EUROPEAN COMMISSION

Brussels, 24.2.2021

SWD(2021) 28 final

COMMISSION STAFF WORKING DOCUMENT

IMPACT ASSESSMENT

Accompanying the

Proposal for a Regulation of the European Parliament and of the Council

on roaming on public mobile telecommunications networks within the Union (recast)

{COM(2021) 85 final} - {SEC(2021) 90 final} - {SWD(2021) 27 final} - {SWD(2021) 29 final}

Table of contents

1Introduction: Political and legal context

2Problem definition

2.1What are the problems?

2.2What are the problem drivers?

2.3How will the problem evolve?

3Why should the EU act?

3.1Legal basis

3.2Subsidiarity: Necessity of EU action

3.3Subsidiarity: Added value of EU action

4Objectives: What is to be achieved?

4.1General objectives

4.2Specific objectives

5What are the available policy options?

5.1What is the baseline from which options are assessed?

5.2Description of the policy options

5.3Options discarded at an early stage

6What are the impacts of the policy options?

6.1Model for the assessment of economic impacts (sustainability analysis)

6.2Impact of Option 2

6.3Impact of Option 3

6.4Impact of Option 4

7How do the options compare?

7.1Effectiveness

7.2Efficiency

7.3Coherence and proportionality

8Preferred option

8.1Sustainable and genuine RLAH

8.2REFIT (simplification and improved efficiency)

9How will actual impacts be monitored and evaluated?

Glossary

1Introduction: Political and legal context

The roaming policy has been one of the European success stories, giving millions of consumers and businesses in the EU the benefits of the Digital Single Market in their daily life. The Roaming Regulation expires on 30 June 2022.

Roaming (in the sense of the Roaming Regulation) is a service that allows a customer of a public Mobile (Virtual) Network Operator (M(V)NO) in one EU/EEA country (the home operator) to access mobile services (voice, SMS or data) when travelling in another EU/EEA country, by connecting to the network of a Mobile Network Operator (MNO) in that country (the visited operator). It comprises a wholesale roaming service (provided by the visited operator to the home operator) and a retail roaming service (provided by the home operator to the roaming customer). The relevant retail and wholesale roaming charges are regulated (among others) in the Roaming Regulation. See also Annex 7 (How roaming works).

The Telecoms Single Market Regulation 2015/2120 mandated, after more than 10 years of regulating the EU roaming market, the end of retail roaming charges in the Union subject to fair use policy and a sustainability derogation by amending Regulation (EU) 531/2012 (Roaming Regulation).

Figure 1. Retail Roaming surcharges in the EU (2007-2017)

Since 15 June 2017, customers have had access to mobile services (voice, SMS or data) at no extra cost when travelling periodically in the EU/EEA. In these cases mobile operators are not allowed to levy any charges in addition to the domestic price for roaming services. These roaming rules are widely known as "Roam-Like-At-Home" (RLAH). To prevent abusive or anomalous use of roaming services, such as permanent roaming at domestic prices, that may have detrimental effects on the domestic markets, mobile operators may apply a fair use policy.

The rapid and massive increase in roaming traffic since June 2017, shows that the RLAH reform has unleashed the untapped demand for mobile consumption by travellers in the EU.

Figure 2. EEA data roaming consumption increase (Q3 2016 - Q3 2019)

The introduction of RLAH for end-users was complemented by measures aiming to ensure that operators can provide regulated retail roaming services in a sustainable way. In this context the term “sustainable” means that operators can either fully recover the cost of providing retail roaming services or at least that the incurred negative retail roaming margin is very small (less than 3%) compared to their domestic profits. At wholesale level, it called for a sharp decrease to regulated maximum wholesale rates (wholesale caps). Regulation (EU) 2017/920 reduced the wholesale caps by 36% for voice (to 3.2 €c/min), by 50% for SMS (to 1 €c/SMS) and by an initial 85% for data (from 50 €/GB to 7.7 €/GB), followed by a glide path with a last step at 2.5 €/GB in 2022. These price caps were designed to ensure that operators could also recover the costs of providing wholesale regulated roaming services.

Figure 3. Wholesale price caps reduction in June 2017 and data cap glide path (2017-2020)

This Initiative is included in the 2020 Commission Work Programme addressing the specific objective “Digital for consumers” and has to be seen in the broader political context of creating a Europe Fit for the Digital Age. The Initiative contributes to the ambition to make the most out of the digital transition to enhance opportunities to connect, communicate, solve societal issues and do business.

The Initiative is coherent with the logic of consistently addressing barriers to the Single Market 1 , and taking actions to ensure that barriers already addressed will not re-emerge, like could be the case for the expiring roaming rules.

One of the main objectives of the Roaming Regulation is to protect consumers by reducing the level of charges that users of public mobile telephone networks have to pay for cross border services. It complements and supports the European Electronic Communication Code (EECC) 2 , the regulatory framework for electronic communications, that Members States have to transpose by 21 December 2020. The EECC not only aims to enable high connectivity and 5G deployment for the benefit of all Europeans, but also to ensure effective protection of consumers through sector specific rules.

The Roaming Initiative is also complementary to the cross-border portability of online content 3 . Thanks to those two initiatives, Europeans are now able to travel throughout the EU without worrying about mobile roaming charges or losing access to music, games, films, learning tools, work platforms, sport events, health applications and other services for which they have already paid. For example, the roaming initiative also facilitates the access to European digital culture 4 since it gives end-users the confidence to stay connected and accessing content as for example Digitised Cultural Archives 5 .

The EU Roaming policy has paved the way for several regions outside the EU to introduce similar approaches 6 to lower roaming charges in cross border settings to address similar market failures.

The European Commission has set as one of its main objectives to respond to the higher connectivity needs of citizens and enterprises 7 and to take advantage of the digital transformation to strengthen the social and economic resilience of the EU and the Member States, their sustainable growth potential and job creation. The Roaming initiative, by facilitating cross-border connectivity, contributes to these objectives.

Europe has already achieved wide 4G coverage (96% average and 99% aggregate 4G coverage in June 2019) and, despite some delays, operators have started deploying 5G networks. The new 5G infrastructure will further increase data consumption. To this end, the roaming initiative enables the seamless use of these infrastructures across borders.

Moreover, the COVID-19 pandemic has accelerated the trend towards the digital transformation, forcing the vast majority of Europeans to work, study and enjoy entertainment online. Once back to normal, learning, working, socialising and e-commerce applications, which we learned to use during the lock-down, are likely to be used increasingly also when travelling in the EU. This requires sufficient quality and affordability of roaming services.

This Impact Assessment (IA) report is a follow-up to the Commission’s Review report 8 adopted on 29 November 2019 (hereinafter the 2019 Review report) and the accompanying Staff Working Document (the 2019 SWD) that examined how the roaming market functions. Roaming Regulation Article 19 requires the Commission to assess the effects of the abolition of retail roaming charges and if appropriate, to present a relevant legislative proposal.

2Problem definition

2.1What are the problems? 

2.1.1Evaluation and evidence base

The elements addressed by this Initiative have been identified in the 2019 Review report and a broad range of data (see Annex 1 Evidence base) used to evaluate how this intervention has performed and how the roaming market functions, based on the Roaming Regulation rules. The backward-looking aspects of the public consultation and the findings of the joint online surveys by the Commission and BEREC (Body of the European Regulators for Electronic Communications) held in 2018 and 2019, complement the conclusions of the roaming Review report (see Annex 2: Stakeholder Consultation).

The Review report confirms the success of the RLAH reform and the overall good functioning of the roaming market under the roaming rules. The report concludes that, despite signs of some competition dynamics on both the retail and wholesale roaming markets, the underlying basic competition conditions have not changed, and are not likely to change in the foreseeable future, to such an extent that retail or wholesale regulation of the roaming market could be lifted. In view of the adequate functioning of the safeguard rules at retail level (fair use policy and sustainability derogation), the Commission indicated that it does not intend at this stage to amend the rules laid down in the Implementing Regulation (EU) 2016/2286.

The problems identified in this section reflect the Review report findings and other evidence collected in relation to:

I)the need to revise wholesale caps with a view to ensuring sustainable provision or retail roaming services;

II)the potential to enhance genuine RLAH experience to end-users not only in terms of price but also addressing issues related to quality of service, access to emergency services and calls to value added services;

III)the potential to clarify access to all network technologies and generations, the possibility to avoid fraudulent generation of traffic towards international numbers and facilitate innovation.

For further details on these findings, see evaluation elements presented in Annex 6.

At present, up to almost 170 million European roaming customers enjoy RLAH 9 . If Roaming rules are not prolonged and if further problems identified in the Review Report are not addressed, end-users would risk losing benefits of staying connected, like at home, while travelling in the EU/EEA. This would lead to lost benefits and lost consumer surplus linked to the roaming rules (See counterfactual analysis in Section 3.3. on EU added-value and Annex 4B). In light of these risks, the no-regulation option has been discarded (see section 5.3). Consumers are now used not to restrain their roaming consumption, as confirmed by the Eurobarometer 10 and the public consultation (See section 3.2 and Annex 2).

2.1.2A. Problems to ensure sustainable provision of RLAH

A.1 Current wholesale measures might not be sufficient to render RLAH sustainable for all operators

The main problem to be addressed in this review is the need to avoid the risk of an increasing number of operators facing sustainability challenges and therefore possibly being obliged to make recourse to derogations to the prohibition of levying roaming surcharges, which would in turn limit the possibility for their consumers to fully benefit from RLAH. RLAH is only possible with wholesale rates which support the provision of roaming services at domestic price level. Consequently, the roaming regulation has lowered and capped the rates. Capped wholesale rates reduce the risk of operators facing a negative roaming margin exceeding their domestic profit margin by 3%, which would make them eligible for requesting derogation to RLAH in case of sustainability problems. In addition to capped wholesale rates, sustainability of the RLAH is ensured through safeguard rules on fair use policies and a possibility for derogations to tackle the unlikely situation where the provision of RLAH is not deemed sustainable 11 .(See Annex 7: How roaming works). 

Figure 4. Functioning of the roaming services

As explained in section 1, the roaming customer connects to the visited operators to use mobile services (calls, SMSs, access to data). The visited operator charges the home operator for this service (wholesale charge) while the home operator in turn charges the subscriber (retail charge). The RLAH rules have abolished the retail surcharge. (See Annex 7: How roaming works).

The main problem that the regulation needs to address is striking a balance between ensuring cost recovery for operators providing wholesale inbound 12 roaming services and minimizing sustainability challenges for outbounder 13 operators and MVNOs. The possibility of negative economic consequences for certain operators of the roaming rules is also linked to the need to have an intervention that is limited in time (see Section 7.3. on proportionality).

The rapid increase in roaming traffic (especially data) intensifies sustainability challenges, especially for operators that cannot balance their roaming costs with wholesale roaming revenues. Operators with low countervailing power (typically MVNOs and small MNOs with very little inbound wholesale traffic) get very low discounts (or no discounts at all) and thus have to pay wholesale rates that are close or at the level of the wholesale price caps. The Commission services have analysed the sustainability of providing regulated retail roaming services over the 6-year period 2020-2025 (see Annex 4). The analysis utilizes operator data collected by BEREC to forecast inbound and outbound roaming traffic and estimate revenues and costs for home operators. According to the sustainability analysis, with current wholesale caps, already in 2023, 27% of roaming providers are likely to have a negative roaming margin equivalent to 3% or more of their domestic margin.

This may lead to an increased use of sustainability derogations, therefore preventing an increasing number of EU citizens to enjoy the full benefits of RLAH. The operators that are at higher risk of sustainability challenges are the ones coming from “outbound” countries where the traffic imbalances are higher, as they are both unable to rely on inbound roaming revenues and also serve large volumes of outbound roaming to retail roamers, which RLAH prevents them from charging for.

Figure 5. Map of in- and outbounder countries

The regulated price caps have been necessary to bring prices down on the wholesale roaming market. The 2019 Review report showed that at wholesale level, since RLAH was put in place, the sharp reduction in price caps has resulted in actual wholesale roaming prices below the price caps (see 2019 Review report and Annex 5: Baseline). The price caps have continued to act as benchmark prices in wholesale roaming negotiations. Any discount on the wholesale roaming market is made from these reference prices. However, most MVNOs and some smaller MNOs without multinational presence and with large outbound roaming imbalances are often unable to negotiate rates significantly below the cap, as exemplified in the Review report and in the public consultation, where 3 out of 4 MVNOs report paying wholesale prices at or slightly below caps level and only 1 out of 4 reports paying significantly below caps level. On the other hand, almost half of the MNOs (46%) report paying significantly below caps for voice and more than half (60%) report paying significantly below caps for data. Less than 1 out of 3 and 1 out of 4 report paying slightly below caps for voice and data, respectively.

The 2019 Review Report concludes that the adequacy of wholesale roaming caps needs to be assessed, but it indicates that both safeguard rules at retail level (fair use policy and sustainability derogation) have worked generally well. Therefore, the Commission does not intend to amend the rules laid down in the Implementing Regulation (EU) 2016/2286. In the absence of fair use policies, sustainability would worsen, both the number of operators with negative roaming margin and the level of negative margins would increase.

According to the Staff Working Document on the findings of the review of the rules on roaming fair use policy and the sustainability derogation (the CIR SWD), the vast majority of operators (86% in total but 95% of MNOs) have been applying a fair use policy. Also according to the CIR SWD, the majority of operators (more than 70%) who have used each fair use policy perceive it as effective or partially effective. Despite the widespread use of fair use policies, the traffic subject to fair use surcharge is limited. According to data collected by BEREC for the International Roaming Benchmark Report, it does not exceed 4% of total roaming traffic for voice and 6% for data. At the same time, the number of sustainability derogations exhibit a declining trend. According to the 7th BEREC Report on Transparency and Comparability of International Roaming Tariffs, in the period from 31 August 2018 to 31 August 2019, NRAs granted 24 sustainability derogations. Furthermore, as the 2019 Review report concludes, operators that have obtained sustainability derogation have been using it in general with parsimony. As shown in the 2019 SWD, voice and data traffic subject to derogation in the EU does not exceed on average 3% and 1.5% of total roaming traffic respectively. Furthermore, the only country, where voice or data traffic subject to derogation exceeds 12% of total roaming traffic is Lithuania.

2.1.3B. Limitations to ensure a genuine Roam-Like-At-Home experience for end-users

While high roaming prices have been addressed by the current regulation limitations still persist from the end-user perspective (see further details in Annex 5).

B.1. Low perceived quality of service and information failure on quality of service and Roam-Like-At-Home

In order to have a genuine RLAH experience end-users should be able to use the service they pay for while roaming as at home. Quality of services (QoS) is already an integral part of the price-regulated roaming service. The Roaming Regulation only implicitly requires that the end-user has access to the same service abroad in the EU/EEA for the same price, as long as such services can be delivered on the visited network. This has caused uncertainty and inconsistency in the market, even if some NRAs already intervened based on this understanding. For these reasons, the 2019 Review Report concluded that regulatory intervention is necessary to clarify the obligations on the providers related to QoS while roaming and to increase transparency.

As regards QoS, there are three underlying factors to define the problem. Firstly, there is evidence that roaming customers in some cases experience lower QoS than domestically. Secondly, there are indications that some operators have difficulties to ensure access to certain network technologies. Thirdly, in the future QoS will be an increasingly important element of the mobile service offer and there is a need to ensure a future proof regulatory framework for consumers and operators. With 5G services, it might become increasingly important for consumers to know if they will be able to use certain applications and services while roaming due to QoS limitations. Operators should be enabled to offer to end-users the same QoS as they offer at home. The last two factors defining the problem are described in section 2.1.4 C on limitations in addressing innovation needs and access to all network technologies and generations from the operators’ perspective.

Insufficient transparency as regards quality of service (QoS) while roaming may lead to insufficient clarity for consumers. The Roaming Regulation does not include any explicit obligation neither in terms of QoS level nor in QoS transparency. However, it mandates that the end-users have access to the same service while travelling in the EU as long as such service can be delivered on the visited network. Low perceived quality: Available data indicates that QoS indeed is sometimes limited while roaming. BEREC data confirms cases where home operators offered 3G even when 4G was available. Results of a Joint Research Center (JRC) roaming study on QoS 14 also confirm this. Analysis of these results shows that at least in 13 cases the roaming customers had lower QoS than at home and in 15 cases frequently lower QoS compared to other roaming customers on the visited network. 6 of these roaming customers had worse experience in both respects. This indicates that in these 6 cases out of 29, the offered QoS was limited in comparison to what other roaming customers could achieve. Furthermore, 21 customers from 11 Member States at least once had worse roaming experience than at home and worse experience than what was technically possible on at least one of the networks that they visited. For further evidence of the limitations, see Annex 5: Baseline.

In its Opinion on the roaming market 15 (hereinafter ‘BEREC Opinion’), BEREC notes that domestic operators should not deliberately lower the QoS compared to what is offered in the home country.

Transparency: The Roaming Regulation includes an obligation on the domestic provider to ensure that a contract includes the main characteristics of regulated retail roaming service provided. QoS is not listed as one of the parameters that should be specified.

The 2019 Review report concluded that transparency on QoS is not sufficiently ensured in the provision of retail roaming services. According to BEREC, operators in 23 countries do not provide information on their websites about the QoS while roaming. 16  

B.2. Information failure regarding higher prices for value added services

Calls to certain numbers involve different fees because they provide ‘value added services’ (VAS) 17 in addition to the mere electronic communications service: for example, free commercial numbers to subscribe to an insurance service, or a paid number to a technical helpdesk. These VAS numbers can generate additional costs when roaming compared to the cost at home; customers can get blocked, or they may face bill shocks. The Roaming Regulation does not specifically address the use of VAS while roaming. Lack of transparency and high surcharges for VAS are not only related to roaming but also to national circumstances. This limitation is however often larger in roaming scenarios.

On retail level, insufficient transparency on the higher charges applied to calls to VAS numbers and the resulting bill-shocks might erode customers’ confidence in roaming, and may reinforce restricted phone use abroad.

The joint Commission-BEREC online survey 2020 shows that 26.5% of the responding operators report having received complaints from their clients about communications related to VAS while roaming in the EEA. Some complaints concerned significant bills (several hundred euros per case) and unexpectedly high compared to domestic charges for the end-user.

B.3. Failure to provide access to emergency services, as provided domestically

In their home country, end-users with disabilities and other end-users can use alternative means of access to emergency services instead of voice calls 18 (e.g. SMS or apps) while also benefitting from caller location. While roaming, access to emergency services through alternative means of emergency communications or caller location is not ensured for them 19 . Even if such access were ensured, it may not be free of charge as required by the EU law for accessing emergency services domestically. The competence of a national law of the visited country does not extend to other countries to solve this issue. 20

Furthermore, caller location information is not provided consistently for all roaming end-users placing an emergency call. In particular, the very accurate handset-derived location solution that is being successfully deployed in the EU is not available for roaming end-users free of charge as for the national end-users of the visited network. Home operators tend to charge alternative means of access to emergency services at retail level also because of the undifferentiated treatment at wholesale level of the various types of traffic (IP data, SMS).

While the Roaming Regulation ensures that end-users are informed about the cost-free call to “112” when entering another Member State, there is no equivalent provision for alternative means of access to emergency communications. Roaming customers with disabilities are not informed about how to contact emergency services when travelling in another Member State. Relevant EU level associations confirm that the lack of awareness on the means of access to emergency services represents a real bottle-neck to the ability to contact emergency services in the visited country. In the public consultation, the majority of answers (76%), including representative NGOs and three public bodies, indicate the total lack of awareness on the alternative means of access.

2.1.4C. Limitations to ensure access to all network technologies and generations, facilitate innovation and avoid misuse from the operator perspective

C.1. Limitations in addressing innovation needs, ensuring quality of service to end-users while roaming and access to all network technologies and generation for operators

Pursuant to Article 3 of the Roaming Regulation, mobile network operators shall meet all reasonable requests for wholesale roaming access, which should cover access to all network elements and associated facilities, relevant services, software and information systems, necessary for the provision of regulated roaming services to customers.

While the vast majority of operators (97%) 21 claim that they do not limit the QoS/data speed of roaming services to 3G for their customers, BEREC data confirms that 43% (68 mobile operators) offered 3G roaming services even when 4G was available. 22  

There are indications that some operators have difficulties in ensuring access to certain network technologies. This is a precondition for them to be able to offer certain levels of QoS while roaming. In particular, this is important for new technologies.

While 46% of the respondents to the public consultation consider that the current wholesale roaming access obligation is sufficient to ensure access to 4G and 5G, 31% of the respondents do not think that the current obligations are sufficient. In particular MVNOs note that they have experienced long delays in being granted access to 4G networks and fear potential bottlenecks on 5G roaming. Some stakeholders also argue that the current regulation does not take sufficiently into account technological evolution, such as 5G services. There is an inherent risk that access to modern technologies may be limited by visited networks.

In the future QoS will be an increasingly important element of the mobile service offer and there is a need to ensure a future proof regulatory framework for consumers, businesses and operators.

While innovative services are leading to an increase 23 in wholesale access requests for machine-to-machine 24 (M2M) communications, market players are asking for more explicit rules or guidelines governing access requests for permanent roaming for the purposes of connectivity for M2M/Internet of Things (IoT). Indeed, connected machines have emerged recently: GSMA 25 describes 26 that unlike the mature voice/data business from standard customers, many M2M/IoT applications and services are still being defined. It has also to be noted that there exist many non-cellular technologies that are not regulated and compete with cellular ones. It has been estimated that cellular connections account for only 14% of all IoT connections 27 .

The majority of MNOs and a small number of MVNOs offer M2M services. However only 1 out of 2 MNOs and 1 out of 4 MVNOs seek to establish specific agreements for M2M 28 . Even though now NRAs are aware of very few ongoing negotiations 29 , we expect the market to grow with technology evolution and to develop with voluntary agreements, since operators often have an interest to host M2M communication traffic on their networks, including on permanent basis, in order to benefit from the related wholesale revenues (See Annex 6: Review Report evaluation).

The Roaming Regulation does not exclude M2M from its scope and the relevant wholesale roaming access obligations, however operators consider that the current, volume based charging model is not suitable for covering network costs like signaling and location updates, in view of the very low data volumes in M2M communications. According to information available, GSMA is considering alternative charging models for M2M communications, involving potentially a charge per SIM per month. Operators should be able to establish flexible roaming agreements, enabling wholesale roaming services, applying tariff schemes, which are not based on the volume of consumed data but on alternative schemes. There is a need for clarification of the possibility to use alternative, non-volume based tariff structures.

C.2. Difficulties in addressing cost of VAS and combating fraud and misuse

In the BEREC opinion two difficulties concerning calls to VAS numbers were identified: (a) the lack of transparency concerning VAS number ranges and wholesale tariffs and (b) the fraudulent generation of traffic towards international numbers with high termination rates subject to revenue sharing between the operator terminating the call and the value added service provider. These difficulties became more important with the introduction of RLAH and have caused and are causing losses for operators.

Operators typically face higher wholesale roaming rates for calls to VAS, since the regulated rates apply only to the connection but not to the whole tariff that is charged for VAS. Due to the lack of transparency, numbering ranges for VAS often cannot be recognised by an operator, leading to unexpected costs incurred upon reception of wholesale roaming bills. Trying to recover these extra costs at retail level leads to consumer complaints, often forcing operators to absorb any extra wholesale costs from calls to VAS. This situation differs from the typical case of domestic calls, where the issue mainly focuses on calls to premium rate services (and more generally numbers with high termination rates). In the case of roaming however, the higher charges may be caused by all types of VAS, including freephone, shared cost and short codes (except 112).

Revenue sharing fraud, i.e. the artificial generation of traffic towards international destinations is the most common case of fraud, described in the interim report and the BEREC Opinion. While revenue sharing fraud is not specific to roaming, it is exacerbated in a roaming context. The home operator has the contractual relation with the end-user making the call, but in a roaming scenario it is the visited network that connects the calls and hence the home operator does not have control over the treatment of the VAS call.

While these are two distinct problems, any measure that increases transparency of high termination rate number ranges may contribute to mitigating the revenue sharing fraud difficulties. See Annex 5 Baseline for further evidence on the problem.

2.2What are the problem drivers?

The analysis of the evidence supporting the impact assessment identified I) Market failure and insufficient competitive dynamics on the roaming market; II) Regulatory gap linked to the cross-border nature of roaming and III) Regulatory gap in addressing technological development and innovation as main drivers contributing to the problems related to roaming. The remaining part of section 2 describes these drivers.

Figure 6. Intervention logic (drivers-problems)

2.2.1Driver I: Market failure and insufficient competitive dynamics on the roaming market

Before the Roaming Regulation, operators charged wholesale prices much above the actual incurred costs and transferred these high costs to the consumers, through equally high retail prices. Operators had very little incentive to compete for low roaming-retail prices and consumers overall seemed to take retail roaming prices as given and less relevant for their choice of domestic mobile subscription. Accordingly, no wholesale competition was observed before the caps on wholesale roaming were implemented and lowered.

The situation in neighbouring regions confirms that without roaming rules, market dynamics are insufficient to reduce roaming surcharges and to bring wholesale rates closer to costs. For example in the Western Balkans (WB), the level of roaming retail and wholesale prices incentivised decision makers from the region to introduce an EU-based approach via a Regional Roaming Agreement 30 enabling them to reduce roaming charges in the six countries of the region. In Switzerland, despite some efforts to incentivize operators to reduce roaming charges, no widespread reduction has been observed. Also rest of the World (RoW) roaming prices are an indicator of market failure. According to operator data, collected for the BEREC international roaming benchmark report, average wholesale rates charged by EU operators are 5x higher for voice and 4x higher for data, and EU customers roaming in the rest of the world pay surcharges of 50 cent/min for voice and 28 €/GB for data.

Identified market failures in the wholesale roaming market:

Wholesale roaming rates are agreed in biannual bilateral negotiations between “un-equal” contractors. Apart from size and multinational reach, the main driver in these negotiations is the capacity of each operator to exchange traffic, i.e. to balance their retail-outbound traffic they buy with the wholesale-inbound traffic they can serve. This means that there are operators with a weaker negotiation position due to no or little inbound traffic to balance off. Those operators are MVNOs or MNOs with large outbound balance, especially when they are small and not part of a multinational group, and are consequently unable or seriously impaired to negotiate better conditions than those in the cap (see also Annex 3 and Review Report).

In addition, national wholesale markets are by nature oligopolistic as only MNOs can offer roaming services in visited market 31 . This situation creates low competitive pressure and has not changed drastically, as confirmed by the study on technological developments and roaming 32 . (See Annex 8 for details).

Wholesale prices provide further evidence that competition remains imperfect. While they are generally charged well below wholesale caps across Member States, there are differences in the wholesale rates charged between Member States 33 , which can be the result of different underlying costs and/or different competitive environments. At the same time, a number of operators still actually pay wholesale prices at the level of the wholesale cap. So, operators with weaker negotiating power are often obliged to pay prices at or close to the wholesale caps 34 .

Market failures in the retail roaming market:

A significant proportion of end-users has a limited interest in the actual retail cost of using roaming services when deciding on a domestic mobile subscription, as roaming is only used while travelling. According to a Eurobarometer survey of May 2018, only 46% of respondents had travelled at least once to another EU member state in the past 12 months and only 28% had travelled more than once. 35 The regulatory framework for the sector includes provisions to facilitate consumers in making informed choices when choosing or changing operators, through an increased level of transparency of information and specific rules on maximum contract duration and one-day number portability 36 . Despite this, mobile end-users do not switch operator only based on roaming related considerations but rely largely on the domestic offers and how these compare. Before introduction of RLAH 37 , the domestic offers between operators mostly did not differentiate on retail roaming prices but focused rather on domestic offerings.

As a result, retail roaming prices were excessive prior to the Roaming Regulation. These market failures without regulation are further exemplified when looking at retail roaming prices in the rest of the world (outside the EEA) as retail roaming charges for consumption in non-EEA countries remain excessively high. 38  

Some technological developments such as over-the-top (OTT) voice and messaging services, WiFi and eSIM, could exercise competitive pressure to retail roaming services as analyzed in the study on technological developments (SMART 2018/12, see Annex 8). However, while each of the technologies can exercise some competitive pressure, they all face substantial constraints that greatly limit their ability to act as substitutes to retail roaming services in the medium term.

2.2.2Driver II: Regulatory failure linked to the cross-border nature of roaming and inability to solve it nationally

In the national settings, national regulatory authorities (NRAs) are responsible for safeguarding and promoting the interests of end-users. However, the NRAs are not able to control the behavior of the visited network operators, situated in other Member States, on whom those customers depend when using international roaming services. This “jurisdictional” obstacle diminishes the effectiveness of measures taken by Member States based on their residual competence to adopt consumer protection rules.

In the past, international roaming was one of the markets susceptible to ex-ante regulation, defined in the Commission Recommendation of 11 February 2003 on relevant product and service markets. However, it was not possible for national regulatory authorities to effectively regulate roaming services as the regulatory powers under this specific framework relies on identifying undertakings with significant market power, which was not possible for the NRAs to do, given the international and cross-border nature of roaming. Only the introduction of common Union-wide roaming rules enabled to address the market failure and the retail and wholesale prices started to decrease.

2.2.3Driver III: Regulatory failure in addressing Innovation, technological and market developments

While constant innovation in the telecoms sector is welcomed and brings benefits for end-users, it also comes with challenges to the existing regulatory measures. Technological developments bring changes to available network generations and commercial offers, which have the potential to disrupt the end-users’ needs and tariff structures in a way not foreseen by the applicable regulation, creating regulatory gaps.

Consequently, commercial roaming arrangements reflecting current rules on wholesale roaming access (Art. 3 Roaming Regulation) might not be sufficient to address needs linked to innovation and technological developments.

