before carrier selection was introduced by law for international calls (1993) and national long distance calls (1994). The development in Sweden thus illustrates that merely opening the market for competitors without securing them access to the incumbent's network does not suffice to create competition. Different starting points for the liberalisation have yielded different results. Before the liberalisation, customers in Finland were served by a range of local monopolies as well as a state owned operator (presently TeliaSonera), which mainly operated in sparsely populated areas but also was the sole provider of long-distance and international calls until 1993. The market opening has resulted in joint ventures as well as competition between the different local incumbents. Denmark, Sweden, Norway and Iceland all had national monopolies prior to liberalising of their markets5. These different market structures may have had an impact on the development of competition. The introduction of call-by-call carrier selection6 and carrier pre-selection7 has enabled entrants to specifically target profitable market segments for fixed line traffic. Prices for fixed to mobile and international calls have generally been reduced. National calls to fixed line telephones, particularly regional and long-distance calls, have seen at times dramatic price reductions. In some countries (Sweden and Finland) local call prices have increased, which can reflect both a rebalancing of prices towards more cost-oriented prices (which eventually will facilitate entry) as well as incumbents' market power. In Finland, the incumbents generally have market shares of almost 100 percent on local calls. Iceland has had uniform tariffs during the entire period in question. But in Denmark, Sweden and Norway, differentiated tariffs for local, regional and long distance calls have been rebalanced into uniform national tariffs. The rise in Swedish local tariffs is therefore an integral part of the price reductions for long distance calls and average prices have fallen significantly in Sweden. For all the Nordic countries, international call charges have been reduced significantly since markets were opened. The Finnish market for international calls was opened already in 1993. Since 1995, average prices for international calls have fallen 42 percent in Finland. The same applies for long distance calls in Finland where recent price increases in part reflect a recovery in prices after fierce competition pushed down prices dramatically after the market was liberalised. Even though international calls have seen dramatic price reductions there are still significant differences in the pricing of national and international calls. Even in countries bordering each other prices for calls from a border town to just the other side of the border far exceed prices for a national long-distance call. Differences in minute prices for local calls and a call to a bordering country range from 57 to more than 500 percent; cf. table 1.3. A common Nordic market for fixed line traffic apparently has not arisen so far. Table 1.3: Indexed price comparison for a 3 minute peak hour fixed line call (2004)
Source: Incumbent operators' official price lists. General increases in subscription fees and reductions in minute charges illustrate the so-called rebalancing between the two revenue components putting an end to previous crosssubsidisation of subscription fees from minute charges. Previously, such pricing regimes were intended to make fixed line telephony available to low-income users. But this artificial pricing regime had to be changed as markets were opened for competition and access and traffic were separated into individual markets through introduction of carrier selection and carrier preselection. Otherwise, entrants would have been unable to enter the cross-subsidised segments and, on the other hand, would have eroded the incumbents' revenues, which otherwise should have covered the costs of providing universal services. Cross-subsidies between rural and urban customers still prevail. The risk of cream-skimming is kept at bay though through uniform national wholesale prices. Current end-user prices across the Nordic countries are compared in table 1.4. Table 1.4: PSTN tariffs excl. VAT (2004)
Note 1: Average call tariffs are calculated as €-cent per minute for call pattern as defined in the OECD PSTN residential composite basket. Norway has, by far, the highest subscription fee. Norwegian minute charges are in the high end as well leading to the highest monthly spending for the OECD fixed line composite basket (see box 1.1 for further details on composite basket comparisons). Box 1.1: OECD composite basket
Differences in underlying costs and regulation as well as in terms for competition can be the reason for the apparent differences in overall prices and the distributions between the fixed and the variable components of total end-user expenditure. Monthly expenditures for a domestic fixed line basket are compared in figure 1.1. Before turning to the results of the cross-country comparison, a word of caution is necessary. When comparing prices for a consumption basket, potential national differences in calling patterns are neglected. Prices quoted for a consumption basket therefore do not necessarily reflect actual spending for an average consumer in the respective countries. In the following, the prices indicated show what a consumer with one particular call pattern would pay if he bought the cheapest possible fixed line subscription of the incumbent operator in each country. Figure 1.1: Monthly spending for a residential fixed line basket excl. VAT