Innovation drives consumers’ needs as regards QoS. 5G is a technological step change for mobile services and it is expected that certain services and applications will require 5G technology. In particular, the deployment of 5G networks and services is expected to enable internet access services with different levels of QoS (QoS classified connections), and specialised services. Lack of effective measures to ensure that operators can gain wholesale access to all network generations and related facilities on equal terms might prevent the users to use innovative services and applications that require 5G connectivity and higher QoS, not only at home but also while roaming.

Technological developments also drive consumers’ needs as regards emergency communication. In particular, the migration from circuit switched solutions (essentially calls) to IP solutions, triggers the need to ensure accessibility and free of charge use by consumers when these solutions are deployed in EU jurisdictions (see relevant problem in Section 2.1.).

Innovation also drives increase in data volume needs, with widespread use of data-heavy applications and services that we might witness in particular with the development of 5G. This trend might create an increase in roaming data consumption that might exacerbate sustainability problems for operators, that need to be addressed through further regulatory measures (see problem B1 Section 2.1). Expected increase in data volumes are considered in the sustainability model, and impacts of possible higher data increases are assessed via the sensitivity scenarios as presented in Section 6 and Annex 4A.

Innovation drives an increasingly important M2M communication market, where we register an exponential increase in the number of Internet of Things (IoT) devices, which often involves permanent roaming. The 2019 Review report notes that the Roaming Regulation does not exclude M2M communications from its scope and that wholesale roaming access obligations apply in case such access is sought for the purposes of M2M communications. It further acknowledges that the relevance of volume-based maximum wholesale charges for low-volume, narrow-band M2M communications requires further attention. M2M often requires the use of roaming services on a permanent level. Although the Roaming Regulation does enable alternative wholesale tariff structures between operators, there is an inherent risk that the measures in place which are foreseen to eliminate the negative effects of permanent roaming, in turn acts as a hindrance to M2M, as confirmed by BEREC.

2.3How will the problem evolve?

As regards the possible impact of technological developments on roaming services, a study on technological developments and roaming has considered developments which could impact competition in wholesale and/or retail roaming markets over the medium term (5-10 years). In particular, it has examined the following developments 39 :

(i) Developments which enable end-users to bypass data roaming or roaming calls and SMS by using alternative technologies to traditional mobile: Wi-Fi and Wi-Fi aggregation services; Over-The-Top (OTT) services; and Rich Communication Services (RCS).

(ii) Technological developments and platforms which could facilitate competition in mobile roaming and cross-border connectivity: Virtual SIM (VSIM); Embedded SIM (eSIM), 5G and 5G network slicing; Voice over LTE (VoLTE); Internet of Things (IoT); Wholesale trading negotiating platforms; and Local data break-out.

(iii) New business models and players entering the roaming space: Multi-MVNO agreements and cross-border MVNOs; entry of equipment, content and service providers into the roaming space.

The study acknowledges that some of the technologies above could exercise a certain competitive pressure on the rates charged for roaming wholesale, especially OTT voice and messaging services and to a lesser degree eSIM and (especially for IoT) 5G and network slicing. However, it concludes that there is no case for significant changes to the regulatory rules applying to international roaming under the current review (without prejudice to review of maximum wholesale rates). These conclusions are confirmed by operators’ responses to the Joint Commission-BEREC online survey 2020 and the public consultation (See Annex 2).

A: How problems from the end-users’ perspective would evolve

Section 5.1.2 further describes end-user rights related to QoS, emergency services and VAS and indicates the European Electronic Communications Code provisions on contract information, transparency, QoS and emergency communications that will affect consumers rights in a roaming environment.

If an operator is not able to negotiate a roaming wholesale agreement covering the latest technology, their end-users will have access only to 4G, also when they roam in areas where 5G is available for domestic users (see limitations in sufficient QoS to consumers as at home in Section 2.1). This could limit end-users from fully profiting the possibility to use 5G and IoT mobile services, for example services connecting vehicles and road infrastructure, which would limit his possibility to avoid congestion and road accidents.

As regards emergency communications, the evolution of 5G networks and the obligations of the European Accessibility Act 40 that provide for simultaneous text and video as the emergency communication for disabled end-users (total conversation 41 ) will further increase the need for end-users with disabilities to be aware of means of access and to be ensured free of charge use when roaming.

B: How problems from the operators’ perspective would evolve

As regards sustainability of RLAH for operators, if wholesale caps are retained at 2022 levels, we can expect that wholesale prices will slightly decline, roughly in line with the current reduction in wholesale rates for voice calls. 42 However, operators that currently pay wholesale rates close or at the level of the wholesale caps will most likely not benefit from such reductions. This will happen for the same reasons why they have not benefitted from such reductions so far, i.e. their lack of negotiating power. As a result, MVNOs and small MNOs with limited inbound traffic will continue to pay wholesale rates close or at the level of the caps. This, coupled with the continuous increase in retail roaming data volumes, is expected to lead to a steady reduction in their sustainability, forcing them to resort to sustainability derogations. For further details see section 5.1.2 on the baseline description also including indications on the possible impact of COVID on roaming volumes.

As regards QoS, the difficulties linked to prolonged negotiation for the access to all technology generations in wholesale roaming agreements might increase with the introduction of 5G, as confirmed by some operators in the public consultation.

As regards M2M, The deployment of 5G networks and services, and especially the development of use cases based on M2M/IoT is expected to have a deep impact in international roaming. The provision of M2M connectivity services or IoT services will require upgrading current roaming agreements to 5G, taking into consideration specific quality characteristics. In practice, commercial voluntary agreements should be able to address this emerging need. In case there would be a future unwillingness on the part of MNOs to conclude such agreements (as host networks), this may impede the development of the M2M/IoT market as well as that of the 5G market.

As regards VAS, numbering ranges are set in the national numbering plans of the Member States and are not harmonised at EU level. Operators may therefore not be able to recognise the numbering ranges for VAS in all countries in advance, which leads to unexpected additional costs incurred upon reception of wholesale roaming bills. Currently, in the case of many VAS numbering ranges termination rates are neither regulated at EU level, nor at national level.

As regards future evolution, it has to be noted that termination rates for VAS, which constitute only one element of the overall revenue in the case of VAS, will not be included in the upcoming delegated regulation setting maximum Union-wide voice fixed and mobile termination rates in 2020 43 , therefore no changes are expected with the entry into force of the delegated act’ in relation to VAS. Since in the roaming scenarios visited networks are unable to identify VAS numbers abroad in the EU, the unexpected additional costs incurred upon reception of wholesale roaming bills will persist.

3Why should the EU act?

3.1Legal basis

The current Roaming Regulation, which will expire on 30 June 2022, is based on Article 114 of the TFEU. This Article is the legal basis for measures adopted in accordance with the ordinary legislative procedure with the aim of establishing or ensuring the proper functioning of the internal market, an area without internal frontiers in which the free movement of goods, persons, services and capital is ensured as foreseen in Art. 26 TFEU. As the present initiative concerns the prolongation and review of the Roaming Regulation, the same legal basis should be used.

According to the case-law of the European Court of Justice, the object of measures adopted on the basis of Article 114 (ex Article 95 TEC) must be to improve the conditions for the establishment and functioning of the internal market 44 . The Union legislature may have recourse to it particularly where there are differences between national rules which are such as to obstruct the fundamental freedoms and thus have a direct effect on the functioning of the internal market 45 or to cause significant distortions of competition 46 .

The following sections explain how the proposed review improves the conditions for the functioning of the internal market, in line with the subsidiarity requirements set by EU Law and the relevant CJEU case-law.

3.2Subsidiarity: Necessity of EU action

In the international roaming, only action by the Union is effective as the problems could not be solved at national, regional or local level. EU action is strictly needed for the enhancing the Single Market for electronic communication. As observed by the Advocate General in the landmark case C-58/08 Vodafone, “the differences in price between calls made within one’s own Member State and those made while roaming could reasonably be regarded as discouraging the use of cross-border services such as roaming. Such discouragement of cross-border activities has the potential to impede the establishment of an internal market in which free movement of goods, services and capital is ensured. Indeed, there is no clearer cross-border activity in the mobile telecoms sector than roaming itself.” 47

This cross-border character justifies the intervention at the EU level because Member State actions cannot by themselves address the issue effectively and National regulatory authorities have accordingly been unable to autonomously tackle this problem 48 .

Also the Court of Justice recognised that, in the past, “the high level of retail charges had been regarded as a persistent problem by NRAs, public authorities and consumer protection associations throughout the Community and that attempts to solve the problem using the existing legal framework had not had the effect of lowering charges”54.

The proposed review includes measures both at wholesale and at retail level. In this respect in the relevant case law the CJEU had found that wholesale regulation of roaming market is compliant with the subsidiarity principle in view of the fact that "the interdependence of retail and wholesale charges for roaming services is considerable, so that any measure seeking to reduce retail charges alone without affecting the level of costs for the wholesale supply of Community-wide roaming services would have been liable to disrupt the smooth functioning of the Community-wide roaming market"55.

The issues addressed by the measures included in the proposed review are strictly linked to the cross border character of roaming, and can potentially result in either discouraging the use of roaming, or creating barriers in the use of mobile services and applications while travelling in the single market, or in disrupting the smooth functioning of the EU-wide roaming market, which, also according to the relevant case law, is an objective that must be pursued and could best be achieved at EU level 49 . The reasons why MSs alone cannot tackle these problems and it is necessary an EU action are reported in section 2.2.2.

A more detailed assessment of the compliance of this proposal with the principle of subsidiarity is reported in the Subsidiarity grid, accompanying the legislative proposal.

3.3Subsidiarity: Added value of EU action

EU level action has clear benefits as confirmed by the results of the public consultation. The vast majority of respondents (including citizens, consumer organizations, and academic institutions) strongly agree that they can enjoy the benefits that the Roaming Regulation aims to bring. 65% of respondents in all respondent groups replied that the Roaming Regulation has significantly promoted the interests of the citizens and businesses in the EU/EEA.

Based on the data collected in the International Roaming BEREC Benchmark Report, in Q3 2019 (i.e. July-September 2019), almost 170 million Europeans roamed abroad to another EU/EEA member state and enjoyed the benefits of RLAH. They generated more than 6.4 billion minutes of voice traffic, more than 2.1 billion SMSs and more than 240 million GB of data traffic that were not subject to any kind of roaming surcharge.

To size the benefits brought to consumers by RLAH we can first of all consider the huge increase in roaming volumes (see also Figure 2 in Section 1) that measures the unleashed and untapped demand for mobile consumption among travelers in the EU. Even if compared with the expected growth in roaming traffic without RLAH, the increase is exponential. Furthermore, the Commission (JRC) has performed an econometric exercise aimed at quantifying the extent to which European consumers have benefited from the implementation of RLAH, that is, the additional Consumer Surplus they got from the regulation. In brief, consumers now pay less for roaming services and use more roaming services. In this analysis, the counterfactual volume increase has been used to compute the change in the consumer surplus, which can be quantified in the range of around 5400 million euros for the 2 years after the implementation of RLAH from 15 June 2017 to 15 June 2019 (see Annex 4B for details).

As regards the impact on competitiveness of EU operators, the analysis shows that European operators are generally not hindered in their abilities to compete in the international setting as a result of RLAH.

Firstly, the analysis of the domestic revenues before and after the entry into force of RLAH reveals that, on average, the increase of domestic revenues has been around 25% 50 . A similar analysis on domestic market shares (ratio between subscribers for each operator and number of subscribers in the country) shows that market shares have not changed much within the Member States. Had certain operators been more (negatively) effected by RLAH, they could have resorted to higher domestic prices, which in turn would incentivise customers to change domestic provider. This has not been observed.

Second, the median ratio of roaming revenues over domestic revenues in 2016 Q4 (i.e. before the introduction of RLAH) was 3.3%. From this, it appears highly unlikely that a regulatory change affecting less than 3.5% of the revenues could have the potential to hamper operators' profitability overall 51 . This is further confirmed, as the structure of the European mobile market has remained mainly unchanged since 2017 with no big mergers observed or larger operators exiting the market.

Third, the current global pandemic has drastically reduced the possibilities for travelling, and this resulted in heavily declining wholesale roaming revenues. Operators have indicated this as a negative impact of the pandemic, which shows that, in general, hosting roaming customers is a benefit to the operators. 52  

In addition to the above the proposed review aims at further improving the functioning of internal market also contributing to the possibility for end-users to have a genuine “roam like at home experience” and addressing the possible obstacles to this goal.

The added value of the different EU actions proposed in the review of the Roaming Regulation is explained in the following paragraphs:

(I)Reviewed wholesale measures aim to ensure the cost recovery for all operators and to address the sustainability challenge, thus representing the adequate balance between the needs of the different Member States (in- and outbounders) that only an EU instrument can ensure. The proposed solution aims to also ensure flexibility and a future-proof approach. Previous attempts to regulate roaming through ex ante regulation have not been successful and failed to provide appropriate solutions as illustrated in Section 2.2

(II)New measures on VAS aim to establish at EU level solutions to provide single information point on the relevant VAS number ranges. Such a common solution will also enable additional transparency for roaming end-users and, in the future, more effective actions against misuse and fraud.

(III)Introducing new measures on QoS and on access to emergency services: the value added of an EU intervention is strictly related to the cross-border nature of the underlying problems as better explained in the previous section. With regard to both kind of measures, national solutions are limited, since the QoS depends on the provided connectivity/service of the visited network. 

4Objectives: What is to be achieved?

4.1General objectives

A. Ensure sustainable provision of RLAH

At wholesale level the overall objective is to ensure, through competition and regulated wholesale caps, sustainable provision of RLAH so that operators domestic pricing models are not affected. This objective is essential to the prolongation of RLAH.

This first objective aims to enhance the sustainability of RLAH while ensuring recovery of their costs by network operators (MNOs) who offer roaming at wholesale level. Ensuring cost recovery at wholesale level, preserves incentives to invest in new networks and avoids distortion of domestic competition in the visited markets. It is consistent with the European Electronic Communication Code (EECC) objectives, the regulatory framework for electronic communications, which promotes connectivity as the most fundamental building block of the digital transformation and focuses on infrastructure competition and return on investment for operators. It is also a complement to the connectivity policy 53 to accelerate rollout of 5G and fiber networks, reduce the cost of network deployment 54 and provide operators with timely access to pioneer spectrum for 5G.

B. Ensure a genuine RLAH experience for end-users

At retail level the overall objective is to give roaming customers a genuine RLAH experience and increase transparency. The Roaming Regulation establishes a common harmonised approach that ensures that roaming customers do not pay additional roaming charges for Union-wide roaming when periodically travelling within the Union. The enhanced RLAH experience aims to ensure that consumers can benefit from the same quality of service while roaming as at home, a high level of transparency to avoid bill-shocks, and access to emergency services for all in the same way as at home. The Regulation is built on the general objective of giving end-users the confidence to stay connected when they travel within the Union, and to become a driver of convergent pricing and other conditions in the Union.

This objective complements and is supported by the European Electronic Communication Code (EECC), measures that not only aims to enable high connectivity and 5G deployment for the benefit of all Europeans, but also to ensure effective protection of consumers in the context of e-communications, boosting their choice through an increased level of transparency of information and specific rules on maximum contract duration and number portability. These provisions play a fundamental role in facilitating consumers in making informed choices when choosing or changing operators, depending on the services and prices they offer. The rules further ensure the free of charge harmonised single European emergency number and caller location.

Furthermore roaming contributes to Shaping Europe’s digital future. The digital transition should work for all, putting people first and opening new opportunities for business. The COVID pandemic has proven that technology that works for people is highly important. The cross-border connectivity further benefits the creation of a European social, educational, cultural and entrepreneurial area based on the mobility of individuals and digital data that facilitates communication between people. This will further strengthen the efforts towards the creation of a Digital Single Market where free movement of persons, services and capital is ensured, where the individuals and businesses can seamlessly access and engage in online activities under conditions of fair competition, and a high level of consumer and personal data protection.

C. Facilitate innovation, ensure access to all network technologies and generations , avoid misuse and reduce burden

The wholesale regulation should facilitate innovation and technological developments, and avoid misuse related to VAS. The review also aims to simplify and reduce the burden on operators and other stakeholders.

There is a need to ensure a future proof regulatory framework for consumers, businesses and operators to facilitate the access to next generation connectivity and modern technologies. Europe is investing in more strategic capacities, that allow us to develop and use digital solutions at scale and to strive for interoperability in key digital infrastructures, such as extensive 5G (and future 6G) networks. In the future QoS will be an increasingly important element of the mobile service offer and there is a need to ensure a future proof regulatory framework for consumers and operators. With 5G services, it will become increasingly important for consumers to know if they are able to use certain applications and services while roaming due to QoS limitations. For a genuine RLAH experience, operators should offer to end-users the same QoS as they offer at home.  

Roaming is also relevant to facilitate innovation, benefiting users of connected objects. Indeed, the free movement of goods entails that objects connected by an operator in one Member State can be sold or used in another Member State. Users sometimes carry their connected objects with them when travelling. Therefore, machine-to-machine (M2M) connectivity is by nature a cross-border market. Since M2M is a critical enabler of 5G 55 and Internet of Things 56 , facilitating M2M roaming is important to digitise EU industry and enable EU policies for sectors including health, the environment, transport and energy.

In the transport sector, roaming agreements might also favor the organization of business models conducive to innovation for Connected cars. The roaming initiative is therefore expected to contribute to the relevant EU Policy in the area, in particular for Cooperative Intelligent Transport Systems, which aim to increase road safety. It is also expected to help industrial organization to deliver cooperative, connected and automated driving. 57

The Roaming Regulation further complements actions for the European Green Deal, enabling the use of “green” applications in the roaming situation, supporting energy efficiency and the use of smart mobility solutions, enabling environmentally friendly behavior by individuals.

In the health sector, roaming facilitates the efforts to fight against COVID-19 using tracing applications. Indeed, the relevant Commission Recommendation 58 recommends that Member States ensure interoperability of health applications in cross-border scenarios. More generally, this initiative is an enabler of the European Health Union 59 since it facilitates the availability of health services when travelling.

4.2Specific objectives

The general objectives oriented towards consumers and operators and other businesses are further divided into specific objectives following the overall Intervention logic.

Figure 7. Intervention logic (drivers-problems-general and specific objectives)

4.2.1A. Ensure sustainable provision of RLAH

Objective A1: Ensure enhanced sustainability of RLAH and cost recovery at wholesale level preserving incentives to invest in visited networks and avoiding distortion of domestic competition in visited markets. In order to ensure retail roaming services at domestic retail prices, wholesale roaming inputs must be available at a level that allows home operators to provide RLAH. This should enable a sustainable prohibition of retail roaming surcharges for operators in the EU, avoiding distortions on the home markets.

4.2.2B. Ensure a genuine RLAH experience for end-users

Objective B1: Increase transparency and ensure the same quality of service in roaming as at home. Consumers and businesses should be well informed about QoS, so that they know what Quality of Service they can reasonably expect while roaming. While end-users in some cases experience lower quality of service than domestically, they may not always have sufficient understanding about the level of QoS they could reasonably expect, because of insufficient transparency concerning the provided QoS while roaming. Roaming customers should be able to use the mobile services that they pay for, and innovative services and applications that require 5G connectivity and higher QoS like at home, to the largest extent possible. This objective clarifies the obligations in terms of QoS and RLAH at retail level, addressing the end-users’ perspective. This specific objective is complemented by specific objective C1 addressing the same problem from operators’ perspective clarifying obligations at wholesale level.

Objective B2: Increase transparency regarding VAS to avoid bill-shock. Calls to certain numbers can generate additional costs when roaming in the EU/EEA compared to making the same calls at home. Mobile operators need to transparently inform end-users in the welcome message that roaming customers receive when connected to another EU/EEA operator, about the possibility to incur higher costs when calling certain numbers while roaming. Well informed end-users have less incentive to restrict their use of roaming services in fear of bill-shock.

Objective B3: Ensure access to emergency services as provided at home. End-users should be able to enjoy free of charge access to emergency services while roaming in the EU/EEA as this access is also ensured domestically, especially for end-users with disabilities. All end-users need to be adequately informed about the possibility to access to emergency services through emergency communications, including alternative means of access, while roaming in the EU/EEA.

4.2.3C. Ensure the same QoS as at home and access to all network technologies and generations, facilitate innovation and avoid misuse/fraud from the operator perspective

Objective C1: Ensure the same quality of service while roaming as domestically and access to all network technologies and generations for operators and respond to technological and business developments. The home operators should not deliberately lower the QoS when roaming compared to the QoS offered in the home country. In addition, operators should facilitate access to all network technologies and generations on wholesale level, and ensure interoperability of roaming services on all available network generations, including 5G. To ensure that operators can offer, and end-users can efficiently enjoy, innovative services and applications that require 5G connectivity and higher QoS, wholesale network access to modern technologies needs to be facilitated to avoid limitations and delays. Adequate wholesale roaming price schemes are necessary to enable M2M roaming service, which require permanent roaming.

This objective clarifies the obligations in terms of QoS at wholesale level, addressing operators’ perspective. Regulatory clarity on wholesale level is a precondition for resolving the problem at retail level, but also for meeting innovation needs and future proofing the Regulation. In the future QoS will be an increasingly important element of the mobile service offer and there is a need to ensure a future proof regulatory framework for consumers and operators. Clarifying the preconditions for the operator to effectively establish wholesale roaming agreements will facilitate both QoS like at home for roaming customers, and meet the specific objectives of innovation. This specific objective therefore complements specific objective B1, addressing the same QoS problem from end-users’ perspective clarifying obligations at retail level.

Objective C2: Higher level of transparency on wholesale level for VAS to reduce misuse and fraud. A higher level of transparency on the value added services’ numbers and charges applied to them needs to be ensured for mobile operators at wholesale level to hinder fraudulent use of roaming services.

Objective C3: REFIT: Simplify and improve efficiency of the Regulation. The repeal of inefficient obligations, rationalized monitoring obligations on operators and NRAs and lighter procedures for the revision of wholesale caps, should be considered to reduce the administrative burden.

5What are the available policy options?

5.1What is the baseline from which options are assessed?

5.1.1Choice of the Legislative Instrument

A regulation is the only suitable legislative instrument to effectively ensure that roaming customers continue benefitting from RLAH. The tangible benefits that citizens and enterprises currently enjoy as a result of RLAH should be ensured simultaneously in the entire EU, imposing equal obligations to all operators and fostering legal certainty.

The regulatory intervention in the form of a review of the existing regulation is therefore necessary to ensure the continuation and the functioning of a single telecom market across the Union. The Roaming Regulation has been substantially amended several times so that recasting is necessary in the interest of clarity.

5.1.2Option 1 - Baseline: Prolong the Roaming Regulation without amendments

Under the baseline option, the Roaming Regulation would be prolonged, maintaining its current provisions both at retail and at wholesale level. Given that the 2019 Review report has confirmed the need of the current approach, deregulation of the market is discarded (see Section 5.3 on reasons for discarding options).

Annex 5 complements Section 5.1 and for each content area presents the measures already in place that would be prolonged. It also presents data assessing the initial situation, supporting evidence of existing problems, referring to the public consultation and survey feedback collected by the Commission.

Baseline A. Sustainable provision of RLAH

A1. Sustainability of RLAH and cost recovery at wholesale level preserving incentives to invest in visited networks and avoiding distortion of domestic competition in visited markets.

Under the baseline scenario, RLAH will continue unchanged from 30 June 2022 onwards. The caps applicable on 30/6/2022 will continue to apply (3.2 €cents/min, 1 €cent/SMS and 2.5 €/GB). The two safeguard mechanisms (fair use and sustainability) 60 , provided in the Roaming Regulation remain in place.

Regarding the application of fair use policies, these have generally been stable and have not exceeded 4% of total roaming traffic for voice and 6% for data. Derogations are in general concentrated in only a few countries. The 2019 Review report, observes a continuous decrease in the number of derogations requested and granted, following the decline in wholesale caps. According to data collected by BEREC, sustainability derogations were granted to 8 MNOs and 16 MVNOs between 31 August 2018 and 31 August 2019 61 .

How the situation would evolve under the baseline scenario

The rapid increase of roaming volumes (especially data) is expected to create sustainability challenges for several operators, especially outbounders and MVNOs. This is reflected in the results of the public consultation (see Annex 2), the BEREC opinion and the inputs collected from the market through the joint Commission-BEREC on-line survey.

Under the baseline scenario, the current measures at retail level (fair use policy and sustainability derogation) as well as on wholesale level (wholesale caps) would remain in place.

However, the current wholesale caps no longer appear to be fit for purpose, since maintaining the current caps would lead to non-negligible sustainability challenges. According to the sustainability analysis (see section 6.1 and Annex 4A), the provision of RLAH would be unsustainable for 27% of operators in 2023. This would likely imply that more operators (especially MVNOs) are likely to request and obtain sustainability derogations, to apply a surcharge to (parts of) roaming traffic.

The COVID pandemic and the baseline scenario

The COVID pandemic has a major impact on the tourism industry in 2020. Eurostat data on the nights spent at tourist accommodation establishments by foreign residents per member state show a sharp year-to-year reduction for the first period of the lock down (96.6% for April 2020 and 94.7% for May 2020) that slowly recovered with the partial lifting of measures (85.4% for June 2020 and 64.6% for July 2020). The overall year-to-year reduction for the first 7 months of 2020 is estimated to 56%.This massive collapse of tourism indicates a strong impact of the COVID pandemic to international roaming traffic.

While we do not have readily available data for Q2 2020 (the first peak of the pandemic) we can reasonably expect a significant reduction to roaming traffic, hence also to roaming revenues and costs. For inbounder operators 62 this means reduced roaming profits, following the drop in wholesale roaming revenues. On the other hand, for outbounder operators 63 and MVNOs it means reduced sustainability challenges, due to lower outbound roaming traffic (and the ensuing wholesale costs). This impact could well extend into 2021, depending on the roll-out of vaccination, the removal of travel restrictions and any new waves of the pandemic.

Given the review of the Roaming Regulation will enter into force in June 2022, the baseline should focus on values from 2022 onwards. The medium-long term effect of COVID is however unclear. The OECD “Tourism policy responses to the coronavirus” warns that travel restrictions and containment measures are likely to be in place for longer, and are expected to be lifted only gradually, with the possibility of reversal should new waves occur. Demand-side recovery will also take some time, given the interlinked consequences of the economic and health crises. This leads to growing expectations that recovery to pre-crisis levels may take two years or more. Yet, all the aforementioned views point towards an at least partial recovery, before the new Regulation enters into force in mid 2022.

At the same time, the pandemic has led to a substantial increase of digital services and an ensuing increase in connectivity demand 64 . The increased digital interaction is expected to affect the overall data usage and contribute to an accelerated increase in mobile roaming data consumption per travelling consumer with a potential long-term effect. This may at least partially compensate for any delay in the recovery of the tourism industry and any long-term reduction of business travelling, also linked to the increased use of digital channels.

On the other hand, we do not have any evidence or indication that telecom operators have changed their business practices (including on wholesale roaming tariffs), because of the COVID crisis. In view of the above, we do not expect a disruptive impact, which would affect the proposed policy options or the assumptions on the parameters used in the sustainability analysis.

As regards the impact from COVID on the roaming wholesale caps and the estimated costs of providing roaming wholesale services, it is recalled that mobile operators do not dimension their networks to solely serve roaming customers but rather dimension to domestic needs. If indeed operators in a Member State would need to take special account of roaming customers in the network dimensioning, this is termed a seasonal Member State, where peak consumption (and needed dimensioning of the network) is, to some degree, driven by roaming customers. If this is the case, the Member State is termed seasonal and this is taken into account in the modelling exercise performed by the consultants. For five Member States (Croatia, France, Greece, Malta and Spain), seasonality has been taken into account, meaning that these Member States have shown a need to further update their networks to facilitate increased consumption from roaming customers. Therefore, operators in these Member States appear to have some constraints coming from roaming customers, which in the case of COVID would imply that a lack of roaming customers could have some effect on network dimensioning. Looking at the estimates from the Axon cost model for these Member States reveals that current costs estimated is well below the maximum caps proposed in both 2022 and 2025, indicating that the caps proposed should also ensure cost recovery in these Member States despite uncertainties for roaming consumption.

For the 22 other Member States, it has not been shown that roaming customers exerts pressure on the capacity required in the network. Therefore, the networks constructed in the 22 Member States without seasonality are built to meet domestic demand, indicating that declining roaming customers as a result of COVID would equally not impact the estimated costs of providing roaming wholesale services.

The sensitivity analysis (Annex 4) developed some additional scenarios, which seek to examine how a prolonged impact of the COVID pandemic on international travelling could influence the ability of roaming providers to offer RLAH services in a sustainable manner. The analysis reflects that COVID ameliorates the sustainability challenges for outbounder operators and MVNOs in all four COVID-19 scenarios (COVID-19 High impact, COVID-19 Medium impact, COVID-19 Low impact and COVID-19 Minimum impact). In general the COVID-19 scenarios improvements does not surpass the results under the high sustainability scenario presented in Annex 4.

Baseline B. Ensure a genuine RLAH from an end-user perspective

B1. Perceived Quality of Service and transparency 

The Roaming Regulation does not explicitly ensure that roaming end-users have access to retail roaming services of the same quality as at home (domestic QoS) even if the QoS is considered as an integral part of the price-regulated product. BEREC reports that 3 out of 30 NRAs have initiated procedures against operators not ensuring same QoS.

Article 6e (4) of the Roaming Regulation includes an obligation on the roaming provider to ensure that a contract which includes any type of regulated retail roaming service specifies the main characteristics of that service. This transparency obligation does not mention explicitly QoS.

At wholesale level, i.e. between operators, the regulation simply requires that an operator shall provide all necessary network elements along with roaming access, but does not require that the operator provides certain levels of QoS.

How the situation would evolve under the baseline scenario:

Customers would continue to be uninformed about the expected level of QoS for retail roaming services. Contracts are an important tool for end-users to ensure transparency of information and legal certainty. With the upcoming deployment of 5G retail services, and future technological developments, it is important for roaming customers to be well informed, since the lack of appropriate QoS might hinder the usage of certain applications while roaming (The business side of this issue is further explained in Section 2.1. problem C1).

The fragmented approach among NRAs on QoS monitoring as well as lack of transparency on QoS is expected to continue.

B2. Transparency on higher prices for value added services. At retail level, VAS can be subject to surcharges while roaming or can be blocked. According to the BEREC, operators indicate that VAS are excluded from wholesale contract negotiations and that the visited operator decides whether to block access to premium services or to impose additional charges for such services. This means for example that when calling a freephone number abroad consumers may incur surcharges. According to operators, VAS numbering ranges cannot be recognized in all countries in advance. This leads to unexpected termination costs and/or customer experience degradation. In addition, the fact that in the roaming situation is the visited operator that decides on the treatment of VAS prevents home operators from giving their customers transparent information on charges, as they do not know the associated wholesale costs.

Consumer complaints are currently handled on a case-by-case basis in the absence of clear rules with regard to the regulatory treatment of VAS in roaming scenarios.

How the situation would evolve under the baseline scenario

If no further action is taken, the difficulties as confirmed by the operators and consumer complaints are to remain or even increase following a general trend of growing use of roaming. The number of related complaints reported by NRAs remains limited but is increasing visibly and calls for measures 65 . (See Annex 5 for details) This can certainly have a negative impact on the end-users’ confidence in the RLAH regime. The lack of transparency at wholesale level in particular cannot be adequately tackled at national level.

B3. Access to emergency services

Subject to Article 14 of the Roaming Regulation, end-users shall receive information on access to emergency services by dialing the European emergency number 112 free of charge. The Roaming Regulation does not include an obligation to inform end-users with disabilities about access to alternative emergency communications that are deployed at national level (these alternative means of access are not harmonised at EU level).

Provision of caller location is mandated through the Universal Service Directive and the Electronic Communications Code, but there is no obligation in the Roaming Regulation to ensure the exchange of the technical and regulatory information between the roaming partners, ensuring that caller location provision is free of charge for the end-user.

Caller location is the most important contextual information that allows emergency services to locate end-users and intervene efficiently. Roaming customers are at higher risk than domestic customers to not be able to determine their location when they request the assistance of emergency services. Therefore, the instant provision of accurate caller location information is all the more important in case of roaming end-users.

Alternative means of access are not harmonized 66 , as a consequence, end-users with disabilities are not aware of access means and sometimes do not have access to emergency services while roaming. Disabled end-users are more at risk of not being informed about the means of access and, eventually, not having access to emergency services. The latest data provided by Member States (September 2020) indicate that the visited Member States do not have the jurisdiction or monitoring capability to ensure that the use of the means of access deployed in their jurisdiction is not charged by the home operator.

How the situation would evolve under the baseline scenario

The visited and the home operators do not systematically exchange relevant technical data to ensure proper, free of charge functioning of emergency communication and caller location (handset derived) for the end-user. As a consequence, home operators do not recognize emergency communications traffic and charge end-users.

In view of the fact that IP-based communication, using real-time text and multimedia is expected to be deployed in the medium term (as mandated by the European Accessibility Act), the lack of implementation of these solutions at roaming wholesale level could lead to a lack of access to emergency services (including when a prepaid customer exhausts their credit), lack of caller location and retail charges.

Baseline C. QoS, access to networks, innovation, and avoid misuse/fraud from the operator perspective

C1. QoS while roaming, innovation and access to networks 

Article 3 of the Roaming Regulation obliges mobile network operators to meet all reasonable requests for wholesale roaming access, which shall cover access to all network elements and associated facilities, relevant services, software and information systems necessary for the provision of regulated roaming services to customers. The Roaming Regulation Article 16 (5) stipulates that end-to-end connectivity and interoperability of roaming services has to be ensured.

These provision aim to ensure that mobile operators can deliver retail roaming services to their customer when they travel. However, the current rules do not sufficiently ensure that operators should request and be granted access to the technologies that will enable them to provide the roaming services with the same QoS as domestically. There is still a small number of 3G-only wholesale agreements. According to the joint Commission-BEREC online survey 26% of MVNOs have 3G-only roaming available either in certain specified countries in the EU (4%) or in certain networks in certain countries in the EU (14%) or in general in the EU (8%).

In the Regulation, the wholesale roaming access obligations and wholesale data price cap apply in case such access is sought for the purposes of M2M communications as long as they are charged on a per unit, volume-based, basis like person-to-person communications. Permanent roaming is not prohibited as such by the regulation and can be agreed by two roaming partners in the wholesale roaming contract.

How the situation would evolve under the baseline scenario

While the available data and consultation results indicate that operators in general do not deliberately lower the QoS for their roaming customers, there is evidence that some operators, in particular MVNOs have difficulties in gaining access to 4G networks. These observed difficulties in ensuring access to different network generations (in particular 4G), may continue. For the deployment of commercial 5G services, difficulties in ensuring access presents an even greater risk.

The baseline scenario might not sufficiently ensure that operators will have access to modern network technologies and be able to offer e.g. 5G services to their end-users when roaming. Stakeholders interpret the wholesale access obligations differently. Many stakeholders, in particular MNOs, consider that the current wholesale access rules are technology neutral. However, in the public consultation MNOs expressed that modern technologies, such as 5G and IoT narrowband do not fall under the scope of the rules because these technologies are not dedicated to assure retail roaming mobile services and are connected with huge investments costs. These doubts may limit the access of these technologies, posing a risk to innovation and competition. Unless access is ensured, operators and service providers will not be able to compete on the market.

Moreover, the 2019 Review report concludes that 5G technologies are likely to change the nature of roaming services in the long term. Among others, it could potentially affect the commercial model applied, e.g. basing pricing on bandwidth as opposed to usage. 5G could also provide options for MNOs and MVNOs to use access agreements as an alternative to traditional roaming, allowing them to enhance their flexibility on service differentiation (latency, security etc), which could prove to be very important for certain vertical use cases.

The new categories of use cases offered by the 5G technology concern, on the one hand, mission-critical cases such as remote healthcare or remote machine control and, on the other, massive object connectivity to get data from sensors in transport, energy, and environment. These new applications of communication networks are expected to be reflected in future roaming agreements.

Doubts on applicability of roaming rules to M2M might slow down the conclusion of wholesale agreements addressing the M2M emerging market needs.

C2. Transparency on VAS, misuse and fraud.

The general provisions in the Roaming Regulation enable the termination of wholesale roaming access agreements, in case of anomalous or abusive usage of roaming services on retail level.

VAS as such fall outside the scope of the Roaming Regulation, only the tariff component corresponding to the connection to such services is subject to the RLAH rule 67 . Operators indicate that VAS are generally excluded from wholesale contract negotiations and that VAS numbering ranges cannot be recognized in all countries in advance, which may lead to unexpected costs at the wholesale level. Some operators report having taken measures to tackle this situation, including negotiation of wholesale agreements, obtaining information about numbering ranges of other EEA countries, and blocking access to VAS to their customers while roaming.

Where operators are victims of misuse/fraud linked to VAS, an NRA would still be mandated by the Roaming rules to require immediate cessation of a breach of the obligations set out in this regulation, and to the right of the visited network operator to apply adequate measures in order to combat fraud.

According to the 2020 joint Commission-BEREC online survey, more than 30% of MNOs have incurred extra costs resulting from unexpected wholesale charges for communications related to VAS by their customers while roaming in the EEA. Almost half of the MNOs reported being aware of abusive use of SIM cards in voice and/or SMS roaming communications. More than 40% of these operators report about considerable losses linked to fraud and misuses related to value added services with a median of € 70,000.

How the situation would evolve under the baseline scenario:

Operators will continue to be uninformed about VAS number ranges subject to higher termination rates. In the absence of measures promoting transparency, operators will continue to face substantial losses. Therefore, they will continue to take unilateral measures, mostly through fragmented efforts to obtain information about VAS number ranges in other EEA countries but also by blocking calls to VAS, which also leads to consumer complaints.

5.2Description of the policy options

The analysis of alterative measures to the baseline was firstly developed per thematic area, covering sustainability (A), Quality of Service and innovation (B1 and C1), value added services (B2 and C2) and emergency communications (B3), comparing all possible alternative measures to address problems for each thematic area (including measures that were later discarded) and assessing their possible impacts and benefits, considering synergies between measures. Based on this preliminary analysis the aggregated options presented below were identified. The discarded alternative measures per each objective area are presented in Annex 9.

5.2.1Option 2: Continuity, clarifications, increased transparency and competition

Option 2A. Sustainable provision of RLAH

A1 – Enhance operators’ capacity to sustain RLAH while respecting cost recovery for network operators offering wholesale roaming access. The level of wholesale caps would remain at the level valid until 30 June 2022 for calls made, SMS messages and data at the respective values of 0.032 € per minute, 0.01 € per SMS, and 2.5 € per GB.

In addition, it is proposed to complement the price caps by encouraging emerging possibilities for the operators to trade wholesale roaming traffic in a non-discriminatory way. This could prove to improve competition in the roaming wholesale market and is as such welcomed by the Commission, that will keep monitoring these developments and assess the situation in the context of the report by the end of 2025.

Option 2B. Ensure a genuine RLAH

B1 – Increase transparency regarding QoS: Operators should increase transparency regarding QoS, by providing clear information to the end-user, e.g. in the contract about the QoS that the end-user can reasonably expect, while roaming in the EU.

In the public consultation 63% of the respondents, across the stakeholder groups, replied that it would be very relevant to include a transparency obligation on QoS. It is supported in particular by citizens and consumer organisations, while M(V)NOs are less positive to such an obligation (15 out of 34 do not think it is relevant and 6 are positive). BEREC supports additional transparency measures regarding QoS in roaming. 68  

B2. - Increase transparency regarding VAS: Operators should increase transparency regarding the risk for bill-shocks from calls to VAS number, e.g. by providing a clear warning in all contracts that include roaming services and explanations about the types of services that may be subject to increased costs.

B3 - Increase transparency regarding access to emergency services: Operators should increase end-user awareness about available means of access to emergency services, by including relevant information in the automated “Welcome SMS”. This transparency measure would require that Member States make relevant information on alternative means of access to emergency services readily available.

Option 2C. QoS, access to networks, facilitate innovation and avoid misuse/fraud

C1 – Clarify technology neutral access obligations for wholesale roaming services. The basic principle of neutrality implies that operators should remain neutral towards technology when granting access to their networks. This clarification would imply that the visited operator has to grant access to any requested network generation as long as the request is reasonable and hereby facilitate access to all network generations, including modern technologies. The visited MNO should not refuse an access request or impede the access process to a certain network technology e.g. 4G and instead offer only access to 3G, or 4G instead of 5G.

It would further enable home operators to access the same networks generations that they offer domestically and hereby tackle the issue of some M(V)NOs only granted access to 3G services. This measure is supported by BEREC. 69  

C2 - Increase transparency on wholesale level for VAS. Option 2 proposes an obligation for each Member State to publish national VAS number ranges.

C3 – Introduce a minimum level of simplification by removing obligations and repealing acts that have become redundant with RLAH 70 . In addition, the burden on operators outside the Eurozone would decrease by aligning the revision mechanisms for maximum charges linked to the use of currencies other than the Euro that are used for roaming charges and intra-EU communications charges 71 . Measure requested and supported by some authorities outside the Eurozone.

5.2.2Option 3: Sustainable and genuine RLAH

This option would consist of proposing a review and prolongation of the Roaming Regulation (in the form of a recast). It would contain the similar provisions at retail and wholesale level as the Roaming Regulation in force, but with essential provisions to improve sustainability of operators to provide RLAH through reduced caps at wholesale level (A1 below).

Some additional important enhancements could be considered to allow a genuine RLAH experience for end-users and to respond to technological and business developments (B and C below)

Option 3A. Sustainable provision of RLAH

A1. Ensure enhanced sustainability of RLAH and cost recovery at wholesale level - Reduction of caps and measures to promote competition

EU-wide wholesale roaming caps would be set at lower levels than the caps valid until 30 June 2022 for calls made, SMS messages and data. In line with the policy choice made already in the Roaming Regulation in force, the new levels of the caps would allow recovery of costs of providing wholesale roaming services in all Member States, including a relevant share of joint and common costs. Reducing the caps would also affect elements that are determined based on the level of wholesale caps, as the maximum level of surcharges imposed under the fair use policy mechanism and the sustainability derogations as well as the data allowance of the fair use policy (FUP) on open data bundles and pre-paid limits 72 .

The caps have been defined considering the estimates of the costs for a hypothetical efficient operator in the highest cost Member States, using a variety of evidence sources such as the cost model developed for the Commission by Axon Partners, assumptions underlying regulated termination rates for voice evidence from domestic and roaming wholesale markets and from domestic retail prices.

This initiative aims to strike a balance between ensuring cost recovery and minimizing sustainability challenges. To reconcile the two objectives this option consists of a two-step reduction of caps, as indicated in Table 1, the proposed caps are derived from the estimated costs from the cost model (see Annex 4C for a description of the model and an outline of the results). Overall, the cost model indicates room for proposing a decrease of the three applicable caps. Across wholesale voice, data and SMS, the current caps are well-above the estimated costs in all Member States and the cost model estimates decreasing costs in 2022-2025 for the provision of wholesale voice and data services.

For data roaming services, from 1 January 2022 until 1 July 2022, the applicable wholesale cap stands at 2.5 EUR/GB. The highest estimated cost from the cost model including transit in 2022 is 1.74 EUR/GB. In Q1 2020, the average wholesale price charged has been 1.53 EUR/GB but ranges in the member states from 1 EUR/GB to more than 2.5 EUR/GB.

Regarding voice roaming services the wholesale cap has been 0.032 EUR/min since 15 June 2017 (and will remain applicable until 30 June 2022). The cost model estimates a decreasing cost from 2022-2025. Taking also into account the forthcoming Eurorate for mobile termination 73 , the highest estimated cost decreases from 0.022 EUR/min in 2022 to 0.0184 EUR/min in 2025.

The wholesale cap setting the maximum charge for one SMS while roaming has been constant since the introduction of Roam-Like-at-Home (15 July 2017), at 0.01 EUR/SMS. The cost model estimates a fairly constant cost for delivering this service in the modelled period of just under 0.003 EUR/SMS for the period 2022-2025.Considering the above observations, this option proposes a continuation of the decreasing cap observed since 2017. Specifically, the option proposes to set a two-step glide path for the caps for data, voice and SMS. The intention is to accommodate the decreasing costs observed in the cost model whilst gradually reducing the caps to minimize disruptions for the operators.

A two-step glide path therefore balances the two objectives of ensuring sustainability and allowing recovery of costs for all operators. It also allows for a gradual reduction to the estimated efficient costs of providing the relevant service. Further, for voice roaming service, the two-step glide path takes into account the decreasing costs for termination as set by the delegated act on maximum Union-wide termination rate.

Specifically for data, the cap of 2 EUR/GB proposed in 2022 is slightly above the maximum efficient cost of 1.74 EUR/GB estimated for the same year. This cap is proposed to balance the transition to the cost proposed in 2025 of 1.5 EUR/GB and ensure that the operators have sufficient time to negotiate wholesale agreements reflecting the decreasing caps. This cap proposed will ensure a gradual reduction to the cap proposed in 2025 in line with the reductions of data caps seen in the previous years.

Summarising the elements outlined above, Table 1 presents the two-step glide path for wholesale roaming data, voice and SMS.

Table 1: Wholesale caps glide path from 2022 onwards

The preparatory work for this initiative has assessed the possibility to present alternative options for the new wholesale caps, however given the analysis presented above and in particular the need to ensure cost recovery leaving also some room for negotiation, those alternatives were not considered relevant for the impact assessment.

For further updates of wholesale caps beyond 2026 based on an updated cost model see proposed possibility to amend the caps through a delegated act under the simplification/REFIT measure C3 presented below.

The Commission will monitor pro- competitive developments linked to non-discriminatory trading (as discussed in option 2, see also Section 9) and assess the situation in the context of the report by the end of 2025 and consider additional measures, if deemed necessary.

The majority of respondents in the public consultation agree or strongly agree that the EU intervention had a positive effect in ensuring the sustainability of the wholesale roaming markets for voice, data and SMS, compared to what Member States could achieve acting alone. The majority of M(V)NOs agree (28 out of 52 respondents), as well as the consumer organisations (5 our of 5) and public authorities (5 out of 5). Opposing views do not exceed 15% of respondents. In parallel, almost half of the respondents express the view that retail roaming services are not sustainable with the current wholesale roaming caps, effectively supporting a reduction to wholesale caps. By contrast, only 1 out of 5 respondents support the opposite view. All responding MVNOs (9 out of 9) express that retail roaming services are not sustainable. MNOs are more divided and 11 out of 25 think that retail roaming services are not sustainable with the current wholesale roaming caps, 9 think that they are and 5 remain neutral.

Stakeholders seem to have divided views on the level of price caps. Large inbounder network operators (and especially multinational groups) tend to favour maintaining the wholesale caps at the current level. On the other hand, small outbounder network operators and virtual operators seem to support significant lowering of caps. We expect a similar trend to apply also to the views of Member States and public authorities. However, the public consultation did not render a sufficient number of data to present the views. BEREC in its opinion supported a reduction of price caps, as long as it ensures efficient cost recovery by the MNOs 74 .

Option 3B. Ensure a genuine RLAH

In addition to the transparency obligation in Option 2, Option 3 includes:

B1 – Prohibiting home operators from deliberately offering lower QoS, (e.g. limiting access to 3G instead of 4G in wholesale agreements), compared to the QoS offered in the home country. It foresees an obligation on retail and wholesale level. The QoS that a home operator needs to provide to its end customer while roaming depends on the technically available QoS level that is possible in the visited network, i.e. some networks may not be able to support the same QoS level as the home operator’s network.

Such an obligation would allow the NRAs of the home operator to monitor the compliance with the rules. It would also ensure that end-users are able to use innovative services over modern technologies such connected devices and 5G services.

73% of all respondents to the public consultation support a measure prohibiting the home operator to deliberately lower the QoS while roaming, compared to the QoS offered in the home country. 60% (86 out of 95) of the citizens and consumer organisations think that such an obligation is relevant. 6 out of 25 MNOs have expressed that such an obligation would be relevant, while 10 do not think it is relevant and 6 are neutral. Among the MVNO/Es 4 do think it is relevant while 3 do not and 1 is neutral. Among the other business stakeholders, including SMEs/entrepreneurs/vertical industries/IoT & M2M and industry associations 4 out of 5 think such a measure would be relevant.

BEREC in its opinion supported such a measure, suggesting that the Roaming Regulation must make clear that conditions of the domestic offer must not be altered by the home operator during roaming. 75 .

B2. Increase transparency regarding VAS. Two obligations will be included for all operators, in addition to the one in Option 2:

(a) Include in the “welcome SMS” a warning about calls to VAS numbers and the risk for bill-shocks, including a link to the web page of point (b).BEREC supports this, as a more targeted solution. In the public consultation, the majority of respondents (62%) supported the proposal to include information on VAS in the “Welcome SMS”.

(b) Provide a dedicated web page warning about the risk of bill shock when using VAS while roaming and detailed information about the types of services (calls and SMS) and the number ranges that may be subject to increased costs or blocking. This is intended to serve as a reference, for any end user who sees the warning in the “Welcome SMS” and wants more information.

The proposal to introduce an obligation on mobile operators to include in the “Welcome SMS” an alert informing that these types of communications may not be under the RLAH principle was supported by a majority of respondents (85, i.e. 62%) in the public consultation. BEREC in its opinion supported these additional transparency measures regarding VAS. 76  

As regards access to emergency services, in addition to the transparency measure in Option 2, option 3 includes measures to:

B3: Ensure access to emergency services in the visited country. It would clarify that, in order to enable the home operator provide free access to emergency services and free of charge provision of caller location information for all roaming end-users, as stipulated in Article 109 of the Electronic Communications Code, the visited operator should provide necessary technical and regulatory information to this effect. A way forward could be adding the necessary provision in the wholesale agreements.

In addition, home operators will have an obligation to inform customers at retail level on the possibility to access emergency services through alternative means.

B3: Ensure access to emergency services free of charge as at home. Not charging the emergency communications at wholesale level by the visited operator, would enable home operators to replicate the free of charge provision of emergency communication at retail level. As indicated in its additional input to the Commission, BEREC supports a clarification about the provider responsibility for bearing costs at the wholesale level would give more regulatory certainty especially for NRAs. 77 In the public consultation, 26% of the respondents expressed that this measure would be relevant, while it was considered not relevant by 28% of the respondents.

Option 3C. QoS, access to networks, facilitate innovation, and avoid misuse/fraud

On QoS and innovation, in addition to the measure proposed in Option 2, Option 3 includes the following measures:

C1: Ensure the same QoS while roaming as at home and respond to technological and business developments by clarifying the obligation on visited MNOs to give access to all network technologies and generations (2G, 3G, 4G, 5G etc.), upon a reasonable wholesale roaming access request. Clarifying the principle of access to all network technologies and generations on will allow all operators (MNOs and MVNOs alike) an equal and fair prospect to access the networks available to offer retail roaming services like at home. It would contribute to more equal terms for competition in the retail market, in particular for MVNOs. Whether or not an operator seeking access can offer reciprocal access in his network should not be a limitation. Furthermore, a reference offer should not be made in such a way that access to a certain technology is made more/less favorable.

The QoS for roaming services shall not be limited or hampered by access restrictions. The principle of access to all network technologies and generations ensures that MNOs do not limit wholesale roaming access seekers from offering retail roaming services to their end-users on any given network. This would allow wholesale roaming operators seeking access to effectively replicate the domestic retail offers, when technically feasible, in a roaming context. It aims to address also innovation and business developments, ensuring the widest use of 5G and modern connected services and to minimize the risk that end-users would not be able to use certain applications requiring 5G technology while crossing borders.

This option is supported by BEREC 78 and other stakeholders. 75% of all respondents are positive to an obligation MNOs to give access to all network technologies and generations (2G, 3G, 4G, 5G etc.) non-discriminatory access, upon a reasonable wholesale roaming access request. 89% (85 out of 95) of the citizens and consumer organisations think that such an obligation is relevant. 7 out of 25 MNOs have expressed that such an obligation would be relevant, while 8 do not think it is relevant and 6 are neutral. Among the MVNO/Es 7 do think it is relevant while 2 are neutral. Among the other business stakeholders, including SMEs/entrepreneurs/vertical industries/IoT & M2M and industry associations 4 out of 5 think such a measure would be relevant.

To address emerging needs of the evolving M2M market, the option will clarify that operators could employ charging schemes other than the volume based tariffs that are typically used (e.g. per SIM card and month) through alternative wholesale tariffs to be applied on voluntary basis. Given the expected market developments, operators are encouraged to accept all reasonable requests for establishing (or amending) roaming agreements for M2M communications services. Commercial agreements where operators explicitly allow permanent roaming for M2M are an expected trend for roaming wholesale agreements. Operators might increasingly agree not to consider caps to such charging schemes nor to any tariffs that apply in cases of permanent roaming and foresee offers that are better adapted to requirements of this market, that is still in its early stage of development.

The Commission will monitor the developments of roaming wholesale services for the M2M market and related commercial agreements (See also Section 9) and assess the situation in the context of the report by the end of 2025.

C2: European database on VAS number ranges to ensure higher level of transparency on wholesale level.

This measure envisages establishing a centralized European database for VAS numbering ranges but without information on wholesale tariffs (as this would significantly complicate the maintenance process). It is intended as a transparency tool that would enable NRAs and operators to have direct access to information about which numbering ranges can generate higher costs (termination rates) in all Member States. Accessibility will be limited, only to roaming providers and NRAs. The Regulation will not stipulate its usage but rather leave it to the discretion of NRAs and operators. NRAs could use it to confirm if high wholesale rates requested for a call to a specific number is indeed justified. Operators could use it to control calls to VAS, to warn consumers that the call might have increased cost. They could even block calls to certain number ranges, in case of suspicious fraudulent activity.

The task of establishing and maintaining the database can be assigned to BEREC. It is estimated that a maximum of 3 years from the adoption of the Regulation should be sufficient for developing such a database. BEREC has already initiated a similar (though much simpler) project for building a list (xls file) with all number ranges for voice-based Premium Rate Services and directory enquiry services in the EEA 79 .

In its opinion BoR(20) 131, BEREC supports this approach. Several operators have also suggested establishing such a database, in their responses to the joint Commission-BEREC online survey of 2020. More than 70% of respondents to the public consultation confirmed the need for introducing measures in the Roaming Regulation against bill-shocks from calls to VAS. The idea of setting up a European database for VAS was generally welcomed by stakeholders (more than 70% from all respondents groups), and in particular by operators and their associations, while 9% of respondents (a few companies and citizens) considered this solution not relevant. All consumer organizations (5) remained neutral.

C3. REFIT Horizontal simplification and improvement measures.

In addition to measures proposed in Option 2, under Option 3 the following would be included to further simplify the Regulation and reduce administrative burden.

It should introduce a possibility to use a lighter system for any future revision of wholesale caps. A full legislative process to revise the wholesale caps could be considered non-efficient. Regulatory efficiency could be achieved through an empowerment included in the new Regulation, setting the details of the methodology to be applied by the Commission for the revision of caps. Based on an updated cost model from 2026 onwards it would become possible to amend the caps through a delegated act. BEREC should have a role in this process, similarly to the delegated act setting the Union-wide voice termination rates under Art 75 of the EECC.

Such a light procedure would only be used, if the report on the functioning of the roaming market (foreseen for 2025) shows a need for regulatory intervention. Alternatively, the caps could remain unchanged after 2025, if competitive conditions show a well-functioning market. Only if there is a need for more structural changes to the Regulation itself, the Roaming Regulation would be reviewed.

Further simplification might be introduced in the monitoring process, possibly reducing the burden on operators and NRAs. In the framework of the coordination with BEREC, the Commission services will examine the possibility of merging and streamlining monitoring processes, including the obligation to publish the yearly report on transparency and comparability of roaming tariffs with the international roaming benchmark report and respective questionnaire. However, it is not planned to reduce considerably the data collected, since this would have a negative effect on the Commission’s ability to provide an adequate level of monitoring of the roaming market and analysis in future reviews and impact assessments, especially as regards regulation of wholesale pricing and roaming market dynamics. In its additional input to the Commission BEREC suggests simplifying the monitoring obligations by covering the data collection of the transparency and comparability report within the scope of the international benchmark report. 80  

5.2.3Option 4: Expanded wholesale obligations for an enhanced RLAH experience

Option 4A. Sustainable provision of RLAH

A1. Ensure enhanced sustainability of RLAH and cost recovery at wholesale level. This option is identical to option 3 but with the addition of measures aiming at alleviating the burden on MVNOs. Specifically, it mandates MNOs to pass the discounts they get on the rates for regulated wholesale roaming services to the MVNOs they host. This measure is relevant for those MVNOs that obtain wholesale roaming services through some sort of a resale agreement and not through bilateral negotiations (92% of full MVNOs and all light MVNOs, according to the June 2019 BEREC Opinion).

Around 30% of the respondents expressed a positive view on this proposal, another 30% indicated a negative view and around 40% did not reply or had a neutral view. Out of the respondents who expressed a view 16 out of 17 MNOs do not think such an obligation would be relevant, while 7 out 9 MVNO/Es think it is relevant. A limited number of other respondent group replied to this question, however the small number of responding consumer organisations and public authorities in general support this option.

Option 4B. Ensure a genuine RLAH

As regards QoS and innovation, in addition to the transparency obligation and obligation on prohibiting the home operator from deliberately offering lower QoS while roaming in Option 3, this option foresees also an obligation on the visited operator.

B1: Visited operators would be prohibited from deliberately offering lower quality of service for roaming customers in the visited network (e.g. limiting access to 3G instead of 4G), than what can technically be offered in the visited network.

This measure envisages introducing an explicit obligation on the visited mobile operator. The visited operator shall not deliberately limit access to its network for roaming customers and should not discriminate between their own customers and roaming customers by e.g. throttling the speed for roaming customers.

It is up to the home operator to ensure that access is requested for at least the same level of quality, if technically possible, in the visited network. The visited operator will be prohibited from deliberately lowering the QoS while roaming (e.g. limiting access to 3G instead of 4G or limiting the bandwidth) compared to what can technically be offered in the visited network and what has been agreed between the operators. This obligation would entail additional responsibility on the visited MNO to ensure that roaming customers in the visited network can enjoy the same QoS as at home (depending on what has been agreed in the wholesale agreement), without discrimination.

78% of all respondents are positive to an obligation on the visited mobile operator at retail and wholesale level, prohibiting deliberately lowering the quality of service while roaming (e.g. limiting access to 3G instead of 4G), compared to the quality of service offered in the home country. The majority of these respondents are citizens and consumer organisations (63%). 96% (91 out of 95) of the citizens and consumer organisations think that such an obligation is relevant.

B2. Increased control regarding VAS via dedicated (opt-in) mechanism. The opt-in mechanism means that customers who want to use VAS also while roaming would need to inform their roaming provider.

BEREC has considered this option as a possible solution, indicating however that its technical feasibility and its implementation costs for the roaming providers would have to be further assessed. In the public consultation, more than half of the respondents (57%) supported this proposal. However, operators were less positive, which reflects its technical complexity.

As regards access to Emergency services, and in addition to the measures proposed in Option 3 and in alternative to the transparency measure in Option 2, this option includes:

B3: Opt-in functionality for additional information on available alternative means of access to emergency services. It would enable end-users with disabilities to opt-in to receiving additional information about alternative means of access to emergency services. Information could for example be received through an additional SMS with information specifically about the available means of access to emergency services in the visited Member State.