Source: Prices calculated by Nordic competition authorities on the basis of incumbent operators' price lists. Customer expenditures in Sweden and Denmark are approximately in line with one another as is the case for Finland and Norway. The balances between fixed fee and usage are slightly different in both cases with a greater proportion of revenue generated from usage in Sweden than in Denmark and in Finland than in Norway. Differences in overall expenditure between the cheapest country (Iceland) and the most expensive country (Norway) are significant. Average spending in Norway is approx. 40 percent higher than in Iceland for a comparable consumer profile. These differences can be the result of different degrees of competition as well as differences in underlying costs and terms for competition. The actual intensity of competition is difficult to measure. One of the most tangible measures for competition is competitors' ability to acquire market shares. Concentration is relatively high in all the countries compared to other sectors. In all the countries, concentration is higher on the market for fixed line subscriptions than on the market for fixed line minutes, cf. table 1.5. This can in part be a reflection of the fact that markets for minutes in practice were opened for competition before markets for subscriptions. Another, maybe even more important, explanation can be a far stricter regulation of wholesale inputs for minutes than for subscriptions as illustrated later in this report. Table 1.5: Fixed line market shares as of December 2003
Note 1: "CR" refers to concentration rates. CR1 is the market share of the biggest company on the market. CR2 is the sum of the two biggest companies' market share etc. Disregarding Finland, the incumbents' market shares range from 54 percent in Sweden to 87 percent in Iceland measured by minutes sold. For subscriptions, the incumbents' market shares span from 83 percent in Denmark to approximately 100 percent in Sweden. So even though entrants have been somewhat successful in obtaining viable market shares on the markets for minutes, they are still struggling to gain foothold on the markets for subscriptions. The apparently weak competition on subscriptions is particularly worrying because not even general increases in subscription fees during the past 5 years seem to have permitted entrants to gain market shares to an appreciable extent. This may both reflect that wholesale terms have worsened as prices have increased and entrants therefore have been unable to benefit from the price increases or just indicate that even though terms for entry have improved they are still not good enough. Finland is a particular case because of the many local monopolies prior to the liberalisation. Market data gathered at a national level therefore does not necessarily reflect the true market situation because the former monopolies still may hold strong positions on their former "home markets". In fact, in a decision taken by the Finnish regulatory authority March 1st 2004 concerning the market for local calls, 41 operators have been found to possess significant market power on their home market, which corresponds to a dominant position. An indication of customers' propensity to switch operator can be found in the European Commission's 9th implementation report. In 2003, the Commission found that 5 pct. of Finnish customers used an alternative operator for access. 5 pct. of the customers used carrier selection for local calls compared to 65 pct. for long-distance or international calls. In all the countries, incumbents are subject to universal service obligations (USO). Incumbents are thus required to provide basic telephony services to anyone requiring such services. Only in Denmark can the incumbent apply for financial compensation for the costs of USO. In Denmark and Sweden, USOs have been coupled with price-caps on subscription fees thereby implicitly cross-subsidising customers in scarcely populated areas with revenues from densely populated areas. Competition therefore implies a risk of cream-skimming since entrants can choose only to focus on the profitable market segments. Box 1.2: Local loop unbundling
So far, problems of cream-skimming have not arisen since competition on access products is ailing and local loop unbundling (see box 1.2) in general has been unsuccessful in providing competition on PSTN8 access. By the end of 2003, Iceland performed best of all the Nordic countries with 12 percent of the local loops unbundled. Finland lead the rest of the countries with merely 3.9 percent of the local loops unbundled followed by Norway that has experienced an impressive growth in the number of unbundled lines so that 2.5 percent of the total stock of loops were unbundled by the end of 2003. Denmark trails Norway with 2.1 percent of the local loops unbundled cf. table 1.6. Table 1.6: Deployment of local loop unbundling (2003)
Source: Broadband access in the EU: situation at January 2004 plus statistics from Nordic regulatory authorities. Most of the loops unbundled are used for providing broadband access alone. Data from March 20039 suggests that only 0.6 percent of the total loops in Finland were unbundled with the aim of providing telephony at that time. In Denmark, the corresponding figure was 0.01 percent. In Sweden all the unbundled loops were used for broadband. So far, broadband thus seems to be the primary beneficiary of local loop unbundling whereas competition on markets for PSTN seems largely unaffected. IP-telephony10 has already been introduced in Norway to residential users and is expected in Denmark and Sweden shortly. This may change the otherwise gloomy outlook for competition on fixed line access through local loop unbundling. IP-telephony as an add-on to broadband access may spur demand for broadband while providing a more solid revenue basis for