This measure had some support by stakeholders participating in the public consultation. Amongst those respondents that expressed their opinion on the relevance of the proposal (52%), 16 found it relevant and 18 did not find it relevant. However, among those who considered the proposal relevant (relevant (11) or very relevant (5)) the European Disability Forum, European Union for Deaf and European Emergency Number Association consider it very relevant.

Option 4C. QoS, access to networks, facilitate innovation, and avoid misuse/fraud

C1: Obligation on the home operator at wholesale level to request the same QoS as offered at home, for all wholesale agreements, i.e. no preferred network. It would ensure that the roaming customers receive equivalent roaming services to the mobile services enjoyed at home, irrespective of the visited network(s) in the same visited country.

C2. Higher level of transparency on wholesale level for VAS. This an expanded version of option 3. It envisages establishing a centralised European database for VAS numbering ranges, including tariff information, allowing end-users to access information about VAS numbering ranges and charges. The task of establishing and maintaining the database will be assigned to BEREC.

As underlined in section 5.2.2, more than 70% of respondents expressed their agreement to the development of a database with VAS number ranges and relevant termination rates that is open to the public.

C3. REFIT Horizontal simplification and improvement measures, are the same as under Option 3.

5.3Options discarded at an early stage

The non-prolongation option has been discarded upfront based on the following reasons:

I.Findings of the 2019 Review report and BEREC opinions

The Review report has gathered a broad range of data and presents evidence to evaluate how the Roaming Regulation has performed and how the roaming markets are functioning and has confirmed the validity of the current approach. Deregulation is therefore not considered, since it would seriously risk the successful achievements the roaming intervention has brought to consumers and businesses. For additional details on the Review report and evaluation elements see Annex 6.

II.Unsustainability of RLAH without regulation

The Review report and section 2.2 of this IA both explain why RLAH cannot be sustained absent regulation given the observed market and regulatory failures. RLAH and its sustainable provision by operators have been ensured through the regulation of EU wide wholesale caps. If the Regulation is left to expire and the wholesale market would be left unregulated, there would be a market failure. Prices of “rest of the world roaming” (RoW roaming) demonstrate that without regulation prices would be considerably different.

III.Cost of non-Europe in this area:

There are 170 million European roaming customers currently enjoying RLAH that would risk to lose consumers benefits as presented in Section 3.2 that includes an estimate of the Consumer surplus as quantified by JRC (see also Annex 4).

IV.Stakeholders feedback and Eurobarometer confirming benefits

The public consultation on the review of the Roaming Regulation confirms that the Roaming Regulation for EU citizens and businesses is still needed, and none of the respondents explicitly proposed to lift the regulation. 96% of the citizens strongly agree (87%) or agree (9%) that they can enjoy the benefits that the Roaming Regulation aims to bring. The benefits include staying connected without having to restrict their usage of roaming, not worrying about having to pay excessive costs for the use of mobile services and continuing to use mobile services like at home . 74% of all respondents consider the regulation significantly relevant, while 10% are neutral and 8% expressed that the relevance is moderate. The 2018 Eurobarometer survey showed that 81% of travelers were aware that roaming charges had ended in the EU/EEA and 69% of all Europeans responded that they, or someone they know, benefit or will benefit. The feedback received by the Commission from consumer associations since June 2017 is overall very positive.

V.Political feasibility and misleading consumers

There seems to be a high level of political consensus to the continuation of the current policy. It would be therefore unrealistic to consider that the policy would come to an end. Consumers have also adapted consumption patterns to RLAH and would risk to incur bill shocks.

A complete list of options which have been discarded based on the exclusion criteria is included in Annex 9.

6What are the impacts of the policy options?

This section will assess the social, economic and environmental impacts of the options based on the assessment criteria indicated in Table 2 below.

Table 2: Criteria to assess the impacts of each option compared to the baseline

6.1Model for the assessment of economic impacts (sustainability analysis)

To assess the impact of the various policy options on the sustainable provision of RLAH (sustainability analysis), the Commission services have designed a model that evaluates the economic impact of RLAH on operators, taking into account the different policy options and the (forecasted) functioning of the roaming market. The analysis seeks to estimate the number of operators for which the provision of RLAH would be unsustainable 81 .

The model utilises data collected by BEREC in the framework of the International Roaming Benchmarking Reports, covering the period Q4 2016 to Q1 2020. The analysis covers a set of 96 operators, 72 MNOs and 24 MVNOs. It produces consumption forecasts for the period 2020-2025, based on the information on outbound and inbound volumes from 2017 Q2 up to 2019 Q4, as well on the EUROSTAT monthly data on “Nights spent at tourist accommodation establishments”. It then utilizes the above forecasts to assess sustainability at operator level over a 6-year horizon.

Sustainability, as defined in the Commission Implementing Regulation, is the ratio (in percentage) that compares the “Roaming Margin” (the difference between roaming revenues and roaming costs) with the “Domestic Margin” (operator profits from domestic services).

The analysis also examines the robustness of the forecasts by performing a sensitivity analysis. It specifically examines how sustainability develops under different scenarios leading to improved or deteriorated sustainability results as well as by taking into account the impact of the COVID pandemic (See Annex 4A for details).

6.2Impact of Option 2

6.2.1Sustainable provision of RLAH and other economic impacts on operators

Option 2 fails to make a difference compared to the baseline, alleviating sustainability challenges for operators, but ensures cost recovery for the provision of wholesale roaming.

According to the sustainability analysis (see section 6.1), in 2023 the provision of RLAH would be unsustainable for 26% of operators (compared to 27% in the baseline). Sustainability improvements could lead to a reduction of the total negative roaming margin 82 , compared to the baseline option by 23% in 2023 and 22% in 2025.

Concerning the risk of fraudulent generation of traffic towards international numbers (see section 2.1.4, problem C2, ‘misuse’), option 2 proposes a simple solution to increase transparency of VAS number ranges. This is intended as a tool that operators could use in a way that can reduce the risk for (wholesale) bill shocks from calls to VAS numbers and mitigate the risks linked to the artificial generation of traffic towards international destinations. However, it will be complex and potentially costly to use, because information on VAS will be dispersed in various sites. As a result, we expect that several operators (especially smaller ones) will be unwilling to use it.

6.2.2Genuine RLAH experience and social impact

The expected positive impact on end-users capacity to benefit from RLAH is limited since sustainability measures are not expected to have a major impact (as analysed in section 6.1.2). The analysis done by the Commission services estimates that in 2025, the percentage of EEA end-users who could be subject to sustainability derogations, hence not enjoy the full RLAH benefits is expected to remain practically unchanged (14.6% compared to 14.8% in the baseline).

Option 2 contains additional transparency measures on QoS, emergency services and VAS. As a result of these measures, consumers are likely to see an improved QoS and have a clear picture about the QoS they should expect when travelling abroad. Improving awareness on the expected QoS could help reduce consumer complaints. On the other hand, it cannot improve materially the RLAH experience by itself.

Awareness about the risk of bill-shocks from calls to VAS could reduce the risk of these bill-shocks. However, a transparency measure based on information provided only in contracts is likely to attract limited attention.

Increased awareness about alternative means of access to emergency services can help disabled end users that search for the information and pay attention to the “Welcome SMS”. However, it is unlikely to have equal impact on users that do not pay attention to the message or to users that have opted-out and do not receive it. Most important however, the identified limitations in emergency communications will remain (see also section 2.1.3) and continue hindering access to emergency services. As a result, we do not expect a material impact.

Overall, the positive social and consumer’s benefits impact remains modest with a high risk or derogations and the RLAH experience limited, with unsolved problems related mainly to QoS and emergency communications.

6.2.3Impact on SMEs, Digital Single Market and innovation

SMEs are an important group of business end-users for roaming services and benefits from RLAH are maintained (e.g. productivity gains, increased usage of digital services by consumers etc), along with general business end-users benefits.

SMEs and especially start ups developing new applications, benefit from the possibility for consumers of using applications continuously, also while travelling, without fearing to incur high costs. This is particularly relevant e.g. for applications that offer mobility solution/accommodation/other tourism related services and all the applications that might be particularly interesting while travelling.

However, practically unchanged sustainability challenges compared to the baseline, maintain the risk of sustainability derogations, leading to risk for some end-users of not benefit from full RLAH and this would consequently reduce SME benefits.

Option 2 might facilitate innovation and completion on the market, but only to a limited extent. The clarification that the wholesale access obligation is technology neutral, would eliminate the current doubts on the market in this regard. It might facilitate access to 5G networks but it is not coupled with obligations on the home operator to ensure the same level of services as domestically. As a result, its impact largely depends on the cooperativeness of operators.

It is likely to partially facilitate the utilization of applications requiring access to certain technologies and associated facilities in the mobile network. It would indirectly support innovation by service and mobile applications providers, especially of those that should function seamlessly across the Single Market.

The above positive albeit small impact for enterprises and business users in the digital single market cannot be quantified.

6.2.4Administrative burden and compliance cost

Option 2 introduces modest compliance costs, following the additional measures envisaged. In terms of administrative burden, it retains the reporting and monitoring mechanism as under the Roaming Regulation in force (see section 9). This not only saves the burden of new reporting processes but can benefit from the current experience to improve the efficiency, through the REFIT process.

Sustainability (A): It introduces a minor reporting requirement for operators (data on wholesale traffic exchanged in a non-discriminatory manner).

Quality of Service related measures (B1 and C1): The proposed measures (transparency and clarification on neutral access obligation for wholesale services) introduce a minor additional burden to operators and NRAs. Operators already have systems in place to manage their own customers’ roaming traffic dynamically, to ensure the best QoS, as confirmed in the public consultation. As a result, the compliance cost will be minor. Similarly, NRAs already monitor the QoS offered while roaming, so they will not incur an additional enforcement cost.

Access to emergency communications services while roaming (B3): The proposed transparency measures imply a minor compliance cost for operators. An additional compliance cost comes from the need to inform subscribers that have opted-out from receiving the Welcome SMS. However, failure to ensure the same level of access to emergency services while roaming as at home could have a considerable indirect cost that cannot be estimated. The inability to access emergency services could have a significant negative impact to the lives of travelers, including possible loss of life (see also section 6.3.1 on the impact of emergency calls to the lives of people).

Calls to VAS (B2 and C2): The proposed measures (transparency and publication of VAS number ranges at national level) have a minor impact on compliance costs and no impact on the administrative burden for operators. Also, they do not incur any additional enforcement costs for NRAs. The publication and update of information on VAS number ranges will have a modest cost for the NRAs (responsible for these tasks).

On the other hand, option 2 includes a minimum level of horizontal simplification (REFIT), including on the reporting process, so as to reduce compliance costs and the administrative burden of operators without compromising the collected data.

6.2.5Environmental impacts

The proposed measures do not have substantial environmental impacts compared to the baseline, thanks to slightly reduced derogations it might only marginally improve roaming customers ability to enjoy full RLAH and use new generation and IoT mobile services while roaming. For example, it could also enable consumers to continue using their “smart home” applications, allowing them to save energy.

Therefore, we expect option 2 to have a slight positive environmental impact but we are not able to quantify it.

6.3Impact of Option 3

6.3.1Sustainable provision of RLAH and other economic impacts on operators

Option 3 will lead to a substantial reduction to wholesale caps as well as to actual wholesale prices. This will ensure cost recovery for the provision of wholesale services as described in Section 5.2.2, preserving incentives to invest in visited networks and avoiding distortion of domestic competition in visited markets. It will also further affect operators in different ways:

Inbounder MNOs will see their positive roaming margin reduce, as a result of the reduced wholesale prices, due to evolving competition dynamics, also in light of potential emergence of non-discriminatory trading. However, this reduction is in practice balanced, thanks to the anticipated increase in roaming traffic. When the caps set are above the costs, an inbound operator will always have a positive margin. As roaming traffic increases and estimated costs decreases, the caps set based on cost recovery will likely also decrease. Therefore, the margin between costs incurred and the cap set might shrink, but due to the increasing traffic, the inbound operator will balance the reduction of the cap by the increase in traffic.

On the other hand, outbounder MNOs and MVNOs will see their negative roaming margin decrease (and in some cases become positive), as a result of the reduced wholesale prices. For an outbound operator, any reduction to the cap will always have a positive effect on the roaming margin if traffic is unchanged, as outbound operators by definition has to purchase more roaming traffic than it can sell from its own network. With an increase in traffic, the outbound operator will on the one hand likely see decreasing caps but will at the same time also need to purchase more roaming-traffic to serve the increased demand from its customers. So while increasing traffic leads to reduced unit costs, there is some uncertainty for the marginal outbound operator as to how this effects sustainability. As operators are not allowed to charge extra for roaming consumption, they do charge for overall consumption, including roaming consumption. Therefore, when the share of roaming traffic to domestic traffic increases, this might also affect how the retail revenue is allocated between roaming and domestic consumption. The outbound operator will therefore always be better off when wholesale caps decreases.

However, as a result of the reduction in the caps, operators might see their revenues from authorized roaming surcharges decrease (fair use policy retail revenues). This is due to the fact that rules on Fair Use Policies 83 , are based on the level of wholesale caps and therefore the limitations that operators are authorized to apply in case of particularly advantageous offers (data allowances based on the open data bundles limit) and in case of pre-paid subscriptions are automatically increased with the reduction of the caps. Also the maximum fair use policy surcharge (equal to the wholesale cap) is automatically decreased with the reduction of the cap.

According to the sustainability analysis (see section 6.1), in 2023 the provision of RLAH would be unsustainable for 19% of operators (compared to 27% in the baseline). This marks a 22% improvement, compared to the results produced by the model for 2019 and indicates that option 3 can lead to a further decrease in the number of sustainability derogations 84 . . Sustainability improvements could lead to a reduction of the total negative roaming margin, compared to the baseline option by 42% in 2023 and 52% in 2025.These are, however, conservative estimates and competitive dynamics might further lower down wholesale prices and hence the sustainability challenge in reality might be smaller than what is indicated by the model.

Furthermore, the proposed solution for the European database for VAS number ranges foreseen in option 3 is expected to reduce losses from fraud and misuse and reduced bill-shocks from calls to VAS numbers. However, it is only a tool and it is up to the operators to use it in a way that can minimize the risk for (wholesale) bill shocks from calls to VAS numbers. Estimating and monetizing the actual impact is difficult. A modest 20% reduction to misuse could save European operators a median of 14,000 € per operator yearly.

6.3.2Genuine RLAH experience and social impact

The expected positive impact on consumers’ and business end-users capacity to benefit from RLAH is considerably improved compared to the baseline. Option 3 strengthens sustainability and contains additional measures at both retail and wholesale level to ensure an adequate RLAH experience and to have a positive social impact.

Subscribers of inbounder MNOs and MVNOs are less likely to need sustainability derogation (hence end-users are more likely to enjoy the full benefits of RLAH), thanks to the reduced wholesale prices and the ensuring improved sustainability. For business end-users this means less risk of having to pay for the use of on-line productivity and business tools while roaming. The Commission analysis estimates that in 2025, the percentage of EEA end-users who could be subject to sustainability derogations, hence not enjoy the full RLAH benefits could be reduced from 14.8% to 8.6%.

Those consuming substantial volumes while roaming and being at risk of facing fair use policy surcharges, will see a reduced cost for two reasons. Reduced caps cause thresholds used in the open data bundle and pre-paid fair use limits to drop, which means less amount of consumed data becoming subject to a fair use policy surcharge. The reduction in wholesale caps would imply a 20% reduction of the maximum fair use surcharge for data from 1 July 2022 and an additional 25% reduction from 1 January 2025.

As a result of option 3 QoS measures, consumers are likely to see an improved QoS and have a clear picture about the QoS they should expect when travelling abroad. This can result in increased customer satisfaction and reduced complaints. According to the joint Commission/BEREC online survey, 18% of operators have received complaints that at most 3G was available and 22% have received complaints that full 4G speeds were not possible. The number of complaints could be reduced but we cannot quantify the relevant impact.

Consumers will also have better awareness about the cost of calls to VAS, thus, be less likely to face bill-shocks. This can in turn lead to reduced frustration from bill-shocks and fewer complaints that again cannot be quantified.

As regards social impacts, similarly to option 2, increased awareness about alternative means of access to emergency services can help disabled end-users in emergency situation reducing the baseline risks or not being able to use emergency services while roaming. In comparison to the baseline option and Option 2, this option is more likely to enable operators to offer equivalent access to emergency services like at home, including free of charge access and caller location (including handset based localization) that might be particularly important for e.g. end-users with exhausted pre-paid credit. This can have a major indirect benefit. We cannot give a specific estimation of live saved due to proposed measures, since it is not possible to estimate the number of cases where emergency communication in a roaming environment would be hindered. It is however, possible to give an indication of the average number of lives saved per year for every 100,000 mobile emergency calls. Just by implementing handset based Advanced Mobile Location (AML) average saved life per year is 0.45 and the average number of lives impacted every 100,000 mobile emergency calls is 4.37 (lives impacted are those persons that have a diminished or prevented injury as a consequence of accurate location). 85 Free of charge provision at wholesale level of the caller location transmission and wholesale level transparency, as proposed in option 3, would contribute to enabling of the presented benefits.

6.3.3Impacts on SMEs, Digital Single Market and innovation

Option 3 substantially reduces risks of SME not being able to benefit from RLAH, as it quite effectively alleviates sustainability challenges. This implies a reduced use of sustainability derogations (compared to options 1 or 2), hence a reduced number of end-users that might be subject to (the small) derogation surcharges. This means that under option 3, consumers and business end-users will continue using digital services while roaming and that SME employees will continue to have access to on-line, corporate productivity and collaboration tools without any surcharges while roaming.

The QoS measures are likely to facilitate future needs linked to technology and innovation developments, while ensuring the widest possible use of 5G and minimization of the risk for end-users not being able to use certain applications requiring 5G technology while crossing borders. As such, a non-discrimination obligation would ensure a level playing field for the European mobile operators. Ultimately, the home operators are empowered to ensure that their subscribers enjoy the same QoS while roaming as at home.

For application providers and start-ups, including SMEs, this implies an increased possibility for consumers to use applications continuously based on an improved and known QoS, also while travelling, without fearing to incur high costs due to surcharges. This is particularly relevant for applications that offer mobility solution/accommodation/other tourism related services and all the applications that might be particularly interesting while travelling.

This option might bring additional indirect innovation benefits to developers and end-users of application that require high QoS, especially if those applications and services might operate exclusively on 5G networks. Enjoying the same QoS when travelling as at home could indirectly support the uptake of applications and services, that can function and be used seamlessly across the Internal Market, without cross border interruption. However, the above positive impact for enterprises and business users in the Digital Single Market cannot be quantified.

The proposed clarification of the possibility to use alternative, non-volume based tariff structures can offer clarity and contribute to the development of the wholesale roaming market for M2M communications (including agreements that allow permanent roaming) and related innovation benefits. However, its effectiveness remains to be proved, because it largely depends on the cooperativeness of operators. The foreseen reporting and monitoring activities on the M2M market will allow NRAs to develop expertise on this emerging market. The additional administrative burden for market players and member states is minimal, as it does not introduce new reporting and monitoring obligations.

6.3.4Administrative burden and compliance cost

Option 3 introduces some additional compliance costs compared to the baseline and in addition to Option 2 following the additional measures envisaged:

Sustainability (A1): It introduces a minor reporting requirement for operators (data on wholesale traffic exchanged in a non-discriminatory manner).

Calls to VAS (B2 and C2): The proposed measures (transparency obligations and European database with VAS number ranges) have a minor impact on compliance costs and no impact on the administrative burden for operators. Also, it does not incur any additional enforcement costs for NRAs. The development and maintenance of the database will have an implementation cost for BEREC (responsible for both tasks) and a modest cost for the NRAs (responsible to update the database contents).

Access to emergency communications services while roaming (B3): The proposed measures imply implementation costs linked to amending wholesale roaming contracts, mostly linked however with the implementation of rights and obligations already defined in the European framework (EECC). Compliance costs relating to the transparency obligation are minor.

On the other hand, it includes horizontal simplification (already analysed under option 2), including on the reporting process, so as to reduce compliance costs and the administrative burden of operators (C3). It also introduces a lighter mechanism (a delegated act instead of a normal legislative procedure) for the potential future revision of wholesale caps. This is a substantial improvement, since by setting the parameters of pricing methodology in primary law, regulatory efficiency would be achieved by delegating the power to set the wholesale caps in accordance with this methodology to the Commission. If compared to a standard legislative procedure a delegated act would reduce excessive administrative burden on all subjects involved.

Annex 3 gives an overview of measures of option 3 and costs stakeholders incur.

6.3.5Environmental impacts

Option 3 will have a similar environmental impact as option 2, with slightly higher positive impact due to reduced derogations and improved consumers capacity to benefit from a genuine RLAH experience (including 5G) that makes it easier for them to use environmentally friendly mobile services and applications while roaming.

For example, roaming customers will be able to use increasingly new 5G and IoT mobile services connecting vehicles and road infrastructure (Cellular Vehicle-to-Everything or (C-V2X) 86 , while travelling through other member States. This could prevent human casualties and increase traffic efficiency, reducing the cost per km of travel in urban areas, fuel costs, CO2 and other air pollution emissions, lower need to repair or recycle wrecked vehicles, with a smaller ecological footprint 87 .

We expect option 3 to have a slight positive environmental impact compared to the baseline scenario.

6.4Impact of Option 4

6.4.1Sustainable provision of RLAH and other economic impacts on operators

Option 4 will lead to a substantial reduction to wholesale caps as well as to actual wholesale prices, while maintaining cost recovery for the provision of wholesale services. Its only difference with option 3 is the MNO obligation to pass any discounts they get on wholesale rates to MVNOs. This might improve MVNO sustainability (compared to option 3). Its impact on MNO cannot be measured with the available data.

This difference is reflected in the results of the sustainability analysis (see section 6.1), according to which, in 2023 the provision of RLAH would be unsustainable for 17% of operators (compared to 27% in the baseline). Sustainability improvements could lead to a reduction of the total negative roaming margin, compared to the baseline option by 45% in 2023 and 53% in 2025.However, this improvement might not happen, since it is based on an intrusive measure (MNO obligation to pass discounts on wholesale roaming tariffs to MVNOs). Indeed, hosting agreements can be subject to commercial negotiations or mandated by national decisions (e.g. dispute resolutions, remedies following approval of mergers, or even terms and conditions of procedures to award rights of use for spectrum). As a result, it is difficult to justify and implement. In addition, it will introduce substantial administrative overhead for both operators and member states, to monitor pricing and demonstrate the passing of discounts. Furthermore, it is not transparent and entails complex implementation and supervision resulting in substantially increased burden for market players and member states. Finally, it could have other negative repercussions that could effectively increase the roaming cost of MVNOs negating to some extend the anticipated gains. Examples would be including additional administrative or implementation charges (e.g. for reporting, negotiating wholesale rates, implementing FUP etc.) or requiring (possibly increased) commitments for purchasing wholesale roaming traffic (as is the case in direct negotiations of wholesale rates.

Concerning misuse, this option is practically similar to option 3 as both are based on establishing a central database with the VAS number ranges. The only difference (database open to the public under option 4 as opposed to operators and NRAs only under option 3) is unlikely to have an additional impact on misuse.

6.4.2Genuine RLAH experience and social impact

We expect option 4 to have a similar impact on genuine RLAH experience as option 3 in general:

Concerning QoS, option 4 provides an additional measure (prohibiting the visited operator from deliberately offering lower quality of service than what can technically be offered).

There is a slight additional positive impact, compared to option 3, that cannot be quantified.

Concerning VAS, option 4 requires opting-in to use VAS while roaming. The impact is considered similar to option 3. While we could foresee a stronger impact in terms of reduced frustration from bill shocks and complaints, the same measure would also hinder the use of VAS. This is equally likely to create frustration and complaints, reducing the overall impact as well as the benefit for consumers. According to the joint Commission-BEREC on-line survey of 2020, 16.4% of operators have received complaints for blocking calls to VAS. In either case we cannot quantify its impact.

Concerning emergency communications, option 4 includes an additional transparency measures, with an opt-in mechanism for additional information on available alternative means of access to emergency services. This constitutes a simple facilitation that cannot make a measurable difference, and might even reduce the impact since end-users might fail to opt in for this mechanism.

6.4.3Impact on SMEs, Digital Single Market and innovation

We expect option 4 to have a similar impact on SMEs and on the Digital Single Market as option 3. It substantially reduces the risks of SME not being able to benefit from RLAH compared to the baseline (with even higher impact than option 3) and has equally positive impacts on innovation and use of innovative mobile services, based on latest technologies.

There is a difference, since Option 4 contains an additional obligation on home operators to request same QoS as at home, for all wholesale agreements. This could increase the number of networks in each country that can deliver the desired QoS. It could facilitate end-users to enjoy the expected QoS. However, this is not necessarily a bottleneck, as long as there is at least one network with sufficient coverage and offering the expected QoS.

In any case, the impact of option 4 for enterprises and business users in the Digital Single Market cannot be quantified.

6.4.4Administrative burden and compliance cost

Option 4 introduces additional compliance and reporting costs compared Option 3. In terms of administrative burden, it retains the reporting and monitoring mechanism as under the Roaming Regulation in force (see section 9), however following the additional measures envisaged, mainly linked to the open access VAS database, and the additional MVNO measures, it increases considerably the overall reporting burden as explained below.

Innovation and Quality of Service related measures (B1 and C1): The option of not having a preferred network may be burdensome for the home operators as they would have to negotiate the same wholesale agreements with all operators that can technically fulfill the requested QoS level. It would limit bargaining power as well as the possibility of selecting the best wholesale offer. Overall, the proposed measures introduce a minor implementation cost for operators and a minor enforcement cost for NRAs but do not impact the administrative burden.

Access to emergency communications services while roaming (B3): The proposed measures imply increased implementation costs, mostly linked with the implementation of rights and obligations already defined in the European framework (EECC). Compliance costs relating to the transparency obligation are minor.

Calls to VAS (B2 and C2): The development and maintenance of an open access European database with VAS number ranges and tariff information. However, will have a substantial cost for BEREC (responsible for both tasks), especially due to the increased requirements (because of the open access). The inclusion of tariff information introduces a substantial additional cost for the NRAs (responsible to update the database contents) and a high administrative burden for operators that will have to report tariffs, as soon as they change.

Sustainability (A): The proposed measure (MNO obligation to pass discounts on wholesale roaming rates to MVNOs) will introduce complex and burdensome reporting requirements for both operators and NRAs. Its monitoring would require extensive reporting on the wholesale prices paid and the corresponding prices charged to MVNOs, where the linking between the two is not always straightforward. As a result, it can introduce a substantial increase to the administrative burden. Option 4 includes similar simplification gains (REFIT), as option 3.

6.4.5Environmental impacts

We expect a similar impact as described for options 2 and 3, which cannot be quantified.

Table 3: Summary assessment of impacts of options compared to the baseline

Level of Impact:

♦ Minimum impact

♦♦ Small Impact

♦♦♦ Strong Impact

♦♦♦♦ Very Strong Impact

♦♦♦♦ for the Administrative burden the scoring indicates a negative impact with high, non-proportionate costs, higher the scoring, bigger the negative impact

7How do the options compare?

7.1Effectiveness

Proposed scoring: ● Minor contribution towards objectives, ●● Major contribution but without fully achieving objective, ●●● Achieving objectives.

Objective A: Sustainable provision of RLAH 

Option 1 (the baseline) cannot contribute to ensuring the sustainable provision of RLAH. The sustainability challenges encountered under the baseline are described in detail in section 2.1.2. According to the Commission’s sustainability analysis (see Annex 4A), in 2023 27% of operators are expected to have negative roaming margin exceeding 3% of their roaming margin. This figure is expected to increase to 29% in 2025. (score ●)

Option 2 makes a minimal contribution towards ensuring the sustainable provision of RLAH as explained in section 6.1.2. (score ●).

Option 3 contributes towards the sustainable provision of RLAH without however entirely eliminating sustainability challenges, as explained in section 6.2.2. This confirms the need of maintaining the safeguards mechanisms already included in the Roaming Regulation (derogations and Fair use policy). Competitive dynamics might however further lower down wholesale prices and sustainability challenges might be smaller than what is indicated by the model. (score ●●●)

Option 4 contributes towards the sustainable provision of RLAH without however, eliminating entirely sustainability challenges, as explained in section 6.3.2. Its impact is slightly better than that of option 3. However, it is based on a measure that is difficult to implement and monitor, which compromises its effectiveness and confirms the need of maintaining the safeguards mechanisms (score ●●●).

Objective B: Ensure a genuine RLAH experience for end-users

Option 1 cannot contribute towards ensuring a genuine Roam-Like-At-Home experience for end-users. The relevant challenges are described in detail in section 2.1.3. (score ●)

Option 2 cannot effectively contribute towards objective B. The relevant measures can improve end-user awareness but do not materially improve the RLAH experience and do not address the identified limitations in emergency communications, discussed in section 2. Furthermore, they do not reduce considerably the risk of end-users facing a sustainability derogation surcharge (see section 6.1.1). (score ●●)

Option 3 contains additional measures that can contribute towards improving RLAH experience. Thanks to measures under 3A, the risk of subscribers facing sustainability derogation is expected to reduce from 14.8% to 8.6%, while the maximum level of FUP surcharges for data will be reduced by 20% from 1 July 2022 and an additional 25% from 1 January 2025 (as a result of the reduction of the wholesale caps). Consumers are likely to see an improved QoS and have a clear picture about the QoS they should expect when travelling abroad, which can decrease complaints. They will also have better awareness about the risk of calls to VAS, which can reduce bill-shocks and again decrease complaints. Finally, the proposed measures under option 3are more likely to enable operators to offer equivalent access to emergency services like at home, making a measurable impact to the quality of life, as explained in section 6.2.1. (score ●●●)

Option 4 is expected to have a similar impact on genuine RLAH experience as option 3 in general. The additional measures are expected to make only a minor difference, as discussed in section 6.3.1. and might go beyond what is necessary to achieve the objective satisfactorily (score ●●●)

Objective C: Ensure the same QoS as at home and access to all network technologies and generations, facilitate innovation and avoid misuse/fraud

Option 1 cannot contribute towards objective C. The relevant challenges under the baseline are described in detail in section 2.1.4. (score ●)

Option 2 cannot effectively contribute towards objective C. The impact of the measure on wholesale roaming agreements depends on the cooperation of operators, as it does not include any obligations on the home operators. Finally, the publication of VAS number ranges will be complex and potentially costly to use, because information on VAS will be dispersed in various sites/databases. As a result, several operators (especially smaller ones) might opt for not using it, which will limit its effectiveness. As also concluded in section 6.1.3, the positive impact of option 2 in this regard is small. (score ●)

Option 3 includes a set of measures that can contribute towards objective C. The proposed European database for VAS number ranges provides operators with a useful tool to combat losses from fraud and bill-shocks from calls to VAS numbers. Operators confirm this need in their responses to the public consultation. Also, according to the joint Commission-BEREC online survey, many operators have in fact sought to obtain information on VAS numbers and support this measure. The access obligation to all network technologies and generations empowers the home operators to seek the appropriate roaming agreements that can ensure the same level of QoS as at home, and the wide use of 5G technologies. The proposed clarification of the possibility to use alternative, non-volume based tariff structures can offer clarity and contribute to the development of the wholesale roaming market for M2M communications and related innovation benefits. However, its effectiveness remains to be proved, because it largely depends on the cooperativeness of operators. The foreseen reporting and monitoring activities on the M2M market will allow NRAs to develop expertise on this emerging market. (score ●●●)

Option 4 is equally effective as option 3, as regards objective C. The additional measure prohibiting the visited operators from deliberately offering lower QoS for roaming customers in the visited network than what can technically be offered, would introduce an additional level of safeguards to ensure that end-users are protected from QoS restrictions both from the home and visited operator, but a major limitation of this option is effective monitoring. The roaming end-user has a retail agreement with the home operator. The national regulatory authority of the roaming customer is not competent to monitor if the visited operator ensures any specific QoS level. Similarly, the regulator of the visited operator would not be able to monitor if the QoS offered is equivalent to what the roaming customer is offered by its home operator domestically. The main differences in the proposed measures are not expected to improve its effectiveness, as explained in sections 6.2.2 (additional features of database with VAS number ranges) and 6.2.3 (additional obligation on wholesale roaming agreements) and the option goes beyond what is necessary to achieve the objective satisfactorily (score ●●●)

7.2Efficiency

Proposed scoring: ● Considerable additional costs non-proportionate to the benefits and difficult implementation ●● Neutral or Increase in costs proportionate to the additional benefits ●●● Increase in costs largely outweighed by the benefits

Option 1 is a simple continuation of the current Regulation. As a result it is straightforward to implement and does not incur any additional implementation costs. However, it also fails to reduce the administrative burden. At the same time, it fails to ensure the same level of access to emergency services while roaming as at home. This could have an indirect cost that cannot be monetized. The inability to access emergency services could have a negative impact to the lives of travellers including possible loss of life (see also section 6.2.1 on the impact of emergency calls to the lives of people) (score ●).

Option 2 introduces a set of rather light measures that are easy to implement and incur a modest compliance cost. Like option 1, it implies a major indirect cost (failing to ensure access to emergency communications while roaming). However, it introduces simplification measures that reduce the administrative burden. Overall costs and benefits are balanced (score ●●).

Option 3 introduces sustainability measures that are based on the continuity of the approach already tested since RLAH introduction in 2017 and strengthen efficiency. They bring important benefits in terms of reduced negative roaming margin overall and reduced cost for derogations requests. It also increases consumer benefits in terms of RLAH (due to fewer derogations and reduced surcharges – see also section 6.3.2). These benefits outweigh the cost of the measure. As explained in section 6.3.1, increase in inbound, wholesale traffic mitigates losses of wholesale revenues (due to the reduction of wholesale caps). Also, the reduction in wholesale costs outweighs the reduction of revenues from surcharges (also due to the reduction in the wholesale caps).

If the additional measures are implemented those impose new implementation and compliance costs (e.g. measures to ensure access to emergency communications as at home, the VAS database, QoS requirements). Unlike the previous two options, it does not imply indirect costs while also introduces additional simplification measures that further reduce the administrative burden. Overall, its benefits, in terms of reduced sustainability challenges, improved QoS, reduced risks from calls to VAS for both consumers and operators, and unhindered access to emergency communications outweigh the cost (score ●●●).

Option 4 introduces a number of measures that are difficult to implement, costly and complex to monitor, thus creating substantial implementation costs for both operators and NRAs (measures to ensure access to emergency communications as at home, MNO obligation to pass any discounts on wholesale prices to MVNOs, database on VAS number ranges including tariffs and open to the public). The additional measure introducing wholesale level obligation requiring the home operator to request the same QoS as offered at home, for all available networks in the visited country is considered overly burdensome for operators, in particular for smaller operators, both in terms of economic burden and the cumbersome process of setting up wholesale agreements. It would limit bargaining power as well as the possibility of selecting the best wholesale offer. It may lead to a situation where end-users have the option to select a visited network but are not aware of which network has the technical possibility to offer equivalent QoS as at home. For the reasons above, the additional regulatory safeguards of option 4 are not considered efficient.

Option 4 complements the proposed transparency measures of option 3 with an additional opt-in mechanism for receiving information on available alternative means of access to emergency services. While this feature ensures that end-users that are not interested in alternative means of access are not overburdened with unsolicited information, it might even reduce the impact since also end-users that would need to use alternative means of access to emergency communication might fail to opt-in for this mechanism. Such an opt-in mechanism is more difficult to implement for operators, costly and complex to monitor, thus creating substantial implementation costs for operators without a corresponding increase of the effectiveness of the transparency measure towards end-users.

Overall Option 4, like option 3, it does not introduce indirect costs and reduces the administrative burden. However, the complexity and costs of the measures are not justified, and the option costs go beyond what is necessary considering the achieved impact (score ●).

7.3Coherence and proportionality 

Proposed scoring: ● Lack coherence and/or proportionality ●● Neutral ●●● Coherent and proportionate.

Option 1 is proportionate, as it is a simple continuation of the current Regulation in force. However, it fails to ensure access to emergency services as at home and to respond to technological and business developments. Therefore, it lacks sufficient coherence with existing policies, since it fails to complement the Accessibility Act and EECC measures on emergency services to ensure alternative access implementation in a cross border context. (score ●).

Option 2 is also proportionate, as it does not introduce any intrusive measure that might go beyond what is necessary to address the problems at stake. Like option 1, it fails to ensure a genuine RLAH experience due to derogations to RLAH reducing benefits for consumers and limitations in the access to emergency services as at home and fails to respond to technological and business developments, especially concerning 5G developments. Therefore, like option 1, it lacks sufficient coherence with existing policies, in particular with 5G strategies. (score ●)

Option 3 contains a set of measures that are neither intrusive nor disproportionate. In particular, the measures proposed are limited to those aspects, which have proved to be insufficiently addressed at national level and for which a harmonised approach is necessary, as explained in detail in the Subsidiarity grid, accompanying the legislative proposal (point 3.2). (score ●●●).

Option 4 includes measures that are intrusive and disproportionate. The obligation of MNOs to pass any discounts they get on wholesale roaming rates to MVNOs is intrusive and lacks EU added value. Hosting agreements between MNOs and MVNOs are governed at national level and NRAs have the necessary tools to address any relevant disputes. The measure prohibiting visited operators from deliberately offering lower quality of service for roaming customers, than what can technically be offered in the visited network goes beyond what is needed, as it could enable roamers to get better QoS than domestically. The same applies for the obligation on home operators to request same QoS as at home, for all wholesale agreements. (score ●)

Proportionality is reinforced for all options, because regulation is provided for a limited period: the Regulation will expire in June 2032. The 10-year duration takes into account that a decade is the typical duration to widely roll out a new generation of mobile communication and develop new business models. In a forward-looking manner, the competition is not expected to change significantly on the market in the next 10 years. The predefined duration is also based on the fact that the rollout of 5G “software-defined networks”, that allow slicing and could potentially have a profound impact on business models, requires the replacement of core network elements and is expected to take more time than the first 5G rollouts. Finally, the 10 year duration aims to ensure certainty in the market and minimize regulatory burden 88 .

Table 4: Comparison of the Options

Effectiveness scoring: ● Minor contribution towards objectives, ●● Major contribution but without fully achieving objective, ●●● Achieving objectives.

Efficiency scoring: ● Considerable additional costs non-proportionate to the benefits and difficult implementation ●● Neutral or Increase in costs proportionate to the additional benefits ●●● Increase in costs largely outweighed by the benefits.

Coherence and proportionality scoring: ● Lack coherence and/or proportionality ●● Neutral ●●● Coherent and proportionate.

For indications on the scoring system see above explanations under each criterion.

8Preferred option

8.1Sustainable and genuine RLAH

Option 3 is, overall, the most effective, efficient and proportionate/coherent way of achieving the objectives as summarized in Table 4 without exceeding what is necessary. It scores highest on all dimensions of effectiveness, efficiency, coherence and EU added value. 

Option 3 considerably improves sustainability results, reducing the number of operators with a negative roaming margin exceeding 3% of their domestic margin. It reduces the need for derogations, allowing an increased number of consumers to fully benefit from RLAH. It seeks to create the preconditions for operators to ensure the same QoS while roaming as domestically and it enables consumers to fully profit from technological development and 5G driven innovation. It seeks to address any lack of clarity on how operators can offer equivalent access to emergency services like at home and delivers useful tools for addressing (wholesale) VAS related misuse. Transparency measures have a positive impact on avoiding consumers’ bill shocks from calls to VAS and end-users awareness on equivalent means of access to emergency services. As a result, it can contribute substantially towards a genuine RLAH experience and have a positive social impact. The measures proposed in Option 4 are considered overly burdensome and disproportionate as explained in Section 7 and are therefore not included in the preferred option. Additional regulatory safeguards of option 4 would not outweigh the complexity in ensuring the implementation of the proposed measures and hence option 3 is considered sufficiently effective in addressing the problem related to ensuring a genuine RLAH experience.

While the administrative burden remains stable and is in fact reduced, thanks to the simplification initiatives, some of the proposed measures introduce additional compliance costs. For a detailed indication of how different operators, consumers, NRAs, BEREC, application providers and verticals are affected by the preferred option see Annex 3, that also includes an overview of cost an benefits of the preferred option. Stakeholders opinion on different measures is summarized in the synopsis report of the public consultation (See Annex 2).

Figure 8 below indicates how the measures included in the preferred option address the problems and objectives of this initiative.

Figure 8 The preferred option - Intervention logic

8.2REFIT (simplification and improved efficiency)

As indicated in the Review report, the Roaming Regulation does not face substantial implementation problems as confirmed by overall compliance of mobile operators with the rules, the absence of infringement proceedings, the effective enforcement actions at national level and the small number of complaints (see Review report). As a result, the need for improvements in the current Roaming Regulation is limited, also because many of the costs to implement the monitoring and transparency systems linked to RLAH have been already sustained and only the additional cost of maintaining the system in place are to be considered.

The following simplification and efficiency improvement are proposed:

8.2.1Revision of wholesale caps through delegated act

Instead of reviewing the whole Roaming Regulation when wholesale caps are revised, a lighter system could be introduced. A delegated act could be issued for the definition of wholesale caps e.g. based on an updated cost model (from 2026 onwards) and/or to possibly give a role to BEREC in the process, similarly to the delegated act for the Eurorate 89 .

REFIT impact: Reduction of legislative costs

The measure would allow to lighten the heavy legislative process for reductions of the wholesale caps (after 2025) via amendments of the Regulation. In a context where the economic space for reductions becomes extremely limited if we respect the cost recovery principle, a standard legislative procedure might not be the most efficient procedure. By setting the parameters of pricing methodology in primary law, regulatory efficiency would be achieved by delegating the power to set the wholesale caps in accordance with this methodology to the Commission and reducing the administrative burden on all subjects involved

8.2.2Repeal of the obligation for the separate sale of data roaming services (local data break-out), 

Due to the lack of actual uptake by the market, local data breakout obligations no longer appear to be relevant. The provision obliging operators to provide separate sales of roaming data services at retail level shall be repealed, including the Commission Implementing Regulation (EU) No 1203/2012 of 14 December 2012 on the separate sale of regulated retail roaming services within the Union.

REFIT impact: The systems of providing the possibility for separate sales of data services have already been put in place and operators already sustained the costs for deploying it. The repeal of this obligation may not significantly reduce their economic burden but it may somewhat reduce the maintenance costs, as well as the burden of offering separate sales of data roaming services.

8.2.3Repeal the Implementing Act on weighted average of maximum mobile termination rates

The Roaming Regulation provides for a surcharge on incoming calls based on the weighted average MTRs. The value is defined yearly by the EC in an implementing act based on input from BEREC. With the definition of the Eurorate (the single maximum mobile voice termination rate across Europe) this implementing regulation might be redundant. The Roaming Regulation could refer to the Delegated Act on the single weighted average MTRs directly.

REFIT impact: Such a change would reduce the burden on BEREC to provide input to the EC, as BEREC already has extensive reporting and data collection responsibilities pursuant to the Roaming Regulation (see also 8.2.5. on streamlining the monitoring obligations]).

8.2.4Align the provision on how to determine maximum charges in currencies other than Euro 

The Roaming Regulation sets out rules that oblige service providers in Member States outside the Eurozone to annually revise the maximum wholesale charges and retail surcharges for regulated roaming services and intra-EU communications.

The rule applied for intra-EU communications in accordance with Regulation (EU) 2015/2120 as amended by Regulation (EU) 2018/1971 sets out that the maximum charges in currencies other than the euro shall be revised annually and apply from 15 May using the average of the reference exchange rates published on 15 January, 15 February and 15 March of the same year. The roaming regulation will be aligned to those provisions defining the same date for revising the surcharge for roaming services (15 May instead of 1 May), and the same method for determination of those currencies.

REFIT Impact: The proposed measure would bring clarity and reduce the administrative burden of operators outside the Eurozone, who are obliged to publish their tariffs twice, when revisions of the roaming surcharges or intra-EU communications tariffs are introduced. Furthermore, it would reduce the monitoring burden for NRAs, which have to monitor the correct application of the revised exchange rates. Positive effects at retail level are expected, with retail prices updated once a year, instead twice as its currently the case, which can considerably reduce information obligation of operators for modified contract conditions.

8.2.5Rationalize reporting monitoring obligations

To reduce the regulatory burden for operators, NRAs and BEREC, BEREC proposes to remove the obligation to publish the yearly report on transparency and comparability of roaming tariffs. BEREC is of the view that the required parts of this report could be covered by the international roaming benchmark report.

Additional means of simplifying the monitoring procedure through merging and streamlining monitoring process will be examined in cooperation with BEREC.

REFIT impact: Reduction of administrative costs for BEREC, NRAs, operators.

Table 5: REFIT cost savings for the preferred option

9How will actual impacts be monitored and evaluated?

A well-established monitoring system is already in place. During the functioning of RLAH (from June 2017) the Commission and BEREC have used a wide number of indicators to assess the functioning of the roaming rules 91 . The monitoring system is based on data collected by the NRAs and BEREC, by means of semi-annual international roaming benchmarking questionnaires and an annual transparency and comparability questionnaire. Some indicators were also collected based on an annual joint Commission-BEREC online survey.

To monitor the effectiveness in reaching the additional general objectives illustrated in Section 4, the monitoring tools aim at assessing:

(a)The effectiveness of wholesale measures to ensure the sustainable provision of regulated retail roaming services and the smooth functioning of the roaming market.

(b)The extent to which European end-users enjoy a complete roaming experience while travelling, in particular regarding quality of service, access to emergency services and protection against bill shocks.

(c)Finally the interplay between technological innovations and RLAH.

The new measures introduced by the review would require number of additional indicators:

1.Quality of service indicators (e.g. number of roaming agreements by technological generation);

2.The extent to which roaming customers and operators face problems in relation to value added services

3.Roaming traffic negotiated in a non-discriminatory manner (trading platforms or similar instruments).

4.Indicators on the functioning of roaming in the context of the M2M market (number of M2M roaming agreements, M2M traffic, revenues and payments);

The monitoring system already in place allows the effective monitoring of the roaming market and is well established, therefore it will not impose any additional burden either on operators or on BEREC and the NRAs. Specifically:

·BEREC will continue to collect data regularly from national regulatory authorities on the developments in the retail and wholesale roaming market, utilizing the Benchmarking questionnaire. .

·BEREC will report regularly on the evolution of pricing and consumption patterns in the Member States both for domestic and roaming services, the evolution of actual wholesale roaming rates for unbalanced traffic between providers of roaming services, and on the relationship between retail prices, wholesale charges and wholesale costs for roaming services. BEREC shall assess how closely those elements relate.

·BEREC will also collect information annually from national regulatory authorities on transparency and comparability of different tariffs offered by operators to their customers.

·The Commission and BEREC may continue the joint on-line survey, using it with parsimony and only when necessary, to avoid increasing administrative burden.

The current monitoring system does not address emergency communications, which are monitored through an annual CoCom questionnaire.

Furthermore, the Commission will collaborate closely with BEREC with the view to simplify the process and reduce the administrative burden for operators and NRAs. Elements that could contribute towards this end are:

·The timing of the data collection process,

·The data collection cycles (e.g. semi-annual or annual),

·The stability of the questionnaires, to ensure that operators are familiar with the data they need to collect and have established standard procedures for collecting them,

·An efficient quality control system at NRA and BEREC level, so as to minimise the need for resubmissions,

·Restricting ad-hoc questionnaires to the minimum necessary,

·Refraining from imposing reporting obligations to very small operators.

(1) Single Market Barrier Report COM(2020) 93 final confirms that when the single market is failing to reach its full potential, SMEs and citizens are the most likely to suffer.

(2) European Electronic Communication Code – EECC- Directive (EU) 2018/1972),

(3)

Regulation 2017/1128 of 14 June 2017 on cross-border portability of online content services in the internal market.

(4)   https://ec.europa.eu/digital-single-market/en/news/digitalyou-digital-culture

(5)   https://ec.europa.eu/digital-single-market/en/policies/digital-cultural-heritage

(6)

Mainly two Regional Roaming Agreements (RRA) i) the Western Balkans RRA replicating for the 6 WB countries EU roaming policies (EU acquis alignment) gradually reducing roaming charges (78% average decrease for data already achieved) with RLAH planned from 1 July 2021. ii) EaP RRA introducing a customized harmonised regime based on EU’s regulatory approach gradually reducing during a 5-year transition period retail prices for consumers in the EaP region by 87%.

(7)

 Communication on ‘Shaping Europe’s Digital Future’, COM(2020) 67 final and lately with the “Connect” European flagship included in the Annual Sustainable Growth Strategy 2021, COM(2020) 575 final.

(8)

 Report on the review of the roaming market, COM(2019)616 final, and SWD(2019)416 available  here .

(9)

Based on the data collected in the International Roaming BEREC Benchmark Report, in Q3 2019 (i.e. July-September 2019), almost 170 million Europeans roamed abroad to another EU/EEA member state and enjoyed the benefits of RLAH. They generated a total of more than 6.4 billion minutes of voice traffic, more than 2.1 billion SMSs and more than 240 million GB of data traffic that were not subject to any kind of roaming surcharge.

(10)

In 2018 a large majority (69%) of consumers declared that they benefited or think that they or someone they know benefits or will benefit from roaming rules (Flash Eurobarometer 468, 2018).

(11)

A negative roaming margin of 3% is used in the current roaming rules as a threshold justifying a request for sustainability derogation. The derogation, granted by the national regulatory authorities on operator’s request, allows the application of a small surcharge.

(12)

An inbounder operator has a customer base which consumes less mobile services abroad than those consumed by the partner operators' customer base on its own network.

(13) An outbounder operator has a customer base which consumes more mobile services abroad (i.e. on the networks of partner operators in other EU/EEA countries), than those consumed by the partner operators’ customer base on its own network (i.e. when acting as a visited network).

(14) JRC quality of service study, SMART 2018/0011.

(15)

BEREC Opinion on the functioning of the roaming market as input to EC evaluation, BoR(19)101, 19 June 2019, available here .

(16)

See also BEREC Transparency and comparability report data in Annex 5: Baseline and in the.

(17)

Value added services refer to communications related to premium-rate numbers, freephone numbers and shared cost numbers, see glossary for more details.

(18)

Real time text, total conversation, SMS, emergency applications, web services, relay services. As defined in Article 2 EECC: (35) ‘total conversation service’ means a multimedia real time conversation service that provides bidirectional symmetric real time transfer of motion video, real time text and voice between users in two or more locations. Currently real time text is mandated in the Accessibility act for disabled end-user as of 2025.

(19)

Access to emergency services ensured by the European Electronic Communication Code in particular Article 109(2).

(20)

According to the latest replies to a COCOM questionnaire. These responses feed into the report to the European Parliament and the Council that has to be submitted by the Commission by 21 December 2020 pursuant Article 109(4) EECC.

(21) See the joint BEREC-Commission Survey 2020 (Annex 4).

(22) See the BEREC transparency and comparability Report 2019 BoR (19) 235, p. 25

(23)  WIK Consult estimates that the number of M2M subscriptions in the EU-28 will approach 1 billion in end of 2026 (see Annex 8), while Ericsson forecasts a large 15% growth of IoT connections every year until 2025 at least. Source: Ericsson mobility report 2020, p.23. - https://www.ericsson.com/en/mobility-report/reports/june-2020  

(24) M2M unlike interpersonal communications service (defined in Article 2(5) of the European Electronic Communication Code – EECC- Directive (EU) 2018/1972), these are exchanges of information between machines that involve limited direct interpersonal and interactive exchange of information between natural persons.

(25)  GSM Association, commonly referred to as 'the GSMA' or Global System for Mobile Communications, is an industry organization that represents the interests of mobile network operators worldwide.

(26) GSMA response to EC Public Consultation 2020.

(27) Ericsson mobility report 2020, p.23. - https://www.ericsson.com/en/mobility-report/reports/june-2020  

(28)  BEREC Survey 2020 (see Annex 2).

(29) BEREC survey 2020 (see Annex 2).

(30)

On 1 July 2019, the Western Balkans regional roaming agreement entered into force. It will have a gradual reduction of the roaming charges within the region and Roam like at home will be introduced from 1 July 2021. Already from July 2019 the tariffs were significantly reduced, on average, prices for outgoing calls were lowered by 65% and for data transfer there were a 78% decrease.

(31) Typically, each MS has 3-4 MNOs and among those, often not all of them can provide a full range of services (such as full geographic or services coverage) needed to provide services to all roamers. All failures were already presented in SWD (2016) 202 final (the Impact Assessment for Regulation 2017/92 amending Regulation (EU) 531/2012 as regards rules for wholesale roaming markets) and all persist.

(32) SMART 2018/12 Technological developments and roaming" by WIK Consult, July 2019, available here

(33) According to BEREC’s International Roaming Benchmark Report, in Q1 2019 the average wholesale price for voice and data was respectively 55% and 35% of the wholesale cap. When looking at the price charged per operator, 3 out of 4 of operators charge on average up to 72% of the wholesale voice cap and up to 60% of the wholesale data cap.

(34) The established caps must balance ensuring cost recovery for the visited network (supplying the wholesale service) on the one hand, whilst at best possible reducing the cost to the home operator (buying the wholesale service), as RLAH prevents the home operator of charging the retail customer for the used roaming services.

(35)

See details in the staff working document accompanying the Impact Assessment for Regulation 2017/92.

(36) Directive (EU) 2018/1972 establishing the European Electronic Communications Code.

(37) With RLAH, operators are in principle not allowed to charge for retail roaming services. End-users would not (and should not) take the prices for retail-roaming into account when choosing a domestic subscription and hence one cannot rely on for example the elasticity of roaming retail prices (as this has a marginal price by definition of zero, unless a derogation has been approved or the end-users consumes beyond the FUP.

(38)

As a reference, European incumbents charge anything between 500 and 15.500 euro per gigabyte of data used outside of EU.

(39)

See Study SMART 2018/0012 "Technological developments and roaming" by WIK Consult, July 2019, available here performed for the roaming review in 2018-2019.

(40)

Directive (EU) 2019/882, Annex 1, section IV(a) : (a) Electronic communications services, including emergency communications referred to in Article 109(2) of Directive (EU) 2018/1972:

(i) providing real time text in addition to voice communication;

(ii) providing total conversation where video is provided in addition to voice communication;

(iii) ensuring that emergency communications using voice, text (including real time text) is synchronised and where video is provided is also synchronised as total conversation and is transmitted by the electronic communications service providers to the most appropriate PSAP.

(41)

Art 2(1)(35) EECC: ‘total conversation service’ means a multimedia real time conversation service that provides bidirectional symmetric real time transfer of motion video, real time text and voice between users in two or more locations.

(42)

In this regard, please note that the cost model used by the Commission services indicates a continued downwards trend in costs for providing roaming wholesale services. Therefore, it is reasonable to consider whether these caps should be revisited (for more, see Annex 4).

(43)

Case C‑491/01 British American Tobacco (Investments)and Imperial Tobacco [2002] ECR I-11453, paragraph 60, and Case C-217/04 United Kingdom v Parliament and Council [2006] ECR I‑3771, paragraph 42).

(44)

Case C‑380/03 Germany v Parliament and Council [2006] ECR I‑11573, paragraph 37 and the case-law cited.

(45)

Case C‑376/98 Germany v Parliament and Council [2000] ECR I‑8419, paragraphs 84 and 106.

(46)

See Opinion Of Advocate General Poiares Maduro delivered on 1 October 2009 in Case C‑58/08  http://curia.europa.eu/juris/document/document.jsf?text=&docid=72636&pageIndex=0&doclang=EN&mode=req&dir=&occ=first&part=1&cid=2824416

(47)

See December 2005 ERG letter to the Directorate general of the Commission's DG Information Society.

(48)

C-58/08 Vodafone, CJEU judgment of 8 June 2010.

(49)

The analysis has been performed comparing average domestic revenues before and after the implementation of RLAH, controlling for time invariant characteristics at the operator and country level.

(50) Further underlined by the very few number of derogations of the RLAH regime received.

(51)

See, among others, the ETNO policy note The role of Digital Communications at the time of COVID-19: Building A Digitally-Enabled Recovery, available here .

(52)

 Commission Recommendation C(2020) 6270 of 18 Sep. 2020 on a common Union toolbox for reducing the cost of deploying very high capacity networks and ensuring timely and investment-friendly access to 5G radio spectrum, to foster connectivity in support of economic recovery from the COVID-19 crisis in the Union.

(53)

Initiative for the Review of the Broadband Cost Reduction Directive (Directive 2014/61/EU) in CWP 2020.

(54) 5G Action Plan for Europe - COM(2016)588.

(55) Staff Working Document: "Advancing the Internet of Things in Europe", accompanying the communication Digitising European Industry - Reaping the full benefits of a Digital Single Market COM(2016) 180.

(56) COM(2016) 766 of 30 Nov. 2016 on “A European strategy on Cooperative Intelligent Transport Systems, a milestone towards cooperative, connected and automated mobility”.

(57) Commission Recommendation (EU)2020/518 of 8 April 2020 on a common Union toolbox for the use of technology and data to combat and exit from the COVID-19 crisis, in particular concerning mobile applications and the use of anonymised mobility data (in particular Art. 14 and recitals 14 and 19).

(58) Communication COM(2020) 724 of 11 November 2020 on Building a European Health Union: Reinforcing the EU’s resilience for cross-border health threats.

(59)

Fair use policy is aimed to prevent abusive or anomalous use of roaming services at domestic prices (such as permanent roaming). Sustainability derogations refers to the exceptional and temporary derogations to forestall any risk of domestic price increases.

(60)

See Annex 5 (Baseline) for a list of derogations granted per country. The number of derogations granted from 1 September 2019 is substantially reduced but this period is not considered representative, due to the overlap with the COVID pandemic.

(61)      An inbounder operator has a customer base which consumes less mobile services abroad, than those consumed by the partner operators' customer base on its own network.

(62)      An outbounder operator has a customer base which consumes more mobile services abroad (i.e. on the networks of partner operators in other EU/EEA countries), than those consumed by the partner operators’ customer base on its own network (i.e. when acting as a visited network).

(63) The increasing use of teleworking, teleconferencing and in general digital services has led to substantially increased demand for connectivity and a up to 40% increase in mobile data traffic has been reported by telcos during the pandemic as indicated in ETNO policy note The role of Digital Communications at the time of COVID-19: Building A Digitally-Enabled Recovery, available here

(64)

Joint Commission-BEREC online survey (2019) and BEREC questionnaire for the Transparency and Comparability Report (2020)

(65)

According to the information provided to the Commission by national authorities in the 2019 COCOM questionnaire , 14 Member States have deployed emergency applications, while 24 Members States have deployed SMS as alternative means of access for end-users with disabilities. Some Member States deploy both means of access. https://ec.europa.eu/digital-single-market/en/news/2019-report-implementation-european-emergency-number-112

(66)

 BEREC input on EC’s request for the preparation of the legislative proposal for the new roaming regulations, BoR(20)131, 30 June 2020, available here

(67)  BEREC Opinion on the functioning of the roaming market, BoR(19)101, 19 June 2019.

(68) Indicated in BEREC input BoR (20)131 p. 22.

(69)

Repeal of the Implementing Regulation on weighted average of maximum mobile termination rates (instead the Eurorate will apply), and removal of the obligation for the separate sale of data roaming services (local data break-out) which has become redundant with RLAH. This measure is supported by BEREC, see BEREC input on EC’s request for the preparation of the legislative proposal for the new roaming regulations, BEREC BoR (20) 131, 30 June 2020.

(70)

The Roaming Regulation sets out rules that oblige service providers in Member States outside the Eurozone to annually revise the maximum surcharges for regulated roaming services and intra-EU communications. Currently different dates are used to determine the reference rate for roaming charges and intra-EU charges.