alternative providers of telephony than PSTN has done so far. It is indisputable that local loop unbundling so far has failed to create competition on PSTN subscriptions. One may argue that the original intention of local loop unbundling was to create the prerequisites for broadband competition rather than competition on PSTN subscriptions. Such argument does not take at least one essential feature into account. First of all, in many countries there are no alternatives to using local loop unbundling for an entrant wishing to provide PSTN subscriptions. If local loop unbundling fails to promote competition on PSTN subscriptions competition thus cannot arise at all as entrants are unlikely to achieve lower costs, at least in the short to medium run, by building their own networks. Part of the reason for the difficulties in making local loop unbundling promote competition on fixed line telephony is the margin between the incumbents' retail PSTN price and the wholesale price for access to the local loop. Gross margins for the Nordic countries are shown in table 1.7. Table 1.7: Contribution margin for PSTN using LLU (June 2004)
Note 1: Customer relationships are assumed to last for 60 month. In Denmark, Sweden, Norway and Iceland there are positive gross contribution margins for competitors using the unbundled local loop to offer PSTN subscriptions. But even in those cases there is a long way to profitability since an operator using the local loop also has to pay for collocation11, technical equipment, maintenance, and customer services etc. Only in Sweden and Norway is it possible for entrants to get access to the local loop at reduced rates when providing PSTN/ISDN subscriptions. In all the other countries an entrant that wishes to supply only PSTN telephony to a customer still needs to buy traditional full access to the local loop even though the full potential of the connection is not used. Margins improve considerably if entrants sell both PSTN and ADSL to the same customer. But the group of customers that entrants profitably can compete for is restricted considerably if entrants rely on this possibility alone as only customers that buy ADSL and PSTN from the same supplier are commercially interesting for entrants. Even though ADSL penetration is relatively high in the Nordic countries, this group of customers remains limited. As a more general consideration it is questionable whether it is desirable that entrants have to bundle two products (ADSL and PSTN) in order to be able to compete on the market for one of the products (PSTN). Entrants' only incentive to invest in local loop unbundling is thus expected additional income from traffic or the possibility to exploit the economies of scope associated with offering PSTN and ADSL to the same customer using the same access-loop. In Denmark and Norway entrants can also buy the incumbents' PSTN end-user products with a wholesale discount and resell the product at retail level. Wholesale discounts to resellers range from 18 percent in Norway to 21 percent in Denmark thereby enabling entrants to compete on prices. Resale of incumbent subscriptions is the main reason for the relative success for entrants in Denmark and Norway in gaining market shares compared to entrants in other Nordic (or European) countries. Resale of incumbent products has the drawback that it may discourage investments in independent production platforms for entrants. Too favourable terms for resellers may therefore prevent emergence of network competition. The absence of positive contribution margins for competitors offering PSTN via the local loop is either attributable to too low retail prices or too high access prices. Price differences are biggest at the wholesale level. Average monthly access charges in Norway are more than 50 percent higher than charges in Sweden. Price differences might actually reflect different costs of granting access across the region. It is much more expensive to build fixed line networks in Norway than e.g. in Denmark because of Norway's lower population density and more rugged terrain. It is beyond the scope of this report to evaluate whether prices reflect actual costs but it seems surprising if cost differences can justify that access prices in the Helsinki metropolitan area are more than 30 percent higher than access prices in Danish urban areas. Similar differences in wholesale prices prevail for traffic charges as illustrated in table 1.8. Table 1.8: Peak hour fixed line access and interconnection charges (per minute)
Note 1: Double transit access and termination does not exist in Iceland. As an illustration of the consequences of the different wholesale prices, contribution margins for a three minute local call in the Nordic countries are calculated in table 1.9. New operators are not always present at local level. In some cases, entrants therefore have to buy single or, in rare cases, double transit access and termination to provide a local call. This influences entrants' contribution margins significantly, cf. table 1.9. In Norway local exchanges cover relatively vast areas. Changes currently being implemented to the Norwegian network architecture will lead to only 12 local exchanges across the country. In comparison, there are 111 local exchanges in Denmark. Table 1.9: Peak hour avg. contribution margin in €-cents
Note: Prices are calculated for a 3 min. peak hour fixed line call The higher the contribution margin, the better the possibility for an entrant to undercut incumbents' prices and still make a profit. Denmark and Norway present the highest contribution margins followed by Finland and Sweden where contribution margins are 13-26 percent lower. Lower margins may be attributable to more fierce competition, less strict wholesale price regulation or historical reasons (e.g. end-user price regulation) that has led to relatively low end-user prices. In all countries but Iceland, gross contribution margins fall by 10-23 percent if operators buy access and termination at single transit level instead of local level. The price differences in part reflect greater investment requirements for operators present also at lower levels in the network hierarchy. In all countries but Finland, entrants achieve positive gross contribution margins even when producing a local call implies buying double transit access and double transit termination. This reduces entry barriers as entrants who are only present at a limited number of collocation sites still profitably can offer local telephony. Gross margins constitute from 53 to 74 percent of minute retail prices if access and termination are bought at local level. Gross margins for subscriptions based on access to the local loop range from -27 to 27 percent. One has to be careful making such comparisons but it does seem to be easier to enter the markets for traffic than the markets for subscriptions. Price differences at retail level are in part attributable to different balances between subscription fee and minutes (illustrated by positive and negative contribution margins for traffic and access respectively). Whereas network construction costs are determined by exogenous factors, retail prices do not face similar constraints. Current imbalances in retail pricing are thus to a certain extent the result of political objectives such as a high penetration of fixed line services, which in general means setting subscription charges low and usage dependent charges high. The countries' different regulatory regimes for retail prices are summarised in table 1.10. Table 1.10: Price regulation of PSTN
Source: Nordic competition authorities. Only in Denmark do remnants of end-user price regulation remain in place. The incumbent is subject to a price cap on subscriptions in Denmark. Regulation of minute prices was repealed in 2003. In Sweden, the incumbent, Telia, was subject to a price cap between 1993 and 2000. Since then, prices regarding PSTN have been regulated solely pursuant to the Telecommunications Act and in Telia's licensing conditions. According to the Act, the tariffs for the use of telephony services between fixed network termination points within a public telecommunications network were to be based on costs, if the service was provided by a telecommunications operator that had SMP-status. According to transitional provisions, this price regulation has been in force until 24th July this year. The regulatory authority will, however, take new decisions regarding PSTN prices in accordance with the Electronic Communications Act. In Finland, local incumbents were required to demonstrate that prices for local calls were cost oriented until price regulation was abandoned by April 2004. In Norway, the SMP-operator was subject to end-user price regulation until 2002. Telenor was subject to a price cap regulation on a basket of end user products (subscription, traffic and leased lines) until the end of 2002. In addition, under the previous Norwegian Telecommunications Act, SMP operators were required to provide services at cost oriented prices, also for end user products. Today, no such requirements are applied and regulation is focused on securing cost oriented prices at wholesale level. The markets for mobile telephonyMobile telephony is of increasing importance in the Nordic countries for customers as well as operators. All the countries have mobile penetration rates above the EU average (81 percent) and presumably close to saturation levels. For the telecommunication sector in general, revenues from mobile telephony constitute an increasing share of annual turnover, cf. table 1.11. Table 1.11: Mobile telecommunication revenue as percentage of total revenue
Source: OECD Communications Outlook 2003 Revenues generated by mobile telephony constitute an increasing share of total revenues. In most countries, the share has more than doubled from 1995 until 2001. Iceland tops the list with almost half of total revenue generated by mobile services. Finland trails Iceland by a narrow margin. In Norway, Sweden and Denmark, mobile revenue account for approximately 25 percent of total revenue. As prices for mobile services have generally decreased, it is increasing consumption that drives mobile telephony's increased proportion of total turnover. Usage of minutes and SMS has increased dramatically over the past years. Monthly usage across the Nordic countries in 2003 is illustrated in figure 1.2: Figure 1.2: Average monthly usage of minutes and SMS per user (2003)
Source: National regulatory authorities' statistics. Denmark toppled Norway as top performer measured by usage of SMS in 2003 as Danish SMS usage per customer almost doubled from 2002 to 2003. Norway had first place for at least four years. Finland has a solid lead on usage of minutes as consumption per user is almost 30 percent higher in Finland than in Iceland, who has the second highest usage per customer. MMS usage is still relatively low compared to SMS across the region. Norway has taken the lead in 2003 with an average usage of approx. 4.7 MMS per mobile customer per year. In Sweden and Finland usage was 0.8 and 0.7 respectively whereas usage in Denmark is at 0.5 MMS per user per year. Norway's high mobile usage is followed by leadership in price-levels as shown in figure 1.3. It is difficult to compare prices on such a dynamic market. The following comparison of monthly expenditure for three categories of customers thus only provides a snapshot of the prices in June 2004. Three different composite baskets as defined by OECD are considered in order to compare prices across countries but with the same reservations as in the comparison of fixed line prices. One additional concern needs to be taken into account when comparing prices for mobile composite baskets. In Denmark, Norway and Sweden operators often subsidise customers' handsets. High spending on subscription and usage in these countries may therefore be offset by lower handset prices for consumers. Handset subsidies are not included in figure 1.3. Figure 1.3: Monthly spending for mobile baskets