(71) See Implementing Regulation (EU) 2016/2286 and Annex 7.

(72)

Here, maximum Union-wide mobile termination rates are set to follow a glide path starting in 2021 at 0.7 EURcents/min, decreasing to 0.55 (2022), 0.4 (2023) and 0.2 (2024). (link to final delegated act to be added)

(73)  BEREC Opinion on the functioning of the roaming market, BoR(19)101, 19 June 2019. 

(74)  BEREC Opinion on the functioning of the roaming market, BoR(19)101, 19 June 2019. See also BEREC additional input to the Commission BoR (20) 131, 30 June 2020.

(75)  BEREC Opinion on the functioning of the roaming market, BoR(19)101, 19 June 2019. See also BEREC input on EC’s request for the preparation of the legislative proposal for the new roaming regulations, BEREC BoR (20) 131, 30 June 2020.

(76)  BEREC input on EC’s request for the preparation of the legislative proposal for the new roaming regulations, BEREC BoR (20) 131, 30 June 2020.

(77) BoR (20) 131, p. 37.

(78)

BoR(20) 131, p. 29.

(79)  BEREC input on EC’s request for the preparation of the legislative proposal for the new roaming regulations, BEREC BoR (20) 131, 30 June 2020.

(80)

The Commission Implementing Regulation 2016/2286 (Article 10, par. 1), stipulates that the ability of a roaming provider to recover its costs of providing regulated retail roaming services, would be undermined, only where the negative roaming retail net margin is equivalent to 3% or more of its mobile services margin.

(81) This is the total roaming margin of operators with negative sustainability.

(82)

Commission Implementing Regulation (EU) 2016/2286, see also Annex 7.

(83)

From 31 August 2018 to 31 August 2019, NRAs granted sustainability derogations to 24 operators.

(84)

 HELP 112 II project cost benefit analysis. https://ec.europa.eu/digital-single-market/en/news/112-112-day-locating-emergency-calls-aml-technology-rise . The forecasted scenario is enabled by the legal obligation provided in Art 109.6 EECC of having the handset-based localisation deployed by 21 December 2020. This impact is relevant to the proposed regulatory measures insofar the delivery of handset based caller location is linked to ensuring that the transmission of caller location is free of charge for the end-user. Free of charge provision at wholesale level of the caller location transmission and wholesale level transparency, as proposed in option 3, would significantly contribute to enabling of the presented benefits.

(85)

  https://www.gsma.com/iot/wp-content/uploads/2019/08/Connecting-Vehicles-Today-and-in-the-5G-Era-with-C-V2X.pdf  

(86)

IA accompanying the document “Commission Delegated Regulation supplementing Directive 2010/40/EU of the European Parliament and of the Council with regard to the deployment and operational use of cooperative intelligent transport systems: p. 17: summary of costs and benefits SWD (2019) 96 final of 13.3.2019 https://ec.europa.eu/transparency/regdoc/rep/10102/2019/EN/SWD-2019-96-F1-EN-MAIN-PART-2.PDF  

https://ec.europa.eu/transparency/regdoc/rep/10102/2019/EN/SWD-2019-96-F1-EN-MAIN-PART-1.PDF  

(87)

Delegated Act (a Delegated Regulation) under Article 75 of the European Electronic Communications Code, to be adopted by 31 December 2020.

(88)

For the calculation, we have used labour costs from the Eurostat database and the number of operators per member states responding to the BEREC benchmarking questionnaire, as those are the operators that actually incur reporting costs.

(89)

 Examples of such indicators are: Retail (outbound) roaming traffic (voice, SMS, data) as well as retail roaming traffic per roaming enabled subscriber and per roamer, Retail traffic subject to fair use policy / sustainability derogation / alternative tariff (as % of total roaming traffic in the EEA), Average actual wholesale prices charged per unit (for total, balanced and unbalanced inbound traffic) to the EEA, Average actual wholesale prices charged for roaming traffic to the rest of the world, Average retail roaming prices to the rest of the world, Number of complaints (by type of complaint)Derogations requested and granted, Number of operators applying each fair use policy, Domestic only subscribers (as % of total subscribers), Subscribers subject to derogation (as % of total roaming enabled subscribers).

EUROPEAN COMMISSION

Brussels, 24.2.2021

SWD(2021) 28 final

COMMISSION STAFF WORKING DOCUMENT

IMPACT ASSESSMENT REPORT

ANNEXES

Accompanying the

Proposal for a Regulation of the European Parliament and of the Council

on roaming on public mobile communications networks within the Union (recast)

{COM(2021) 85 final} - {SEC(2021) 90 final} - {SWD(2021) 27 final} - {SWD(2021) 29 final}

Table of contents

Annex 1: Procedural information

1.    Lead DG, Decide Planning/CWP references    

2.    Organisation and timing    

3.    Consultation of the RSB    

4.    Evidence, sources and quality    

Annex 2: Stakeholder consultation

Annex 2A: The Summary of the Stakeholder Strategy    

Annex 2B: The Inception Impact Assessment Feedback    

Annex 2C: The outcome of the Public Consultation    

Annex 2D Joint Commission-BEREC online survey 2020    

Annex 3: Who is affected and how?

Practical implications of the initiative    

Summary of costs and benefits    

Annex 4: Analytical methods

Annex 4A - Sustainability assessment: methodological overview    

Annex 4B - Counterfactual evaluation of the impact of RLAH and assessment of benefits for consumers    

Annex 4C Cost model for wholesale roaming costs    

Annex 5: Baseline

Baseline A. Sustainable provision of RLAH    

Baseline B. Ensure a genuine RLAH from an end user perspective    

Baseline C. Quality of Service, access to networks, facilitate innovation, and avoid misuse from the operator perspective    

Annex 6: Evaluation - Findings from the Review report on the functioning of the Roaming Markets

Annex 7: How roaming works

Annex 8: Technological developments and roaming

Annex 9: Discarded Options

Annex 1: Procedural information

1.Lead DG, Decide Planning/CWP references

The Directorate-General (DG) for Communications Networks, Content and Technology was leading the preparation of this initiative and the work on the Impact Assessment in the European Commission. The planning entry was approved in Decide Planning under the reference PLAN/2020/6784 - Initiative for reviewing and prolonging the “Roaming Regulation”. This initiative is included in the Commission’s Work Programme 2020 under the Policy Objective “A Europe fit for the Digital Age” addressing the specific objective “Digital for consumers”.

2.Organisation and timing

The planned adoption date included in the Commission Work Programme adopted on 29 January 2020, was changed from Q4 2020 to Q1 2021in the revised version adopted on 27 May 2020 following the COVID-19 crisis.

An inter-service steering group (ISSG), was previously established for the review of the Roaming Regulation. This ISSG was consulted for preparing this initiative. The ISSG met three times, in the scope of this initiative, in the period from March 2020 until adoption in November 2020. The ISG was also consulted through written procedure two times in the same period.

3.Consultation of the RSB

An upstream meeting was held with the Regulatory Scrutiny Board (RSB) on 5 October 2020. A draft Impact Assessment was submitted to the RSB on 19 October 2020. Following the Board meeting on 18 November 2020, it issued a positive opinion on 20 November 2020. The Board’s recommendations were addressed as presented below. The Board also noted the useful additional information provided in advance of the meeting and commitments to make changes to the report. These additional changes are also presented in the table below.

RSB Opinion of 20 November 2020

The Board’s Recommendations	IA modification
Improvements addressing suggestions from the RSB in the Opinion
(1) The report could better explain how increasing demand for services leads to economic risks for inbound operators if unit costs decrease with increasing usage of the network.	- The impacts of the chosen caps levels were further explained in Section 6.3.1 (Sustainable provision of RLAH and other economic impacts on operators).
(2) The report should better explain the reasoning behind the choice of the level of the price caps, particular the specific cap proposed for data. In this regard, it should discuss which alternative options were considered and explain why these price cap variants were eventually discarded.	- The choice of level of the wholesale caps was further explained in Sections 5.2.2 (Option 3A. Sustainable provision of RLAH) and in Section 2.1.2 further discussing the sustainability problem.
(3) The report should better explain the various combinations of measures it considered in defining the preferred option. It should set out more clearly, why certain potentially beneficial measures were not included.	- An explanation of the choice of measures for the preferred option was added in Section 5.2 (Description of the policy options). - Measures under option 4 are considered overly burdensome and disproportionate and are therefore not included in the preferred option. This has been clarified in Section 7.2 and in Section 8)
(4) The report should further develop and analyse the impact of the COVID crisis on the baseline. It should assess its impact on the sustainability of RLAH, both for outbound and inbound operators. It should expand the sensitivity analysis by explaining how the discussion of sensitivity results reflects the possible impacts of the COVID crisis.	Impacts of the COVID-19 crisis were further elaborated in sections (5.1.2 Option 1 - Baseline: Prolong the Roaming Regulation without amendments), and the sensitivity analysis (Annex 4A) was elaborated to further present the assessment on the impact of the COVID-19 crisis, with four additional COVID-19 scenarios.
(5) The report should present upfront the broader political and regulatory context of the initiative. It should explain the importance of the initiative and how it contributes to the development of other EU policies.	The broader political context was further explained in Sections 1 (Introduction) and 4.1 (General objectives).
(6) The report should better highlight what key problem is tackled by this initiative in terms of its magnitude and urgency.	The magnitude of problems was further clarified, particularly as regards sustainability and QoS. Additional evidence from the Annex was incorporated in Section 2.1.2 (A. Problems to ensure sustainable provision of RLAH), to further clarify the expected magnitude of the problems related to sustainability and referring to the extensive discussion on safeguards and the limited amount of the overall roaming traffic currently subject to fair use and sustainability derogations. Evidence supporting the regulatory intervention on quality of service (QoS) has been more prominently presented in Section 2.1.3 (B.1. Low perceived QoS and information failure on QoS and RLAH).
Additional improvements addressing elements from the Impact Assessment Quality Checklist
Content and scope	Section 1 and 4 were amended to explain the coherence and complementarity of this Initiative and its objectives with other EU policies. - The possible impact of COVID-19 based on the sensitivity analysis has been further clarified in Section 1 and Annex 4, presenting ad hoc sensitivity scenarios.
Problem definition	- The key problem of sustainability for wholesale operators has been better defined in terms of magnitude and urgency to act. Additional evidence from the Annex has been incorporated in Section 2.1.2 (A. Problems to ensure sustainable provision of RLAH), to further clarify how substantial the problems related to sustainability are, also referring to the extensive discussion on safeguard. - Section 2 also further reports on the findings as to why there is at this stage no need to intervene on fair use and sustainability also based on the data presented in the Annex and Review report. In particular, the impact assessment highlights the limited amount of the overall roaming traffic currently subject to fair use and sustainability derogations. - The factors underlining the need for and the evidence supporting the regulatory intervention on QoS has been more prominently presented in Section 2.1.3 (B.1. Low perceived QoS and information failure on QoS and RLAH). - Additional clarification of the supportive positions of BEREC and NRAs as well as other stakeholders in this area has been included in Section 5.2. (See specifications in Baseline and options below).
Objectives and intervention logic	- As regards the specific objectives B1 and C1, the difference and link between those two objectives related to QoS has been further clarified in Sections 4.1 (General objectives), 4.2.2 (B. Ensure a genuine RLAH experience for end-users) and 4.2.3 (C. Ensure the same QoS as at home and access to all network technologies and generations, facilitate innovation and avoid misuse/fraud from the operator perspective) - The consistency of the three general objectives with the broad policy strategies and other relevant policy initiatives has also been further developed in Section 4.
Baseline and options	- The hypothesis about the sensitivity analysis and the impact over time of the COVID-19 crisis have been clarified in the baseline Section 5 making a stronger link with the sensitivity analysis presented in Annex 4. - The rationale behind the proposal of the preferred lower caps for data has been clarified in Section 5.2.2 - The presentation of the stakeholder views by stakeholder category (operators) for some options as well as NRA views through the BEREC opinion has been expanded in the report to provide an analytical and fully transparent picture on the views expressed. See Sections 5.2.1 (Option 2B Increase transparency regarding QoS and Option 2C Introduce a minimum level of simplification), 5.2.2 (Option 3A Ensure enhanced sustainability of RLAH and cost recovery at wholesale level, 3B Prohibiting home operators from deliberately offering lower QoS, Option 3B Increase transparency regarding VAS, Option 3B Ensure access to emergency services free of charge as at home, Option 3C Ensure the same QoS while roaming as at home and respond to technological and business developments by clarifying the obligation on visited MNOs to give access to all network technologies and generations, upon a reasonable wholesale roaming access request, Option 3C REFIT Horizontal simplification and improvement measures), Section 5.2.3 (Option 4A Ensure enhanced sustainability of RLAH and cost recovery at wholesale level, Option 4B Opt-in functionality for additional information on available alternative means of access to emergency services) and in Annex 2C (The outcome of the public consultation). - As regards REFIT and monitoring, the report clarifies in the report that the main means of simplifying the monitoring procedure is through merging and streamlining monitoring processes. See Section 5.2.2. Option 3C and 8.2.5 (8.2.5 Rationalize reporting monitoring obligations).
Impacts	- Section 6 has been modified to ensure full coherence between the various tables presenting the approach for assessing impacts and the summary table presenting impact results. - Annex 4 further explains how the sensitivity analysis considers the impact of COVID-19 pandemic in the framework of the sustainability analysis. - The coherence between Cost benefit tables in Annex 3 and REFIT tables in Section 7 has been ensured.
Comparison of options and proportionality	- The tables in Section 7 comparing the policy options have been streamlined and refer consistently to the baseline. - The scoring of different options in Section 7 has been amended to a categorical scoring instead of numerical. - The report has further clarified that various combinations of measures have been considered in defining the preferred option. However, option 3 includes measures that were considered preferable for each thematic area, also considering possible interdependencies. - Measures under option 4 are considered overly burdensome and disproportionate and are therefore not included in the preferred option. This has been clarified in Section 7.2)
Future monitoring and evaluation	- The report already presents additional indicators to be monitored on QoS, inbound traffic negotiated in a non-discriminatory manner and on the functioning of roaming in the context of the M2M market. The need of additional progress monitoring indicators has been verified and presented in Section 5.2.2 Sections 8.2.5 (Rationalize reporting monitoring obligations),
Consultation, information base and methodology	- More clarity has been provided on how the sensitivity analysis of the sustainability assessment has taken COVID-19 into account. - Annex 2 on the results of the public consultation has been further verified to ensure that views of specific/disaggregated stakeholders groups is systematically reported when they differ significantly and differences are relevant for the assessment of options.
Presentation	- The glossary has been further elaborated to better assist accessibility of the report to non-specialists - The subsidiarity grid has been taken out from the Annexes of the Impact Assessment. It will be attached to the explanatory memorandum that accompanies the draft legal proposal. Annexes have been re-numbered accordingly.

4.Evidence, sources and quality

The evidence used in the IA have been collected along the following streams:

I.BEREC data collection:

a.BEREC semi-annual international roaming benchmark reports 1 (adopted in March and October each year). Data submitted by the operators on roaming developments at retail and wholesale levels. These are based on a dedicated questionnaire, which is run by BEREC and the NRAs twice a year. It particularly includes data on volumes of roaming traffic (retail and wholesale), wholesale roaming revenues, wholesale roaming rates and information on wholesale roaming contracts.

b.BEREC annual report on the transparency and comparability of roaming tariffs 2 (adopted in December each year since 2017). BEREC is mandated by the co-legislator in the Roaming Regulation to collect data from NRAs on transparency and comparability of roaming tariffs once a year. For this purpose, NRAs are running another data collection from mobile operators in August every year. Based on these data, BEREC publishes country-level results in the BEREC Report on Transparency and Comparability of International Roaming Tariffs in December every year.

II.Joint Commission/BEREC online survey - Market input has been collected through annual online surveys of MNOs, MVNOs and NRAs, which are jointly run by the Commission and BEREC. So far, two such online surveys gathered in June 2018 and March 2019 information on the implementation of fair use policy, of the sustainability derogation and other elements of implementation. A third online survey has been launched in March 2020 covering various elements to be assessed under this review such as QoS, value added services, emergency communications, technology changes and M2M, cost of implementation and administrative burden.

III.Market monitoring and data analysis based on BEREC and Commission data collection - The Commission analysed independently the data, collected by BEREC based on which the following analysis was performed:

a.The Report on the Review of the roaming market COM(2019) 616 final, and its accompanying SWD(2019) 416 final adopted on 29 November 2019, available here , has gathered and presented evidence on how this intervention has performed. It has also assessed whether there were unintended/unexpected effects like for example increases in domestic or international roaming prices.

b. The interim report to the European Parliament and the Council on the implementation of the Roam-Like-At-Home (RLAH) rules over the first 18 months Adopted on 12 December 2018, available here .

c.The Staff Working Document (SWD) on the findings of the review of the rules on fair use policy and the sustainability derogation laid down in the Implementing Regulation (EU) 2016/2286 published on 28 June 2019, available here .

IV.BEREC Opinions and inputs – Consultation of BEREC resulted in extensive BEREC input, including forward looking elements and comprehensive analysis on the monitoring of the roaming market.

a. BEREC provided its formal Opinion on the functioning of the roaming market 3 available here , published on 19 June 2019,

b.as well as a supplementary cost analysis 4 , published on 20 September 2019, available here and

c.an input to the impact assessment on Commission request 5 (June 2020, available here ).

V.AXON Cost Model for Assessment of the cost of providing wholesale roaming services in the EEA. In order to estimate the costs of providing wholesale roaming services in the EU/EEA, the Commission ordered an external study to AXON Partners (SMART 2017/0091), available here . The study was conducted between March 2018 and June 2019. The contractor has developed a cost model to estimate the costs of providing wholesale roaming services in the EU Member State and EEA countries. The cost model was developed in close collaboration with the BEREC International Roaming Working Group and was submitted twice to public consultation. Further, a designated Steering Group composed experts from six different NRAs followed the project closely from start to finish.

VI.WIK Consult (SMART 2018/0012) - The Commission ordered to WIK Consult an external study on technological and market developments that might have an impact on the roaming market, available here . The purpose of the study was to assess the availability and QoS, which are an alternative to regulated retail voice, SMS and data roaming services, in particular in the light of technological developments. The study was conducted between December 2018 and June 2019. The contractor followed a modified greenfield approach to assess the competitive impact of various technological developments in the roaming market.

VII.Joint Research centre (JRC) analysis – JRC has provided extensive support in preparation of the IA, in particular:

a.developing the sustainability model (annex 4A) and the Consumer benefit counterfactual analysis (Annex 4B)

b.JRC initiated a project (SMART 2018/0011) on measuring the quality of roaming services in view of an evidence-based assessment of roaming performance and user experience of QoS of retail roaming services. JRC carried out field tests on mobile broadband in order to collect data to compare roaming performance in several EU MSs. The network performance data collected were used to compare home and roaming network performance.

VIII.The Mobile Broadband Prices in Europe SMART 2018/0045 published yearly by the Commission reports on the level of domestic retail mobile prices in EU Member States and in a number of non-EU countries. For the purpose of cross-country comparisons, the study uses the standard mobile internet usage baskets defined by the Organisation for Economic Co-operation and Development (OECD).

IX.Targeted online survey on potential impact on the roaming market of technological evolution 6 addressed to MNOs, MVNOs/MVNEs and business stakeholders/vertical industries was organised during the first half of March 2018, in the framework of the SMART study 2018/0012 “Technological developments and roaming” – the contractor received 51 responses that informed the analysis and the conclusions of the study.

X.Flash Eurobarometer Survey 468 7 – the survey was conducted one year after the end of roaming charges (June 2018). It provides a better understanding about the impact on consumers of the roaming regulation. It contained questions that covered roaming use when travelling within the EU, asked about the awareness of the end of roaming charges, perceived benefits, perceived costs of roaming among non-travellers and the use of mobile services in other EU countries. Another Eurobarometer survey is planned for 2020 where some specific roaming elements will be investigated, e.g. QoS.

XI.Roaming: One year after implementation, published by the European Parliament in November 2018. It was prepared at the request of the European Parliament’s Committee on Industry, Research and Energy (available here ). The study reviews the retail and wholesale roaming markets during the first year after the introduction of the RLAH regime. It acknowledges its benefits to the retail roaming market and makes certain recommendations focusing mainly on the wholesale regulation, including adjusting the wholesale caps.

XII.Input received to the Inception Impact Assessment feedback period 8  

XIII.Input from Public Consultation on the review and prolongation of the Roaming Regulation (EU) 531/2012 launched for 12 weeks, from 19 June to 11 September 2020. The consultation gathered information fulfilling the following objectives:

(1) Collect views on retail roaming services and on the impact of prolonging these rules. In addition, collect views on the impact of introducing clarifications and improvements to the Roaming Regulation, that are relevant for consumers, e.g. related to QoS, value-added services and emergency communications.

(2) Collect views on the provision of wholesale roaming services, the prolongation of these rules, on the sustainability of wholesale caps and on the need to respond to new technological and business developments (e.g. on Machine-to-Machine services and Internet of Things).

(3) Collect views on the possible reduction of the administrative burden and the impacts of possible simplifications.

Annex 2: Stakeholder consultation

Annex 2 presents main consultation activities’ outcomes and is divided in four sections:

Annex 2A: Summary of the Stakeholders Consultation Activities

Annex 2B: The Inception Impact Assessment Feedback

Annex 2C: The outcome of the Public Consultation

Annex 2D: The outcome of the Joint Commission-BEREC Online Survey 2020

Annex 2A: The Summary of the Stakeholder Strategy

A continuous and active stakeholder consultation strategy was devised and followed for the review of the Roaming Regulation. From the outset, key ideas for the prolongation and review of the Roaming Regulation were outlined in an Inception Impact Assessment (IIA). The published IIA informed citizens and stakeholders about the Commission's plans in order to allow them to provide feedback on the intended initiative and to participate effectively in future consultation activities. This fed into the subsequent consultation activities that ensured an inclusive process with all interested parties having an opportunity to contribute.

Several joint Commission/BEREC online surveys were launched, in June 2018, March 2019 and March 2020. The online survey gathered market inputs from MNOs, MVNOs and NRAs, on the implementation of the Roaming Regulation. The most recent survey covered various elements to be assessed under the review of the Roaming Regulation, such as QoS, value added services, emergency communications, technology changes and M2M, cost of implementation and administrative burden.

A dedicated 12 week public consultation was launched on 19 June to 11 September 2020. The Commission consulted EU citizens, the telecommunication operators, the industry and the administrations in the Member States, to gather information for the Impact Assessment (IA) of a Commission legislative proposal concerning the review of the Regulation (EU) 531/2012 on roaming on public mobile communications networks within the Union (Roaming Regulation). 9 The consultation activities aimed at collecting the views from all relevant stakeholder groups, and in particular of consumers and their associations, national regulatory authorities (NRAs), the Body of European Regulators for Electronic Communications (BEREC), Mobile Network Operators (MNOs), Mobile Virtual Network Operators (MVNOs), business stakeholders (including SMEs), government authorities and industry associations and other stakeholders. All these different stakeholder groups were expected to have important information and insights that could feed into the preparation of the Impact Assessment on a possible legislative proposal to the European Parliament and the Council, extending the Roaming Regulation beyond June 2022.

The other consultation activities were organised along the following streams that are further detailed in Annex 1:

I.The publication of the Inception Impact Assessment for a four-week feedback 10 and of a Public Consultation with targeted questions depending on the type of the respondent

II.Consultation of BEREC and market monitoring, which resulted in extensive BEREC input, including forward looking elements and comprehensive analysis on the monitoring of the roaming market:

a.BEREC Opinions and Inputs: Opinion on the functioning of the roaming market - BoR(19)101, a supplementary BEREC cost analysis of wholesale roaming costs BoR(19)168, and BEREC input on EC’s request for the preparation of the legislative proposal for the new roaming regulations, BoR (20) 131

b.BEREC Reports: the semi-annual international roaming benchmark reports 11 and BEREC annual report on the transparency and comparability of roaming tariffs 12  

c.Data collection by BEREC

III.Joint Commission/BEREC online survey

IV.Targeted interviews on roaming market evolution with actors of various typologies involved in the global connectivity value chain  

V.Targeted online survey on potential impact on the roaming market of technological evolution 13  

VI.Flash Eurobarometer Survey 14 – the survey was conducted one year after the end of roaming charges (June 2018).

Annex 2B: The Inception Impact Assessment Feedback

In total there were 11 feedback replies on the IIA: 5 company/business organisation, 2 business associations, 2 EU citizens, 1 consumer organization and 1 NGO.

The main points addressed by the stakeholders were wholesale caps, QoS, M2M/IoT, emergency communications, VAS and misuse/fraud. In general the stakeholder feedback can be summarised as follows:

One respondent suggests maintaining the wholesale caps in the current Roaming Regulation, while three respondents suggest lowering the wholesale caps. The main argument for lowering the wholesale caps is that MVNOs pay higher wholesale rates than MNOs. On the other hand, the respondent who suggests maintaining the current regulated levels, argues that the wholesale rates are already below the cap and that further reduction of the caps would lead to a disincentive to maintain these discounts.

Five of the respondents do not support obligations on offering the same QoS while roaming in EU/EEA as at home. However, two of these respondents do support transparency requirements in regards to QoS and one supports clarifications as regards to deliberately lowering the QoS. One respondent does support further clarifications as regards to obligations related to QoS.

Four respondents note that the scope of the Roaming Regulation should cover M2M/IoT services, while one respondent does not support any roaming regulations on this market. Another respondent stresses the importance of a thorough impact assessment of any such intervention.

Three of the respondents do not support any intervention as regards seamless access to emergency communications, as the communication to “112” is working well in their view. Two respondents noted that alternative emergency communication solutions vary across EU Member States, thus making it difficult to implement unified rules. Two respondents welcomed a further assessment of emergency communications, but requested a clarification as to why such an intervention would be justified.

Four respondents support further assessment of value-added services, or numbers associated with higher costs. One respondent suggests that the current “best practice” is better suitable to address such issues than regulatory intervention. Three respondents support further assessments of how to address fraud cases in roaming, while one stakeholder suggests that this problem is better resolved by market best practice.

From the consumers’ perspective, BEUC indicates that consumer awareness about the scope of the roaming rules is insufficient, for example there is a difficulty to distinguish between international roaming in the EU/EEA and intra-EU calls, since consumers may not be aware that RLAH does not apply on international calls from the home country. There might be also unawareness of the non-application of RLAH on ships and planes. BEUC furthermore suggests to abolish the price difference between international roaming in EU/EEA and intra-EU calls, address the lack of transparency regarding value-added services, strengthen the transparency requirements on data FUP limits and making the FUP rule less restrictive for consumer so that they can fully enjoy RLAH. BEUC suggests to lower wholesale caps, since this will allow less restrictive fair use policies.

Annex 2C: The outcome of the Public Consultation

The factual summary report summarising the main outcome of the public consultation carried out for the review of the telecoms framework has been published on the Commission’s web page .

Introduction

The European Commission ran a public consultation on the review and prolongation of the Roaming Regulation (EU) 531/2012 for 12 weeks, from 19 June to 11 September 2020. The consultation aimed to gather information for the Impact Assessment of a Commission legislative proposal for the review of the Roaming Regulation. This initiative is included in the Commission’s Work Programme 2020 under the Policy Objective “A Europe fit for the Digital Age” addressing the specific objective “Digital for consumers”.

In particular, the stakeholder consultation had the following objectives:

(1) Collect views on retail roaming services and on the impact of prolonging these rules. In addition, collect views on the impact of introducing clarifications and improvements to the Roaming Regulation, that are relevant for consumers, e.g. related to the QoS, value-added services and emergency communications.

(2) Collect views on the provision of wholesale roaming services, the prolongation of these rules, on the sustainability of wholesale caps and on the need to respond to new technological and business developments (e.g. on Machine-to-Machine services and Internet of Things).

(3) Collect views on the possible reduction of the administrative burden and the impacts of possible simplifications.

This report analyses the replies to the public consultation. The online questionnaire had a mix of closed and open questions in 23 EU languages. The questions were both backward and forward looking, as well as seeking input on impacts of potential policy options. Out of the 175 respondents, 16 respondents sent position papers in attachment to the questionnaire replies. An additional 4 respondents sent separate position and non-position papers via e-mail, and did not reply to the questionnaire.

The consultation targeted a broad range of stakeholders listed below according to their interest and presumed expertise in the subject matter.

Table 1: Respondents to the Open Public Consultation

This report uses the above categorisation of stakeholders in presenting converging or differing views on issues addressed in the consultation. The contributions of the stakeholders who gave their consent to publication are available online .

Out of 175 respondents from 26 countries, 108 were EU citizens and 2 non-EU citizens, 5 consumer organisations, 10 business associations, 5 public authorities, 3 NGOs, 42 company/business organisations of which 25 were Mobile Network Operators or representing them, 9 Mobile Virtual Network Operators/Enablers or representing them and 8 trade or other business stakeholders (e.g. vertical industries, SMEs).

The respondents' profile reflects the self-selecting nature of a public consultation and imposes a certain caution on interpreting the results, since the responses cannot be considered as a representative random sample of all European users. For instance this public consultation does not represent the views of national electronic communication regulators, which were subject to a targeted consultation and did not participate widely in the public consultation (3 national regulatory authorities and 2 government authorities). In fact BEREC, the Body of European Regulators for Electronic Communications, did not participate in the public consultation since they provided extensive input to the Commission on three occasions:

-BEREC Opinion on the functioning of the roaming market as input to EC evaluation, BoR(19)101, 19 June 2019, available here

-BEREC supplementary analysis on wholesale roaming costs, BoR(19)168

-BEREC input on EC’s request for the preparation of the legislative proposal for the new roaming regulation, BoR (20) 131, of 30 June 2020

The number of mobile operators participating in the consultation was limited. For comparison, it can be noted that 25 MNOs and 9 MVNO/Es participated in the public consultation. The respective number participating to the target joint Commission-BEREC online survey 2020 (see annex 2D) was 85 MNOs and 110 MVNOs.