Source: Calculations made by Nordic competition authorities on the basis of incumbents' public price lists. Prices as well as the balances between subscription fee, minute charges and SMS prices vary significantly across the countries. End-user prices vary between peak-time and off-peak in some countries, on-net and off-net in others while uniform prices during the day are standard in yet other countries. There seems to be a tendency towards uniform pricing independent of receiving network and time of the day. Average monthly spending is approx. 60 percent lower in Denmark than in Sweden when comparing the incumbents' best offers for high-using customers. The greatest price differences are found for SMS. The Danish price per SMS is only 15 percent of the corresponding price in Sweden as indicated in figure 1.4: Figure 1.4: SMS prices excl. VAT
Note: The Swedish SMS price is calculated as an average between on-net and off-net prices. The SMS prices compared in figure 1.4 are taken from the subscriptions compared in figure 1.3. For heavy users of SMS flat-rate subscriptions are available in several of the Nordic countries. Other kinds of subscriptions include a given number of free SMS (e.g. 100 SMS per month). SMS prices are therefore difficult to compare for customers with high SMS usage. The price differences for voice and SMS can reflect differences in competition intensity as well as in network costs. The magnitude of price differences across borders seems to exclude the idea of a pan-Nordic common market for mobile services so far. The prerequisites for a common Nordic market ought to be in place with companies such as Telenor, TeliaSonera and Tele2 present in more than one of the countries in question. But roaming prices remain high compared to prices for calls between local customers, cf. table 1.12. Table 1.12: Total costs for 3 min. call during peak hour (mobile to mobile)
Source: Prices calculated by Nordic competition authorities on the basis of incumbent operators' price lists. Total costs for a three minute call to a local customer are in many cases more than several hundred percent higher for a foreigner roaming in the same national network than for a local customer. The price comparison in table 1.12 is made for calls that can be routed using the same network elements similarly. Variable production costs should therefore be equal for a call between two national customers and between a national and a roaming customer. Some price differences may be attributable to higher average total costs for roaming customers as the setup costs for the roaming agreement can only be distributed over a relatively small number of minutes compared to e.g. a national service provider. But it appears improbable that the whole price difference should be attributable to higher costs. The existence of at times very high roaming prices is by no means unknown to regulators. Consequently, national markets for international roaming have been designated by the European Commission for market analysis under the new European regulatory regime for markets for electronic communication. High prices do not per se call for regulatory intervention though. There are competing mobile networks in all the Nordic countries and the prerequisites for competition at wholesale level for roaming are therefore in place. High roaming prices may therefore simply indicate an immature market as customers focus primarily on prices for national traffic. But as prices for national calls are driven down by at times fierce competition, roaming prices becomes an ever more likely future battleground for operators as the spreading gap between prices for national and international calls may spur customer awareness. Prices for mobile originated cross-border calls (e.g. a Swede calling from Sweden to Finland) are relatively high as well. As is the case for fixed line, the costs for providing international calls are unlikely to exceed costs for national off-net calls considerably. Retail margins thus remain high. Part of the reason for this is that e.g. service providers in some countries only can buy international interconnection from their network provider. Network providers can thus limit competition on international calls by keeping wholesale prices for international interconnection high. So far, is seems fairly obvious that the Nordic region is divided into separate national markets. To shed light on the state of competition in the individual countries, focus is turned to market structures across the region. Measured by market share, entrants have generally performed better in mobile markets than in fixed line markets. In Sweden, the second biggest operator is only trailing the incumbent's market share by 4 percentage points. Concentration rates remain high across the board as shown in table 1.13. Table 1.13: Market concentration as of end of 2003
Note 1: Market share is measured by number of subscribers. Concentration rates and HHI indicate that Denmark had the least concentrated market by the end of 2003. But 2004 has brought a rapid consolidation of the Danish market. TDC completed its takeover of the biggest serviceprovider thereby increasing its market share to 42 percent. The second biggest MNO bought another successful serviceprovider increasing its market share to 24 percent and, in July, the third and fourth biggest MNO's announced their intention to merge achieving a total market share of 22 percent. CR4 thereby increases to 98 and HHI is increased by 50 percent to 3011. Denmark remains one of the least concentrated markets though. In Finland, introduction of number portability12 has increased customer mobility considerably. By mid 2004 HHI is thus reduced to approx. 2900. There can be several reasons for the seemingly more level playing field that has allowed entrants to achieve significant market shares in the mobile market. First of all, it is far easier to build competing infrastructure for mobile than for fixed line telephony. This has allowed entrants to develop independently of incumbents. Second, markets have grown rapidly in all countries pursuant to the entry of new competitors allowing entrants to gain market share without actually conquering the incumbent's customers. Third, constant development of new technology and shorter, relative to fixed line, customer equipment lifetime provide additional incentives and occasions for customers to change mobile provider. Mobile markets are therefore inherently more dynamic than fixed line markets. Denmark and Finland have relatively smaller average customer base per operator than Sweden and Norway. Iceland is in a league of its own with the smallest customer basis per network operator. Fewer customers per operator exert two opposing effects on prices: tougher competition between operators to obtain economies of scale, which increases downward pressure on prices, and diminished possibilities to actually obtain economies of scale, which increases the level to which prices can actually fall. The significant geographic differences between the Nordic countries influence costs, which ultimately determine how low prices can get. Accordingly, it is difficult to rank the countries according to the intensity of competition by comparing prices. Neither are differences in market concentration at a level that allows for conclusions about differences in competition intensity. Mobile markets in all the Nordic countries appear dynamic and prices are following a downward trend all across the region. Competition between fixed line and mobile telephony Prices for mobile telephony still exceed fixed line prices for an average customer as defined by the OECD. Despite significant price reductions on mobile telephony, mobile prices have thus not yet reached levels where an average customer would consider giving up his or hers fixed line telephone and rely completely on mobile telephony. But for customers with modest fixed line usage or with high demand for calls to mobile phones, substituting fixed line telephony with mobile telephony has become attractive in many cases. One way to illustrate how close substitutes fixed line and mobile telephony are, is to compare prices for different levels of consumption. Because of differences in price structures (high subscription fee and low minute charges for fixed line and vice versa for mobile), customers with sufficiently low demand for fixed line minutes will be better off giving up their fixed line subscription and use mobile instead. In table 1.14 it is calculated how low a consumer's usage needs to be before it pays to give up a fixed line subscription and use mobile instead. The threshold is indicated as number of minutes with a call distribution as defined in OECD's domestic fixed line composite basket. The higher the number of minutes indicated the more customers will find it advantageous to change. Table 1.14: Threshold for fixed line and mobile substitution (minutes)
Source: Calculations performed by Nordic competition authorities on the basis of incumbents' public price lists. As indicated in the first column, users in Denmark who do not otherwise use a mobile phone will find it advantageous to give up their fixed line subscription if their monthly usage of fixed line is less than 210 minutes. The corresponding usage in Norway is 51 minutes. The picture changes somewhat if the customer instead is assumed to be a heavy user of mobile telephony considering giving up his fixed line phone. In that case, the customer already pays possible subscription fees for his mobile phone and is in some cases able to get lower minute charges than mobile customers with low or medium usage. In that case Norway tops the list with break even between the two types of telephony at a monthly usage of fixed line telephony of 233 minutes. Continued downward pressure on mobile minute charges is likely to make it attractive for even more customers to give up fixed line telephony in the future. But there are at least two countervailing developments on the market. In countries where mobile call termination is not already regulated (as it is in e.g. Norway), likely future regulation of mobile termination charges will lower prices for fixed line customers calling mobile phones and inflict losses on mobile operators. Mobile prices are thus not likely to continue to fall at the same rate as today. In Denmark, Finland and Iceland, it is cheaper to call a mobile phone using a mobile phone during peak-hours than using a fixed line. These price differences are undoubtedly part of the reason for the relatively high substitutability between fixed line and mobile in those two countries. During off-peak hours, only Norway has higher mobile-to-mobile than fixed-to-mobile prices as illustrated in figure 1.5. Figure 1.5: Fixed to mobile versus mobile to mobile charges