Analysis of responses

All online contributions were analysed when drafting this report, and used for further analysis in the impact assessment. The European Commission services have taken into account the position and non-papers received in analysing the contributions to the public consultation.

Citizens could opt-in to reply to sections 3-9. Within in a section, all closed questions were mandatory (it was not possible to skip them), but some questions were only visible to targeted stakeholders. The percentages below refer to the number of actual respondents to the given questions, not to the total number of respondents in the consultation.

The report follows the structure of the questionnaire to the most part. It includes 9 sections addressed to various stakeholders.

The consultation did not contain questions on the combined package including all the measures covered by the four different policy options assessed in the Impact Assessment. It asked for views on the specific measures for each thematic area that have been proposed within the different policy options.

This analysis does not represent the official position of the Commission and its services and thus does not bind the Commission.

Benefits for end-users and overall functioning of the Roaming Regulation

On the perceived benefits of the Roaming Regulation, the overall perception is positive. These questions were primarily targeted at citizens and consumer/citizen organisations (in total 118 respondents).

The respondents of the public consultation confirm the overall success and the effectiveness of the Roaming Regulation. 96% of the citizens strongly agree (87%) or agree (9%) that they can enjoy the benefits that the Roaming Regulation aims to bring. The benefits include staying connected without having to restrict their usage of roaming mobile services when travelling in the EU/EEA, not worrying about having to pay excessive costs for the use of mobile services while travelling in the EU/EEA and continuing to use mobile services like at home while travelling in the EU/EEA. 65% of all the respondents replied that the Roaming Regulation has significantly promoted the interests of the citizens and businesses in the EU/EEA. 21% think that the achievement has been moderate. When asked about the achievement of the development of the Digital Single market, 48% think that the Roaming Regulation has been significantly effective in the development of the Digital Single Market, while 27% think it has been moderately effective, and 8% think it has had little effectiveness.

On the overall functioning of the Roaming Regulation at retail level, the overall perception is positive. 89% of all the 175 respondents across all respondent groups agree that the EU intervention has had a positive effect on ensuring that roaming users do not pay excessive roaming prices in the EU compared to what could be achieved by Member States themselves. The respondents express that there is continued need of the Roaming Regulation for EU citizens and businesses. As regards the relevance of the Roaming Regulation, 74% of all respondents think that the relevance is significant. 83% of all respondents agree that end-users would lose the benefits of using mobile services like at home, without regulation of EU roaming. Approximately 10% of the respondents disagree and 7% do not take a position. The vast majority of citizens and consumer organisations strongly agree that the benefits would be lost. Among the business organisations and associations, which represent 30% of the respondent, half expressed that they agree or strongly agree that the benefits would be lost, while 33% of them disagree or strongly disagree.

The question on the efficiency of the Roaming Regulation was answered by 57 stakeholder, in particular by business associations and company organisations. It appears difficult for many respondents to compare the costs towards the benefits, therefore one third of those who replied declared they do not know (12%) or had a neutral opinion (21%). Of those respondents who expressed a view, 47% (18 respondents) believe that benefits exceeds the cost or significantly exceed the costs, while 52% (20 respondents), believe that costs exceed benefits, or significantly exceed benefits. In total 20 out of 34 M(V)NO/Es and their associations think that the costs exceed the benefits, while 6 think that the benefits exceed the costs. All public authorities (5 in total) agree that the benefits exceed the cost.

Among the MNOs’ that think the costs exceed the benefits of the Roaming Regulation, the main reasons cited are the resources and costs incurred at the time of introducing RLAH, especially with regard to implementing the monitoring mechanism of the fair use policy and the sustainability mechanism. Among the MNOs that indicated that significant benefits for end users exceed the costs, some also indicate that the reduction in roaming charges have been compensated by higher demand, thereby by higher revenues. Some indicated that abolishing roaming charges induced a positive level of additional competition, as for example by increasing the EU-allowance in the tariff plans or including non-EU destinations in the mobile subscription.

Emergency communications and public warning systems

The EU legislation on emergency communications ensures that all EU citizens have free access to emergency services through the single European emergency number "112". End-users with disabilities should benefit from equivalent access through SMS, emergency applications, total conversation, etc. Roaming customers should enjoy the same level of service, when it comes to emergency communications as national customers.

41% of respondents (51 in total) agreed or strongly agreed with the statement that communications with emergency services works well. Respondents that gave a positive answer were individual citizens and one public authority. These answers indicate that the emergency communication through a voice call to 112 seem to work well. On the other hand, two non-governmental organizations (NGOs) for end-users with disabilities, the European Disability Forum and the European Union for Deaf, disagreed and, strongly disagreed with the statement. This answer indicates that, while calls to emergency services may work well, the alternative means of access do not seem to function with the same effectiveness for end-users with disabilities.

The answers on the awareness on the alternative means of access for end-users with disabilities indicate an important information gap in this respect. The majority of answers (76%), including the relevant NGOs (European Disability Forum, European Union for Deaf and European Emergency Number Association) and three public authorities, indicate a total lack of awareness on these means of access. The abovementioned NGOs strongly disagree with the statement that the access to emergency services works well for end-users living with disabilities. The wide majority of citizen respondents did not know or had no opinion.

The proposal to introduce an obligation on mobile operators in the Roaming Regulation at wholesale level to ensure that customers have seamless "free of charge" access, to emergency services by using 112, was considered relevant by 30% of respondents, while it was considered not relevant by 26% of the respondents.

As regards the proposal to introduce an obligation on mobile operators in the Roaming Regulation, at wholesale level, to ensure that the establishment and transmission of caller location in the visited EU Member State is free of charge for the end-user, the results were similar. 26% of the respondents expressed that this measure would be relevant, while it was considered not relevant by 28% of the respondents. As regards the proposal to introduce an obligation on the home operator to inform disabled end-users on the availability of alternative means of access available in the visited EU Member State (opt-in), gained equally mixed reactions. Amongst those respondents that expressed their opinion on the relevance of the proposal (52%), 16 found it relevant and 18 did not find it relevant. However, among those who considered the proposal relevant (relevant (11) or very relevant (5)) the European Disability Forum, European Union for Deaf and European Emergency Number Association consider it very relevant.

Review of wholesale roaming markets

In total 65 respondents replied to questions on the review of the wholesale roaming markets. Respondents in general acknowledge that the EU intervention had a positive effect on (a) the sustainability of the roaming markets and (b) competition on the wholesale roaming markets. In parallel, the vast majority of respondents do not consider that sustainability depends on the Member State. In addition, the vast majority of respondents do not consider that there exist Member States where the EU intervention did not have a positive effect in ensuring that roaming markets are more competitive, compared to what could be achieved by Member States acting alone.

According to the respondents, the main factors that influence the sustainability of the wholesale roaming markets are the existence and size of a positive or negative imbalance of roaming traffic. The same factors plus the number of mobile operators influence most, according to the respondents the competitiveness of the wholesale roaming markets.

Almost half of the respondents express the view that retail roaming services are not sustainable with the current wholesale roaming caps. The majority of M(V)NOs agree with this view (28 out of 52 respondents), as well as the consumer organisations (5 out of 5) and public authorities (5 out of 5). Opposing views do not exceed 15% of respondents. In parallel, almost half of the respondents express the view that retail roaming services are not sustainable with the current wholesale roaming caps, effectively supporting a reduction to wholesale caps. By contrast, only 1 out of 5 respondents support the opposite view. All responding MVNOs (9 out of 9) express that retail roaming services are not sustainable. MNOs are more divided and 11 out of 25 think that retail roaming services are not sustainable with the current wholesale roaming caps, 9 think that they are and 5 remain neutral. The most important obstacles to the sustainability of retail roaming services appear to be the (high) level of wholesale caps and the lack of bargaining power of smaller mobile operators. The vast majority of respondents do not consider that these obstacles depend on the Member State.

More than half of the respondents (54%) agree or strongly agree with the statement that wholesale roaming markets depend on wholesale price regulation to be able to provide RLAH at the retail level in line with the domestic charging model. Less than 1 out of 5 disagree or strongle disagree. A substantial number (1 out of 3) either does not express an opinion or remain neutral. On the other hand, respondents present diverging views on policy options, regarding the wholesale price caps. There is balanced support for maintaining them at the current level or reducing them. On the other hand, there is very little support in favor of lifting wholesale regulation and (expectedly) no support at all for increasing caps. The majority of respondents expect that the possible impacts of lifting wholesale price regulation would be negative, specifically increase of wholesale roaming prices and retail domestic prices and a decrease of the sustainability of MVNOs. On the other hand, they are more reserved towards the possible impacts of lowering wholesale rates and more or close to half either do not express a view or express a neutral one. Respondents were also reserved to express a view on alternative measures, including an MNO obligation to pass on discounts they get on wholesale roaming prices to MVNOs and a measure to ensure that regulated maximum wholesale caps also apply to alternative wholesale agreement solutions. Out of the respondents who expressed a view 16 out of 17 MNOs do not think such an obligation would be relevant, while 7 out 9 MVNO/Es think it is relevant. A limited number of other respondent group replied to this question, however the small number of responding consumer organisations and public authorities in general support this option.

Quality of Service (QoS) of regulated roaming services

The Roaming Regulation Article 3(3) requires that wholesale roaming access shall cover access to all network elements and associated facilities, relevant services, software and information systems necessary for the provision of regulated roaming services to customers. Out of 65 respondents, including MNOs, MVNOs, business associations, citizens and others (including SMEs/entrepreneurs/vertical industries/IoT & M2M and industry associations), 46% consider that the wholesale roaming access obligation in the current Roaming Regulation is sufficient to ensure that M(V)NOs are given access to 4G and 5G for wholesale roaming. 31% of the respondents do not think that the current wholesale obligations are sufficient. MNOs in particular, but also some business associations agree that the current provision in the Roaming Regulation is technology neutral. Several MVNO/Es on the other hand welcome a clarification. They note that they have experienced long delays in being granted access to 4G networks. They fear that bottlenecks on 5G roaming could emerge. Therefore, the Roaming Regulation needs to be reinforced, making clear that visited networks must grant wholesale roaming access (incl. resale) for all technology generations deployed in their network (e.g. 2G, 3G, 4G incl. LTE-M and NB-IoT, 5G). A clear provision could avoid delays and unnecessary costs.

The majority of MNOs and MVNO/Es agree that the current Roaming Regulation is sufficient to ensure that roaming consumers are given access to newest network generations (e.g. 4G, 5G) while roaming when 4G or 5G is/will be available. Business stakeholders tend to disagree. Both stakeholder who agree and stakeholder who disagree are inclined to raise the same argument, in particular that the QoS is dependent on the visited network and as such it cannot always be guaranteed that the same QoS level as at home is offered.

The majority (52% or 23 in total) of M(V)NO/Es and business associations express that they can easily gain wholesale roaming access to all network elements and associated facilities, relevant services, software and information systems, necessary for the provision of regulated roaming services to customers, from MNOs in other EU/EEA Member States. However, 30% of the respondents experience some difficulties in getting access while others do not get it at all because of the difficulties. The main reasons stated for these difficulties among MVNOs are the lack of possibility to negotiate with MNOs on their own (negotiations through wholesale resellers), restricted access to 4G and MVNOs experiencing long delays (several years) in being granted wholesale access to 4G networks, including for wholesale roaming. MNOs note that difficulties to gain access are related to capacity constraints/availability and competing MNOs being excluded from the newest service, by groups of MNO’s with broader EU footprint.

As regards ensuring the same QoS as at home, out of all respondents (141) 49% agree, and 28% disagree, that the current Roaming Regulation is sufficient to ensure that roaming users are given access to newest network generations (e.g. 4G, 5G) while roaming when 4G or 5G is/will be available. MNOs mostly agree (21 out of 24) but MVNOs are more inclined to disagree (5 out of 9, while 3 out 9 agree) with this statement. Among stakeholder groups, for example citizens and consumer organisations 25 out of 84 who expressed a view disagree that the current Regulation sufficiently ensures access to all network generations, while 45 agree.

Mobile operators indicate that they actively try to ensure QoS of retail roaming services through traffic steering techniques (22 out of 30). According to the respondents steering roaming traffic is a dynamic way of increasing QoS for the end-user. This practice enables the domestic operator to offer the end-user roaming services on chosen networks based on quality parameters. This practice does not limit the access to other networks for the end-user, e.g. in case of poor coverage or other service limitations at any given time the end-user will automatically be given access to another network. In this regard, dynamic steering techniques allows the domestic operator to manage the best QoS for the domestic customers when roaming. 68% (28 in total) of M(V)NOs apply roaming traffic steering techniques in the EU/EEA (i.e. routing their own customer traffic while roaming) to one specific visited network. 30% (13 in total) steer traffic from a visited network, to a domestic network.

There is an overall support from stakeholders for strengthening the QoS requirements in the Roaming Regulation. In particular citizens and consumer organisation support additional measures, while the mobile operators are more divided. MNOs are less inclined to agree that additional measures on QoS would be relevant, while MVNO/Es are overall more supportive. As regards the option to include a transparency obligation, e.g. requiring mobile operators to provide clear information about the QoS in the visited country, 63% of all respondents, including all respondent groups, have replied that it would be very relevant. The prevailing respondent group that finds this option relevant are citizens and consumer organisations. M(V)NOs are less positive to such an obligation; 15 out of 34 do not think it is relevant, 9 are neutral and 6 are positive. Among the other business stakeholders, 2 out of 5 respondents think it would be relevant, 2 are neutral and 1 thinks it would not be relevant.

As regards additional QoS obligations on wholesale and retail level respondents are almost equally supportive of introducing obligations on the home mobile operator (73%) as on the visited operator (78%), prohibiting deliberately lowering the QoS for roaming services. The majority of these respondents are citizens and consumer organisations.

MNOs are less inclined to support such obligations. MVNO/Es on the other hand are more positive, and support in particular obligations on the visited operator. Other business stakeholders think such a measure would be relevant.

Lastly, regarding the obligation on MNOs to give access to all network technologies and generations(2G, 3G, 4G, 5G etc.), upon a reasonable wholesale roaming access request, a strong majority, 75% of all respondents, are supportive. The majority of these respondents are citizens and consumer organisations. 7 out of 25 MNOs have expressed that such an obligation would be relevant, while 8 do not think it is relevant and 6 are neutral. Among the MVNO/Es, 7 do think it is relevant while 2 are neutral. Among the other business stakeholders, 4 out of 5 think such a measure would be relevant.

Roaming and Machine-to-Machine (M2M) services and Internet of Things (IoT)

As regards the measures in Article 3(4) of the Roaming Regulation (to be read together with Recital 11), that allows mobile operators to negotiate alternative pricing schemes for wholesale roaming, 19 out of 65 respondents disagree (or strongly disagree) with the assertion that the current rules are sufficient to develop more efficient, integrated and competitive markets for roaming services for M2M, while 14 respondents agree (or strongly agree). Among those who do not think current measures are sufficient, some find that less regulation and more commercial flexibility is needed, while others on the contrary would welcome targeted regulation for the M2M wholesale market. MNOs consider that regulatory intervention would be irrelevant or premature. Vertical companies (i.e. that use the network for industrial applications) and most MVNOs consider that intervention for M2M would be welcome.

A consensus seems to emerge to support the introduction of a clear distinction of M2M from person-to-person communication.

At wholesale level, MNOs rather disagree with the need for specific rules for M2M in contrast to MVNOs and verticals. A vast majority the respondents across categories would not support specific rules at retail level.

Regarding permanent roaming for M2M, MNOs find that it would affect competition and possibly open a back door for unauthorised entry into other national markets. MVNOs either have no opinion or on the contrary they rather support permanent roaming for M2M. Regarding the impacts of such an obligation some MNOs warned of the risks of fraudulent use of the EU roaming regulation would increase with the possibility to roam permanently. 7 respondents noted that an extension of the benefits of the Roaming Regulation to IoT/M2M users, would in practice grant non-EU providers preferential access to EU markets, without reciprocity for EU providers on non-EU markets (i.e. if the regulated wholesale access rights and prices can be used for permanent roaming, in practice the obligations would expand to the national roaming market). MNOs warn that intervention could disadvantage cellular vs non-cellular connectivity.

Regarding the idea of obliging MNOs to negotiate wholesale agreements in good faith or making permanent roaming explicitly enabled with alternative pricing models, MNOs have different and balanced views while MVNOs and verticals rather agree. Most respondents across the categories agree that an obligation on MNOs to provide separate wholesale agreements for permanent international roaming in EU/EEA for M2M/IoT communications would decrease the bargaining power of MNOs and increase the one of MVNOs. While MNOs respond that it would decrease their sustainability but would not lower the rates, nor increase competition, nor increase the level of EU connectivity, nor facilitate access to the market, MVNOs and verticals are of the opposite view.

As regards separate wholesale regulation of M2M/IoT services imposing a cap on wholesale prices for permanent international roaming for M2M/IoT communication, with non-volume based pricing for wholesale access, MNOs see negative effects while MVNOs see positive effects. Vertical industries raise the need to use roaming services for connected objects. One public authority indicated that setting a single wholesale cap to cater for all M2M/IoT connectivity services use cases would be difficult and imply the risk that MNOs do not recover their efficiently incurred cost. Two large associations of telecom operators also consider that the Commission should focus on alleviating the challenges that the operators experience during the cross-border deployment of enterprise and consumer IoT applications, because of the existing fragmentation across Member States. Another association mentions that if roaming is artificially restricted or priced without a clear connection to underlying costs, the benefits will likely fail to materialize. One business respondent suggest electronic trading as part of the solution to offer price competition.

Roaming and technological developments

On the potential impact of technological developments on the roaming markets larger M(V)NOs and associations do not foresee any considerable changes in arrangements of roaming with 5G compared to other generations. Among the impacts that stakeholders have mentioned within the next 2 or 5 years are increased data traffic volumes, network slicing, increased use of M2M/IoT services and increased QoS requirements.

A large proportion of the respondents have not expressed an opinion or remained neutral on expected competitive pressure from various technological and business developments. When stakeholders have expressed an opinion the views are rather mixed between expected pressure and no expected pressure. The results indicate that stakeholders are more inclined to expect competitive pressure from certain technological and business developments. Of those stakeholders who have expressed an opinion, in particular MNOs are expecting that competitive pressure will be caused on voice and SMS service by OTT services and e-SIM. MNOs also expect that e-SIM will cause competitive pressure in particular on data roaming services. Some stakeholders, across the stakeholder groups, expect competitive pressure on data from public Wi-Fi services and Wi-Fi in the home/work place. On 5G and network, slicing the majority of respondent remained neutral or did not express any opinion.

Due to the lack of practical experience and the novelty connected to online trading platforms for wholesale roaming traffic, the majority of the 65 respondents have not expressed an opinion or remained neutral when asked about the willingness to use such platforms or the impacts of online trading platforms. Among the views expressed stakeholders noted in particular that online trading platforms, if introduced, should be on a voluntary basis. Several also noted that trading of roaming traffic does not only depend on the price but that other factors such as QoS and security requirements are significant.

Value-Added Services

There is general support for additional measures in the Roaming Regulation to mitigate bill-shock for end-users due to the use of Value-Added Services (VAS), and avoid inadvertent usage of such services. With regard to additional measures to avoid unexpected additional charges due to the use of value-added services while roaming in the EU/EEA, the majority, 72% of respondents confirmed such need. Only 12% of respondents (17) considered that there is no need to introduce measures to avoid unexpected additional charges. The replies prove that there is a general agreement among all groups of respondents that the measures to avoid unexpected additional charges due to the use of VAS while roaming in the EU/EEA should be introduced in the Roaming Regulation.

The public consultation confirmed a strong support (70% of respondents) for a database of VAS/PRN numbering ranges as a possible measure to address problems linked to value-added services and high termination rates subject to revenue share of roaming services in the EU/EEA, among all groups of respondents.

The proposal to introduce an obligation on mobile operators to include in the “Welcome SMS” an alert informing that these types of communications may not be under the RLAH principle was also supported by a majority of respondents (85, i.e. 62%).

The proposal to introduce an obligation that access to value-added services must be explicitly requested by the roaming end-user (opt-in) was again supported by over a half of respondents (78, i.e. 57%). This solution was mostly supported by citizens and consumer organisations. The respondents who did not support the proposal (43%), were quite evenly divided among those neutral and negative about it (53% and 47%, respectively).

Potential simplification/burden reduction

The relevant simplification measures that have been proposed by stakeholders to explore the potential for simplification and improving the efficiency of that legislation are summarised below. Stakeholders did not raise major problems related to simplification or burden reduction. These proposals have either been taken into consideration in the proposed options in the IA or discarded due to their limited relevance or regulatory limitations.

a. Retail regulation and transparency for customers

·One public authority suggests to simplify the annual revision of the regulated roaming surcharges, for currencies other than the euro. To address this problem the same date should apply for all retail surcharges that are laid down by the Roaming Regulation.

·Price information in the “Welcome SMS” should be the actual price paid, not maximum prices and a warning should be introduced against VAS.

·More efficient transparency obligations, e.g. mandatory information could be presented more transparently and better processed in an app than in an SMS.

b. Wholesale regulation

·Several MNOs suggest to keep price caps at the 2022 levels while MVNOs suggests further reduction of the wholesale caps.

·Separate wholesale charging model for M2M/IoT connections.

·Several M(V)NOs suggest the repeal of the obligation to provide local data break-out services.

·Wholesale FUP instead of retail FUP

c. Supervision/enforcement, dispute resolution, data collection

·Reduce and simplify the data and information gathering requests and extend deadlines for fulfilling the requests.

·Remove the “Welcome SMS” obligation as customers expect to incur no additional charges for EU/EEA roaming. The obligation to send “Welcome SMS” is therefore bringing no tangible benefits for customers anymore, but generates considerable costs and network load for operators.

d. Other

·The treatment of value-added services / premium numbers and price arbitrage under the Roaming Regulation should be clarified and, introduce a common database on VAS numbering ranges.

·Introduction of trading platforms that would introduce self-regulation mechanisms to the wholesale market

Annex 2D Joint Commission-BEREC online survey 2020

Introduction

The joint Commission-BEREC online survey was organised in March-May 2020 and targeted three focus groups:

·National Regulatory Authorities (NRAs)

·Mobile Network Operators (MNOs)

·Mobile Virtual Network Operators (MVNOs)

The survey was announced on 31 March 2020 with the deadline of 24 April 2020. Taking into consideration the difficulties faced by both operators and regulators, due to the COVID-19 pandemic, the survey remained open until 15 May 2020.

The number of responses is shown in the following table:

Table 2: Number of responses to joint Commission-BEREC online survey 2020

NRA Questionnaire

Compliance and infringements

Since 1 January 2019, 8 out of 28 NRAs have started 29 formal procedures for violations of Roaming Regulation. Most formal procedures concerned incorrect application of the FUP (7) or QoS (6). Following these formal procedures, in 9 cases the NRAs imposed administrative fines, ranging between €400 and €15,000 while 3 closed with a warning or moral suasion, 7 closed following the correction of the infringement and 2 closed with a decision that there was no infringement.

Quality of Service (QoS)

Since 1 January 2019, 8 out of 28 NRAs have received consumer complaints about the quality of the roaming services provided by their operators when travelling abroad in the EU/EEA (total 62 complaints). Out of the 8 NRAs, only 1 reports an increasing trend. 22 out of 28 NRAs report that they do not consider that QoS would require any action from their part and only 3 NRAs have either taken formal procedures or are investigating, if there are any violations of the roaming rules concerning QoS. At the same time, almost 2 out of 3 NRAs report that operators in their countries do not provide information on their websites about the QoS (3G vs 4G) offered to their roaming subscribers when travelling within EU.

Wholesale roaming access

In general, NRAs do not report issues concerning wholesale roaming access. Only 1 out of 28 NRAs received a request for authorization to terminate a wholesale roaming agreement according to Art 3 (6) and a request for conflict resolution concerning roaming access. The case concerned permanent roaming.

Technological developments / M2M communications

Only 6 out of 28 NRAs have received complaints from operators or have become aware in any other way of any issues with M2M-based permanent roaming from foreign SIM cards. Almost all concern permanent roaming and only one concerns wholesale charges for M2M roaming. Two concern the same case, of using foreign M2M SIM cards, to provide permanent roaming to consumers (non M2M). At the same time, only 1 out of 28 NRAs was aware of negotiations or requests, in which operators from other member states requested from or negotiated with operators in their country to establish wholesale agreements allowing permanent roaming. The NRA was not aware of any disputes in this process

Emergency communications – 112

The majority of NRAs (25 out of 28) do not seen to be aware of any mechanisms which ensure that roaming customers with disabilities are informed of the available means for non-voice access to emergency services, when they enter the NRA country.

Simplification and administrative burden

The effort that NRAs have put in 2019 concerning the implementation of the roaming regulation varies significantly. For general monitoring, the effort ranges from 20 person days or less to more than 300 but the majority of NRAs have spent between 1 and 6 person months, with a median of 30 person days. For formal procedures during 2019, 16 NRAs have not put any effort while only 1 NRA has spent more than 3 person months. The median of non-zero values stands at 26 person days. Finally for examining sustainability derogations during 2019, 17 NRAs have not put any effort while only 3 NRAs have spent more than 2 person months. The median of non-zero values stands at 22 person days.

Operator Questionnaire (MNOs / MVNOs)

Wholesale Quality of Service (QoS)

Only two MNOs (less than 2.5% of respondents) report including only 3G roaming access in their reference offer. One of them claims not having 4G access.

The majority of MNOs report not having any 3G only wholesale access agreements (62% when acting as host networks and 53.5% when acting as home networks). On the other hand, for a substantial minority of MNOs more than half of their wholesale agreements are 3G only: 16.3 when acting as host networks and 15.1% when acting as home networks. The outlook changes substantially in resale agreements, as more than 80% of MNOs report that they do not have any 3G only resale roaming agreements and less than 5% report that more than half of their resale roaming agreements are 3G only.

The outlook is less positive, when seen from the MVNO perspective, as 26% responded that only 3G is available either in certain specified countries in the EU (4%) or in certain networks in certain countries in the EU (14%) or in general in the EU (8%). However, the main reason cited by MVNOs for not being granted 4G access is the need for technical developments from their part.

Almost 3 out of 4 MNOs report that they intend to include 5G in their reference offer, when it becomes available.

Retail Quality of Service (QoS)

The vast majority of operators (97%) claim that they do no limit the QoS/ data speed of roaming services to 3G for their customers, while roaming in the EU. This is in contrast to the responses on wholesale roaming agreements.

On the other hand, there seem to be some (though not necessarily many) consumer complaints on QoS. From the around 150 operators that have provided data on the number of complaints per category 18% have received complaints on only max 3G available (2.1% report more than 1,000 complaints) and 22% have received complaints on no full 4G speeds possible (1.4% report more than 1,000 complaints). Lack of coverage seems to be most common complaint.

Misuse

Misuse seems to be an important issue, as almost half of the MNOs (but only 16.5% of MVNOs) have reported being aware of abusive use of SIM cards in voice and/or SMS roaming communications in the EEA which cannot be mitigated by the FUP control mechanisms foreseen in the Regulation. International revenue share fraud seems to be a major (but not the only) case of misuse.

According to the data provided in the survey, the financial impact seems to be quite substantial, even though just around 40% of operators that encountered abusive use of SIM cards provided concrete data on revenues lost. Based on the collected data, the median value of lost revenues is €70,000 yearly, while one quarter of operators report lost revenues in excess of 350,000 euros yearly.

The above do not take into account indirect non-financial impacts to operators, including (as reported in the questionnaire) increased resources, loss of business and reputation and increased consumer complaints.

While operators do not report an increasing trend, we can safely assume that inability to effectively combat misuses will inevitably lead to increase of the phenomenon and increasing impact.

Operators offer various suggestions to address misuse. Common suggestions include:

·Transparency measures regarding the numbers related to premium rate numbers and value added numbers, including a database with number ranges with high termination rates.

·Regulation of numbering plans to allow identifying number ranges with high termination rates, including harmonisation of prefixes for premium rate/ value added services in all Member States (which however falls outside the scope of the roaming regulation).

·Simplifying fair use policies (which however is outside the scope of the present review).

·Barring calls to or more generally closing down high-cost destinations/network codes within EU/EEA and globally.

·Increasing wholesale roaming caps (MNOs) or decreasing wholesale roaming caps (MVNOs).

Value Added Services (VAS)

One out of four operators (but more than one out of three MNOs) report having received complaints from their clients about communications related to value-added services while roaming in the EEA. The majority of these complaints concern bill shocks and lack of transparency on the cost of VAS while roaming.

In parallel, 1 out of 5 operators (but more than 30% of MNOs) report having incurred extra costs resulting from unexpected wholesale charges for communications related to value-added services by their customers while roaming in the EEA. There are sufficient data to allow sizing the losses that operators face, but in general we consider that such losses have been in general reported in the section on misuse.

Operators have employed different approaches to remedy such issues. Some have tried to collect information on VAS number ranges but acknowledge that ensuring up-to-date information on VAS number ranges is complex, due to the fragmented numbering landscape and the lack of EU wide rules on VAS. Others seek solutions by renegotiating their wholesale roaming agreements. In some cases, operators have opted blocking VAS but this has often resulted in consumer complaints. Few operators have reported such issues to the competent NRAs but without any result.

Operators propose various approached to address VAS related issues. The most frequently cited proposals are to harmonise VAS number ranges in all EU/EEA member states and to establish a VAS database. Few operators propose excluding VAS from RLAH, allowing operators to block VAS numbers, mandating transparency of wholesale VAS rates, regulating premium rate numbers and free phone usage or mandating relevant provisions in wholesale agreements.