Note 1: The price comparison is made for incumbent customers calling incumbent customers. With a growing proportion of calls going to mobile phones, prices for that particular calldirection become an ever more important parameter in the competition. Whether mobile telephony is a competitor for fixed line operators depends thus to a certain degree on the individual customer's consumption. Second, increased penetration of broadband lowers average costs for supplying PSTN and paves the way for IP-telephony. Lower costs and tougher competition on fixed line subscriptions is likely to push down prices and thus tip the balance in favour of having both fixed line and mobile for many customers. The market for broadband Internet access may therefore influence potential competition between fixed and mobile telephony. The markets for broadband accessThe Nordic countries are performing relatively well in the field of broadband Internet access measured by penetration rates compared to other OECD countries as indicated in figure 1.6. Figure 1.6: OECD broadband penetration
Source: http://www.oecd.org/document/60/0,2340,en_2649_34225_2496764_1_1_1_1,00.html By the end of 2003, all the Nordic countries performed better than both the EU and the OECD countries on average. Iceland leads Denmark by a narrow margin followed by Sweden. OECD statistics indicate that penetration in Iceland has more than doubled from June 2002 to June 2003. Broadband markets are growing rapidly and annual growth-rates above 50 percent are not unusual. Unbundling of the local loop is completed in all the Nordic countries and has acted as a spark that set off an explosion in demand. Access to the local loop has given entrants the chance to move first on the market for broadband services. Incumbents have in turn had no choice but to enter the market as well. Different access products have been developed. Shared access to the local loop and bit stream access have made it cheaper and easier to enter the retail-market as entrants are given the possibility to provide broadband services to end-users buying PSTN subscriptions from other providers. In addition, bit stream access lowers the barriers to entry by diminishing investment requirements for potential entrants, who, by relying on bit stream access, do not have to invest in their own access infrastructure. Incumbent operators have entered the broadband market aggressively and have succeeded in acquiring the lion's share of the booming market, cf. figure 1.7. Figure 1.7: Incumbents' market shares for broadband and ADSL
Note: Finnish market shares are for Elisa Corporation in the Helsinki metropolitan area. Incumbents are generally performing better on the distribution of ADSL-products than on the overall market. This is due to one major difference between competition on PSTN and competition on broadband which is the existence of competing infrastructure and vertically integrated competitors. In all countries but Iceland, cable-TV providers have entered the market for broadband services. The cable-TV providers' networks only cover limited areas, typically densely populated areas, and will probably never attain national network coverage. Whether cable-TV providers will develop into a real threat to the telephony incumbents will thus depend on how the markets for broadband services develop. So far, cable-TV providers have achieved even significant market shares in many countries as indicated in figure 1.6 and further illustrated in figure 1.8. Figure 1.8: Broadband market shares distributed by technology
Source: http://www.oecd.org/document/33/0,2340,en_2649_34225_19503969_1_1_1_1,00.html In neither of the countries do cable-TV-operators grant access to their networks to independent ISPs. ISPs thus rely solely on access to the incumbents' networks either through full access, shared access13 or bit stream access14. Prices as well as upload and download speeds diverge across the countries as illustrated in table 1.15. Table 1.15: ADSL products and prices
Note 1: Customer relationships are assumed to last 36 months. It is difficult to compare prices across the five countries because of different standards for upload and download speeds. Yet Sweden and Norway seem to stand apart from the other countries by offering higher transmission speeds at significantly lower prices. Cost analysis in the Danish investigation into possible predatory pricing of ADSL services showed that production costs are largely independent of bandwidth as far as network equipment is concerned. Higher bandwidth may lead to higher consumption which in turn leads to higher expenditure for IP connectivity though. With some caution, it is therefore meaningful to compare contribution margins across countries. Table 1.16: Monthly ADSL gross contribution margins (June 2004)
Source: Calculations performed by Nordic competition authorities on the basis of data from incumbents' official price lists and standard offers for wholesale network access. The apparently lower prices in Sweden and Norway are accompanied by equally low gross contribution margins. Low prices and contribution margins can indicate fierce competition but may also lead to problems in the longer term as markets develop, market growth slows down and market shares therefore become more and more entrenched. Low retail margins on the incumbents' product implies a greater financial strain on entrants, who often do not have the incumbents' strong financial backing. Low prices and strong growth in the short term may therefore come at the cost of less competition in the future if competitors fail to achieve a sufficient customer basis when markets are booming. In Denmark and Finland, entrants have accused the incumbents of engaging in predatory pricing in the ADSL markets. In Finland, the competition authority intervened in favour of the entrants. In the Danish case, the competition authority did not find sufficient proof to substantiate the allegations of predatory pricing. Similar complaints are currently being investigated in Sweden. In Norway, predatory pricing allegations were under investigation by public authorities but possible problems were solved without regulatory intervention. Successful creation of competition that may pave the way for deregulation of the markets depend to a large extent on creation of infrastructure competition since that is the only way to reduce the competitors' dependence of the incumbent. Cable-TV networks have so far proven to be the most viable substitute. Cable-TV networks (and power utility companies) enjoy at least two major advantages compared to other types of alternative infrastructure. The first advantage is that the companies already have well-established customer relationships with potential broadband customers. The second major advantage is that these companies do not have to rely solely on income from their broadband business because revenues from cable-TV content cover a significant part of operating costs. The potential for competition from cable-TV operators depend on the coverage of their networks. The percentage of homes passed by a cable-TV network (homes that are either connected to cable-TV infrastructure or where cable- TV infrastructure is available in the immediate vicinity) in the Nordic countries is illustrated in figure 1.9. Figure 1.9: Percentage of homes passed by cable-TV network. (2001)
Source: Broadband and telephony services over cable television networks (DSTI/ICCP/TISP(2003)1/FINAL N ot all homes passed by a cable-TV network can actually use the network for Internet access because not all parts of the networks have been equipped for two-way communication. But at least in Finland, Sweden and Denmark, cable-TV networks have reached a sufficient proportion of potential customers to exert significant competitive pressure on DSL providers using the telephone network. Iceland has seen strong growth in cable-TV penetration in recent years. And recent Norwegian figures suggest that the OECD-figures reported in figure 1.9 may underestimate Norwegian availability of cable-TV networks that could be as high as 56 percent. The incumbent telecommunication carriers are also significant cable operators in Denmark, Finland, Iceland and Norway. As long as a significant proportion of cable-TV customers are served by the incumbents, competition can be expected to be limited between cable-TV networks and telephone networks. The biggest cable-TV operator in Sweden (ComHem) was divested from the incumbent carrier (TeliaSonera) in 2003. It is therefore too early to expect notable effects from the divestiture at the present time. But the potential for competition between cable-TV operators and fixed line carriers is clearly illustrated by experiences in Belgium where cable-TV is available to 100 percent the households. The largest Belgian cable- TV operator (Telenet) delivers cable modem services to 17 percent of the homes passed. Belgium has one of the highest penetration rates and download speeds of minimum 3 Mbps. Use of power-lines for data transmission has so far only to a limited extent been introduced commercially in the Nordic countries. But power companies in several countries currently test equipment that later can lead their way into the broadband markets. At least one Norwegian and two Finnish suppliers seem to have overcome initial technical difficulties. Successful use of power lines has at least one significant advantage compared to the possibilities currently offered by cable-TV operators: Power companies can reach every household. If current research into use of power lines for telecommunication is successful it may therefore bring a dramatic push towards fierce competition at wholesale as well as retail level for all fixed line services. Footnotes1 VAT rates range from 22 in Finland over 24 percent in Norway, 24½ in Iceland to 25 in Sweden and Denmark. 2 Olivier Boylard and Giuseppe Nicoletti (2001), "Regulation, Market Structure and Performance in Telecommunications", OECD Economic Studies No. 32. 3 In this report the term "the Nordic countries" refers to Denmark, Finland, Iceland, Norway and Sweden. 4 Exchange rates vis-à-vis Euro (1 July 2003 to 30 June 2004)
Source: Danmarks Nationalbank (Danish Central Bank) daily observations. The average exchange rate is calculated as a simple average of all daily observations. 5 In Denmark regional carriers were merged into one national monopoly before markets were liberalised. 6 Customers can choose which operator to use for each call by dialling an access code/prefix before dialling the called party's phone number. 7 Instead of dialling a prefix at the beginning of every call customers can choose to have the prefix of their operator of choice programmed into the network. Customers then use this operator for all calls. 8 PSTN is an abbreviation of Public Switched Telephone Network and refers to what has become the standard subscription type for residential customers. 9 "Remedies for Broadband Services"; Paper prepared for DG InfoSoc by Martin Cave. September 2003. 10 IP refers to Internet Protocol. IP-telephony is telephony via the Internet build. IP-telephony is feasible even at low bandwidth but to deliver a quality of service comparable to PSTN it would usually take a broadband connection. 11 Location of entrants' network equipment at incumbents' exchanges 12 Customer can retain their telephone number when changing telecom operator. 13 Entrants pay a reduced fee for leasing an access line if the line also is used for e.g. PSTN. 14 Bit stream access gives entrants a possibility to lease access to an access line and the incumbents' equipment at the exchange. Version 1.0 October 2004 • © Danish Competition Authority. | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