Emergency communications – 112

While calls to 112 are routed to the most appropriate PSAP, non-voice communications for end-users with disabilities are, in most of the cases, not. For SMS and applications (59%) and, respectively, (64%) of home operators indicate that adequate routing is not ensured. In addition, home networks do not monitor (83%) whether the routing of the emergency communications is ensured by the host operators. In case of calls to 112, operators tend to rely on initial testing of the roaming services. Monitoring or testing of non-voice access is missing from the practice of MNOs. MVNOs rely mostly on the arrangements of their host MNO.

The provision of caller location is not ensured in the majority wholesale agreements, not for network based location (58%), nor for handset-derived location (74%). The majority of operators do charge at retail level the provision of handset-derived location through both SMS (55%) and data channel (57%). In turn, host MNOs indicate that only 84% of them ensure the provision of network based caller location to customers roaming in their network. In case of handset-derived AML localisation, this is available only in half of the cases when AML is deployed through SMS transmission and even less when AML is deployed through data transmission (36.4%).

In contrast with the calls to 112 (85%) a very low percentage of MNOs ensures through the roaming agreements that end-users with disabilities have access to emergency services. Only 34% of operators provide for the possibility to access emergency services through SMS, where such means are deployed in the visited country. In case of emergency applications for end-users with disabilities the percentage is even lower: 26%. The failure to share information and failure to ensure the adequate division of responsibilities in the wholesale agreement is reflected in the fact that end-users with disabilities are charged for the means of access employed by the home operators: 56% of operators in case of SMS and 68% of operators in case of emergency applications. The responses of the host MNOs confirm that an important share of roaming customers living with disabilities are precluded to have access to emergency services through the means of access deployed in the visited Member States. In case of SMS only 62%, in case of emergency applications only 58% and in case of other mean only 58% of MNOs would ensure access to emergency services.

A significant bottleneck for end-users with disabilities to access emergency services when roaming is the lack of awareness. While other end-users are informed that they could call 112 free of charge when they enter in an EU Member State, end-users with disabilities are not prompted to this relevant and potentially lifesaving information. According to the responses provided 90% of all operators (92% of MNOs and 88.6% of MVNOs) do not inform roaming customers with disabilities on the available means of non-voice access to emergency services in the visited country. This situation puts end-users with disabilities in a comparatively more vulnerable situation than other end-users that are informed when entering and EU country that they may call 112 free of charge in case of emergency. In view of the variety of alternative (non-voice) means of access deployed in the EU it is all the more important that end-users with disabilities are provided with the relevant information on how to access emergency services in the visited country.

Only 16% of both home and host operators are informing roaming customers on the public warning systems deployed in the visited EU country. This information can be potentially lifesaving when visiting another EU Member State that deploys an alerting technology that needs an end-users action to enable his/her device to receive the public warning messages.

Technological developments / M2M communications

The responses of operators to the question about whether technologies or techniques could work as alternatives to the classic data roaming services, confirm the conclusions of the study on technological developments and roaming. The use of eSIM seems to be the most promising development but still does not seem to be mature enough to provide a viable substitute in the medium term.

The majority of MNOs and a small number of MVNOs offer M2M services. However only 1 out of 2 MNOs and 1 out of 4 MVNOs seek to establish specific agreements for M2M, which might imply that often those services are enabled through standard roaming agreements. In the limited cases when specific agreements are requested, most probably also to address the need of permanent roaming, some concerned operators report facing difficulties.

On the other hand 1 out of 4 MNOs report having experienced issues with M2M-based permanent roaming from foreign SIM-cards in their network, including unauthorised use of (M2M intended) SIM cards for non M2M services.

The above reflect that the provision of international roaming services specifically intended for M2M communications could benefit from clarifications.

Simplification and administrative burden

The effort that NRAs have put in 2019 concerning the implementation of the Roaming Regulation varies significantly. We also observe, as expected, a significant difference in the effort put by MNOs and MVNOs.

Reporting on the implementation of the Roaming Regulation: more than two thirds (68.2%) of the operators that provide data, report an effort of up to 1 person month, with a median of 12 person days (20 for MNOs and 12 for MVNOs).

For the remaining categories, a large minority (ranging between 27% and 42%, depending on the question) did not provide data or report that they cannot distinguish these costs from the data provided for the first question. This implies that for a substantial number of respondents the above is the all-inclusive cost for the implementation of the roaming regulation.

Running (not implementing) the transparency obligations: The reported values are equally distributed in 4 groups: 0; 0-1,000; 1,000-10,000; 10,000-100,000 (each with 23-25%). The median is 3,000 (10,000 for MNOs and 1,000 for MVNOs). The third quartile value is at 15,000 (38,000 for MNOs and 15,000 for MVNOs).

Applications for sustainability derogations: Around two thirds of the operators that provide data do not report any effort (which is expected). The median for those who report some effort is 20 person days (18 for MNOs and 20 for MVNOs)

Running (not implementing) the fair use policy: 17.5% of operators (the majority being MVNOs) report 0 effort, which reflects operators that do not implement an FUP themselves. The median for those who report some effort is 20 person days (30 for MNOs and 10 for MVNOs).

Annex 3: Who is affected and how?

Practical implications of the initiative

Summary of costs and benefits

Table 3: I Overview of Benefits (total for all provisions) - Preferred Option

The following two tables present the summary of costs. The first analyses the costs incurred by each measure while the second gives a comprehensive view of all costs involved per measure and category of stakeholder.

Table 4: Measures of the preferred option and costs they incur

Table 5: II. Overview of costs for the preferred option

Annex 4: Analytical methods

Annex 4 presents main analytical methods and is divided in four sections:

Annex 4A – Sustainability assessment: methodological overview

Annex 4B – Counterfactual evaluation of the impact of RLAH and assessment of benefits for consumers

Annex 4C – Cost model for the wholesale roaming cost

Annex 4A - Sustainability assessment: methodological overview

Introduction

The provision of RLAH under specific wholesale caps configurations, may hamper the sustainability of an operator’s charging model. Sustainability is threatened any time the cost for providing roaming services is higher than the expected revenues.

The Commission Implementing Regulation 2016/2286 (Article 10, par. 1), stipulates that the ability of a roaming provider to recover its costs of providing regulated retail roaming services, would be undermined, only where the negative roaming retail net margin is equivalent to 3% or more of its mobile services margin.

In this Annex we describe the analysis we have undertaken to assess the sustainability for mobile operators of the wholesale price caps we propose under Options 2, 3 and 4 of our impact assessment. Our assessment follows the approach described in the Regulation and was conducted by the Competence Centre on Microeconomic Evaluation (CC-ME) of the Joint Research Centre (JRC) 18 under the guidance of DG CNECT's services.

Data used to assess sustainability

The sustainability analysis has been made possible thanks to the data provided by the 19th – 25th waves of the International Roaming Benchmark Data Report conducted by BEREC. The data collected has been cleaned and rationalized in a panel database (BEREC database, hereafter) covering the period 2016 Q4 – 2020 Q1. The consumption forecasts are based on the information on outbound and inbound volumes from 2017 Q2 up to 2019 Q4, as well on the EUROSTAT monthly data on “Nights spent at tourist accommodation establishments”.

The final number of operators for which we have been able to assess sustainability is 96. We have to get rid of some operators because of missing data in the domestic variables, both on volumes and on revenues, as will be explained in the next section.

Moreover, we have been obliged to exclude Ireland and Luxembourg from the analysis because no data on the “Nights spent at tourist accommodation establishments” were available for these two countries.

Table 6: Number of MNOs and MVNOs per country

Roaming consumption forecasts

The sustainability assessment exercise requires to have forecasts on inbound and outbound roaming volumes up to 2025. The forecasts have been produced on the basis of the country quarterly level BEREC database 19 . In particular:

·we use quarterly level data in order to take seasonality into account when forecasting volumes. Therefore, we have 9 quarters available and we use them to forecast volumes up to Q4 2025;

·instead of producing forecasts at the operator level, we prefer to use country level aggregates to reduce the probability of strange outliers dynamics due to misreporting from operators.

We base our forecasts on the Nights spent at tourist accommodation establishments - monthly data by Eurostat. The rationale for the choice relies in the fact that the number of points in time available for the series to be forecasted is too little to produce reliable projections. The Nights series is strongly related to the series on inbound roaming volumes, and therefore we decide to use the former as the basis to forecast the latter. In particular:

1.Making use of an econometric model, we estimate the relationship between inbound tourism and inbound roaming traffic in the period in which we can observe both;

2.we forecast the inbound tourism flow and use the coefficients estimated before to compute the forecasts of the inbound roaming volumes;

3.we compute the ratio between inbound and outbound roaming traffic for each Member State and each quarter of the year, i.e. first, second, third and fourth quarter, on the basis of the information reported in the BEREC database, in order to adjust for seasonality;

4.we use the relationship between inbound and outbound volumes calculated in the previous point to obtain forecasts on outbound roaming traffic;

5.finally, we aggregate the quarterly forecasts to have annual projected volumes at the country level.

Upon country level forecasts, we retrieve operators' level forecasts using reported traffic in BEREC database to compute the ratio between the volumes produced by each operator and the total country level volume for each service, ending up with a proxy for the operator specific market share in each service. We then apply this ratio to forecast country level traffic to redistribute it among operators, and obtain operators' level traffic.

In view of the discussion in section 5, all forecasts ignore the COVID-19 pandemic impact, assuming that it will have elapsed by the time that the proposed Roaming Regulation enters into force. The potential impact of COVID-19 is examined in the framework of the sensitivity analysis.

Sustainability test based on operators’ data

The sustainability will be assessed comparing – at the operator level - the “Roaming Margin” with the “Domestic Margin”. The “Roaming Margin” is defined as the difference between the revenues obtained from the provision of retail roaming services and the costs of providing such services. On the other hand, the “Domestic Margin” is calculated multiplying the revenues from the provision of domestic services by the assumed retail domestic margin.

Therefore, the sustainability index is calculated as:

(1)

It reflects the magnitude – in percentage – of the roaming margin compared with the domestic one. Sustainability will be negative any time the roaming margin will be negative, i.e. when the costs of providing roaming services will overcome revenues.

In the remainder of the analysis we seek to estimate the number of operators with sustainability that is less than -3%, that is the roaming margin is negative and in excess of 3% of the domestic margin.

Roaming Margin

As already mentioned, the roaming margin is computed as follows:

(2)

Retail Roaming Revenues

Roaming revenues are obtained summing up the revenues coming from “exceeding FUP” volumes, those originated by alternative tariffs – for each service k (voice, SMS, data) and the proportion of domestic revenues which can be imputed to roaming traffic.

(3)

where

(4)

In order to retrieve FUP traffic the BEREC database has been used and the following methodology has been implemented:

·for each service, we compute the proportion of reported traffic generated under FUP over total roaming outbound reported traffic for the points in time available and then the average over time has been computed;

·we apply this ratio to the forecasts on outbound traffic to estimate future volumes;

For the sake of simplicity, we are presuming that the proportion of FUP over total roaming traffic will remain constant over time. Indeed, forecasting FUP traffic would be quite challenging because we have little knowledge about the determinants of their dynamics (as we have done with inbound forecasts where we exploited the fact that it is tourism flows what drives roaming consumption).

As for Alternative Tariffs revenues, we assume them to be constant and we take the 2019 values as the reference level.

To estimate the proportion of domestic revenues that can be attributed to roaming consumption () we need first of all to produce forecasts of the domestic revenues series.

Unfortunately, data on the domestic market contain many missing values and a handful of outliers. Moreover, similarly to the case of FUP volumes, we do not have any series which is correlated to the domestic revenues one and it is long enough to be used to produce forecasts as we have done for inbound volumes. It is also worth remarking that in the BEREC Benchmark Survey, domestic revenues are not disaggregated per service, but are reported for the three services together.

First, to tackle the missing values problem, we use the reported data available at the quarterly level and regress the domestic revenues series on the domestic volumes one. Then, we apply the coefficients of the regression to impute values for domestic revenues where these were missing. 20 It is not possible to implement this method whenever both revenues and volumes are missing, and hence we get rid of operators displaying such data configuration.

Second, once filled the gaps in the domestic revenues series, we produce forecast applying the following approach:

1.We compute the European average annual growth rate of domestic revenues as:

(5)

where is the quarter-to-quarter average growth rate from Q1 2018 up to Q4 2019. The quarter-to-quarter growth rate is then multiplied by 4 (the number of quarters) to retrieve an annual estimate, which is equal to 1.03%. The European level growth rate has been preferred with respect to the country level one in order to alleviate the problem of operators - and therefore countries - reporting outlier figures for domestic revenues.

2.Domestic revenues for each operator are computed as:

(6)

The growth rate computed at point 1 works as reference for 2019. Then, it is assumed to slightly increase of 10% on a year-to-year basis. Moreover, we take the operator level domestic revenues in 2019 as the initial level, and use the dynamics of to predict domestic revenues up to 2025.

Once predictions on domestic revenues have been made, we can compute the proportion of such revenues that can be imputed to roaming consumption.

As already mentioned, domestic revenues are not reported separately for service, and therefore it is not possible to use the proportion of each service roaming traffic over domestic one to impute roaming revenues. To tackle the problem we make the assumption that, given the importance data consumption have and will acquire in the future, the main driver of the proportion of domestic revenues to be imputed to roaming is the proportion of roaming data consumption over domestic data consumption, what we call “Roaming-to-Domestic” (roam-to-dom, hereafter) ratio. However, we only have roaming volumes predicted at the operator level, whilst no prediction is possible on domestic volumes at this level.

Therefore, since it is not possible to predict the components of the “roam-to-dom” ratio, we establish a reference level for it, and then estimate its future dynamics according to the following procedure:

1.We calculate the “roam-to-dom” ratio for each operator for 2018 and 2019 and we then take the average of the ratio over the two years as the initial reference level;

2.we calculate the “roam-to-dom” ratio at the European level for 2018 and 2019;

3.we compute the year-to-year growth rate of the “roam-to-dom” ratio at the European level between 2018 and 2019 21 and we take this growth rate as the initial reference level;

4.coherently with the evidence pointing out that roaming volumes will grow in the future faster than domestic ones even though their growth will not be exponential anymore 22 , we assume that the roam-to-domestic ratio will continue growing at a diminishing rate:

(7)

5.Finally, we apply the to the initial level of the ratio we've calculated at point (1) to project the ratio up to 2025.

Roaming Revenues are hence computed as:

(8)

Retail roaming costs

The costs of providing retail roaming services corresponds basically to the wholesale roaming cost beard by operators for the unbalanced traffic.

(9)

Forecasts of outbound and inbound traffic for each service k ( and respectively) have been obtained as described in section 2. In calculating roaming costs we assume unbalanced traffic to be either zero or positive. In this way, we rule out the possibility for a net receiver of wholesale roaming traffic to have a negative wholesale roaming payment balance (i.e. revenues at the wholesale level from the provision of wholesale roaming services

Domestic margin

The domestic margin is computed multiplying the revenues from the provision of domestic services by the assumed retail margin:

(10)

where Domestic Revenues are derived in the previous section.

Sensitivity analysis

In order to analyse how sensible are results on the sustainability index to variations in the main variables we are considering, we perform a sensitivity analysis. In particular, we design two alternative scenarios with the assumptions listed in the Table below:

Table 7: Sensitivity on sustainability scenarios: description

Low Sustainability scenario: For this scenario, we assume a Domestic Retail Margin lower than in the Base case scenario. Moreover, we assume Domestic Revenues do not change with respect to our forecasts, while Outbound and Inbound volumes increase by 10% each year with respect to our forecasts. All these elements represent a threat for sustainability, and for this reason we think this scenario is likely to produce the lowest sustainability.

High Sustainability scenario: For this scenario, we assume a Domestic Retail Margin higher than in the Base case scenario. Moreover, Domestic Revenues increase by 10% per year with respect to our forecasts, while Outbound and Inbound volumes decrease by 10% each year with respect to our forecasts. Sustainability should be enhanced by this combination of elements, and for this reason we think this scenario is likely to produce the highest sustainability.

COVID-19 sensitivity scenarios

The present section examines the sensitivity of the sustainability analysis in view of different scenarios on the impact of the COVID-19 pandemic on the roaming market.

For this purpose, we have developed some additional sensitivity scenarios, which seek to examine how a prolonged impact of the COVID-19 pandemic on international travelling could influence the ability of roaming providers to offer RLAH services in a sustainable manner.

In particular, the scenarios are based on the estimation of the drop in touristic flows in 2020. Eurostat data 23 suggest that, on average, while travelling in Europe, in 2020 European citizens have spent almost 45% of the total nights spent in the previous year. Consequently, we assume that roaming volumes could have decreased – in 2020 – by almost 50%.

On the basis of this assumption, we then speculate roaming volumes recovery speeds through which the gap with the forecasted figures could be filled. The main assumptions of the COVID-19 sensitivity scenarios are presented hereby and summarized in table:

COVID-19 Minimum Impact scenario: We assume that roaming volumes in 2023 will have partially recovered with full recover in 2025. In particular, they will be 10% lower than those expected by the forecasts produced. The domestic market will not be impacted by the crisis and therefore domestic revenues as well as the domestic margin will be as for the baseline scenario.

COVID-19 Low Impact scenario: We assume that roaming volumes will not fully recover by 2025. In particular, they will be 20% and 10% lower than expected in 2023 and in 2025 respectively. As before, we also conjecture that the domestic market will not be influenced and figures will remain at the baseline level.

COVID-19 Medium Impact scenario: We assume here that the recovery in roaming volumes will be slower, and that roaming traffic will be 30% and 20% lower than expected in 2023 and 2025 respectively. Moreover, here we let the COVID-19 crisis to have an impact on the domestic market too. In particular, we speculate that due to the bankruptcy wave caused by the crisis – especially hitting SMEs – domestic revenues could decrease being 10% lower than those expected while the domestic margin will remain unchanged.

COVID-19 High Impact scenario: In this final scenario, the recovery of roaming volumes will be even slower and we assume the roaming traffic to be 40% and 30% lower than expected in 2023 and 2025 respectively. As before, we speculate a high impact of the crisis on the domestic market: besides domestic revenues being 20% lower than expected, we also set the domestic margin at 20% instead of 30%.

Table 8 COVID-19 sensitivity scenarios

Results

The following tables present the sustainability results for the baseline scenario (first row in each option) and for the high and low sustainability scenario (range of second row in each option).

Table 9: Number and % of operators with sustainability < -3% for each policy option and sensitivity scenarios. 2023 estimates

Table 10: Number and % of operators with sustainability < -3% for each policy option and sensitivity scenarios. 2025 estimates.

Table 11 COVID-19 sensitivity results - 2023

Table 12 COVID-19 sensitivity results - 2025

According to the tables on the COVID-19 sensitivity, we conclude that a prolonged impact of the COVID-19 pandemic would ameliorate the sustainability challenges but in most cases without surpassing the high sustainability scenario. Therefore, the COVID-19 sensitivity analysis confirms the choice of not considering the COVID-19 impact in the baseline forecasts.

Annex 4B - Counterfactual evaluation of the impact of RLAH and assessment of benefits for consumers

Introduction

The aim of this section is to quantify how much European consumers have benefited from the implementation of the RLAH.

RLAH brought about huge benefits for consumers. After the introduction on the new roaming rules consumers started to pay less (the domestic price) for something (roaming volumes when traveling abroad) which - in principle - they would have been willing to pay more for. This `extra' benefit is what in Economics is called `Consumer Surplus'.

In order to measure how Consumer Surplus has changed since the RLAH implementation, we first need to estimate the impact of RLAH on roaming volumes which is only imputable to the enforcement of the new rules, excluding all other concurring factors. In other words, what this section is looking for is the causal impact of the RLAH on roaming volumes, upon which the change in Consumer Surplus can be computed. Ultimately, the result will give a precise idea of how much European consumers have gained uniquely from the implementation of the regulation.

The causal impact of a policy is retrieved adopting Counterfactual Impact Evaluation (CIE) methods. CIE replies to the fundamental question “What would have happened if the Regulation were not implemented?” The reply is based on the comparison of the outcomes of two groups. One is composed of treated individuals. The other is formed by all individuals who are totally comparable - under a list of characteristics - to the treated ones but for having not benefited from the policy. It is called control group and serves to mimic what would have happened to the outcome in the absence of the Regulation.

As previously mentioned, in this analysis CIE - namely a Difference-in-differences model (DiD hereafter) is used to derive the causal impact the implementation of RLAH had on a number of outcomes, with a specific focus on roaming volumes. In particular, it will derive the change in volumes attributable solely to the regulation, i.e., through CIE it will be possible to distinguish the effect of the regulation on volumes from all other factors that could have influenced the roaming volumes over time (e.g. technological developments, consumption habits, specific tourism dynamics, etc). Finally, these causal changes are used to compute the corresponding change in Consumer Surplus brought about by the new roaming rules.

Data description

The analysis presented in the following has been made possible by the availability of detailed data on the European roaming market collected by (BEREC. who quarterly sends to NRA a survey through which data on pricing and consumption patterns of both domestic and roaming services - being the latter generated either within the EEA or in the Rest of the World - are asked. The survey also gathers data related to domestic and roaming revenues. Finally, on the basis of the survey results, BEREC publishes the aggregated information on the evolution of roaming traffic on a regular bases in the form of the International Roaming Benchmark Data Reports. For the sake of this analysis, access to the 19th, 20th, 21st, 22nd and 23rd of the International Roaming Benchmark Data Reports has been granted. In particular are used, information on wholesale and retail revenues and volumes for voice, data and SMS services in both domestic and foreign markets (disaggregated for intra EU and Rest of the World countries) for each quarter between q4 2016 and q1 2019 as they were declared by mobile operators of all 28 EU Member States (either MNOs and MVNOs) to their NRAs and then reported to BEREC.

Empirical strategy for the evaluation of the RLAH

The evaluation of the RLAH is made adopting the so called DiD method. The rationale for the double difference resides in the possibility of overcoming several problems related to comparing outcomes either only between groups or only overtime.

The first difference of the DiD method, that is, the before/after difference in outcomes for the treated group, takes charge of all the factors that are constant over time -or at least in the time interval of the analysis - in this group (it is the case, for example, of the structure of the national market, the level of competition).

However, if one were to use this difference to evaluate the impact of RLAH, one would end up biasing our assessment because many other time-varying factors could have contributed to the dynamics of roaming volumes, like for example seasonality, macroeconomic fluctuations, network infrastructure development or increased availability of smart phones.

In order to capture these time varying factors one can calculate the same type of difference for a group that is perfectly comparable to the treated one. In particular, the control group should be comparable specifically for what concerns the time-varying factors influencing the outcome. Finally, the impact of the policy can be obtained by subtracting the second difference from the first one.

To understand if the control group has been appropriately chosen - and to ensure the validity of the methodological choice - it is essential to verify that the outcome between the two groups was similar before the policy intervention. Indeed, if trends were already diverging before the intervention, then it would mean that the time-varying factors influencing the outcome in the treated group were not the same as the ones influencing the controls. This could be the case, for example, if the treated group was influenced by network-technological developments while the control group was not: in this case, we would observe a difference in outcome-trend even before the implementation of the policy, because the technological improvements were already pushing roaming volumes in the treated group. Under these conditions, it is impossible to impute the difference observed ex-post to the implementation of the policy itself.

The RLAH came into force all over Europe at the same time, on 15 June 2017. As for the control group, the natural candidates to be part of it would have been all countries not belonging to the EU (e.g. considering OECD countries), but there is no operators' level information available for them as detailed and complete as the BEREC's one. 24 Therefore the lack of comparable data makes it impossible to use non-European countries as the control group.

The BEREC database collects very detailed data on roaming volumes and revenues from the majority of operators all over Europe. On top of information about European roaming, i.e. the roaming traffic generated among the EEA, the database contains data regarding the roaming traffic produced outside the EEA, that is, the traffic produced by European citizens while traveling outside the EEA.

This type of traffic is produced by the same users consuming roaming volumes in the EU, since each operator reports to BEREC the roaming traffic imputable to its subscribers generated either in the EEA or outside of it. Moreover, being produced by the same devices through which the roaming traffic is generated, it is subjected to almost identical technological improvements and conditions (i.e. time-varying factors) to which the EU one is exposed. Nonetheless, the roaming regulation does not apply to it. Hence, the extra-EEA roaming traffic could be used as the proper `control group' to evaluate the RLAH.

Therefore, to identify the causal parameters of interest we implement a DiD model where the treated units are operators' retail volumes as generated by European customers while traveling in Europe. The control units are those accumulated by European citizens while travelling in extra-EU countries (e.g. outbound extra-EU roaming traffic). As already mentioned, volumes in the control group could not be influenced by the RLAH Regulation because only the traffic produced by European people in Europe is affected by the new rules.

In formal terms, the equation we estimate is:

(1)

where is the outcome of interest (roaming volumes) for mobile operator i in area a – (a can be EU or extraEU) - at time t. PostRLAH is a dummy variable that takes value 1 only in the periods after RLAH implementation, while EU is a dummy variable valued 1 only if the area where the traffic has been generated is the EU. is the coefficient of interest and it is related to the variable which takes value 1 if the outcome is measured in a post RLAH period and the area is EU.

From a formal point of view, the dummy variable EU captures all the possible differences between the treated group and the control group that do not change over time. The variable PostRLAH accounts for all the aggregate factors that would have influenced roaming volumes even in the absence of the RLAH, such as the growing number of smart-phones, the increasing production of new apps for mobile phones, the network improvements, macroeconomic shocks and fluctuations in traveling habits. All these factors impacted in the same way the EU roaming as well as the non-EU roaming.

We increase the equation with operator-area specific dummy variables ( ) that help us to take into account the fact that operators may implement different marketing/productive strategies according to the reference area - either EU or non-EU - which eventually impacts on roaming volumes. This could be the case of differentiated tariffs, for example. Moreover, we make use of time dummy variables that help us remove events happening in specific years and that could have influenced both types of roaming volumes. This is the case, for example, of macroeconomic dynamics: an economic boom taking place in a particular year will most probably lead to higher roaming volumes because users travel more, both in the EU and outside of it.

therefore measures the causal impact of the RLAH regulation. It informs by how much the roaming volumes produced within the EEA have increased with respect to the pre-RLAH period and with respect to what has happened - in the same time span - to the volumes generated outside of it. In particular, the change measured is clean from any other factor that could have influenced roaming volumes.

In order to verify the existence of similar trends in the outcome between the treated and the control group before the implementation of RLAH, a methodology that it is commonly used in the literature and that has been pioneered by the work of David H. Autor 25 is implemented. Basically, we assess whether statistically significant differences in the outcome variable existed in each quarter before the implementation of the RLAH. If no difference is detected, than it means that the two series - EU roaming VS Rest of the World one - were following the same dynamics.

Consumer Surplus

Consumer surplus is the difference between the consumers’ total willingness to pay for roaming services and the amount they actually pay. The RLAH regulation reduces the price for roaming services to the level of domestic services, and this changes consumer surplus in two ways. First, consumers pay less for the volume of roaming services they previously consumed (i.e. before the RLAH regulation). Second, consumers increase their demand for roaming services, which provides an additional benefits.

Figure 1 offers a graphical representation of these two effects, which will both be incorporated in the quantification of consumer benefits. The figure shows the price P for roaming services on the vertical axis, and the consumed volume Q of the services on the horizontal axis, and it plots a downward sloping (linear) demand curve for roaming services.

Before the RLAH regulation, consumers paid a price for roaming equal to P1 and correspondingly consumed a volume of Q1. Consumer surplus was the yellow triangle ABF, i.e. the difference between the total willingness to pay (area ABDE) and the actual expenditures P1 (FBDE). After the RLAH regulation, the price drops to the domestic price P2, so that demand increases to Q2. Consumer surplus has increased to the larger triangle ACE.

The change in consumer surplus because of the price drop is therefore the red area, FBCE, and this indeed consists of two parts. First, consumers gain because they experience a reduced price on the volume consumed before the RLAH, Q1. This reduction in expenditures is the area FBDE, and it can be easily calculated without a causal effects analysis. Second, consumers gain because they now consume a higher amount of roaming volumes (increase from Q1 to Q2). This gain is given by the triangle BCD, and is calculated from the causal effects analysis.

Figure 1: Consumer Surplus graphical representation

More formally, based on the linear approximation of demand, one can show that the change in consumer surplus is the sum of consumer expenditures before the regulation, plus half of this amount multiplied by the percentage increase in demand caused by the RLH regulation, that is, the area of the FBCE polygon. The latter - for each roaming service k, i.e. voice, data and SMS - can be computed summing up the area of the FBDE rectangle and the area of the BCD triangle. In mathematical terms:

(2)

Even though the roaming price is not observed, it can be derived dividing revenues by volumes . Consequently, Equation 2 becomes:

(3)

The term represents the percentage change in roaming volumes between the pre RLAH implementation and the post implementation period. The latter term can be estimated through the coefficient in Eq. 1, therefore obtaining the causal change in Consumer Surplus, that is, the variation in the CS exclusively imputable to the enforcement of the RLAH regulation.

Results

This section presents the results about the counterfactual analysis and the Consumer Surplus calculation which have been estimated referring to the period 2016 Q4 up to 2019 Q1. Specifically, the computation of the Consumer Surplus has been implemented on an annual basis, and the results have to read as the change in CS only for the year after the RLAH enforcement.

Table 13 displays the counterfactual estimated changes in voice and data roaming volumes.  26

Table 13: Impact of RLAH on roaming volumes [3]

Note: Robust standard errors in parentheses. Errors clustered at the country and operators’ level.

***p<0.01, **p<0.05, *p<0.1

Specifically, it says that because of RLAH, voice and data volumes have increased by 110% and 267,2% respectively 27 . The numbers here represent the change in volumes that is uniquely imputable to the implementation of RLAH, excluding all other concurrent and confounding factors.

Indeed, if we were to calculate the simple change in EU roaming volumes before and after the new roaming rules, controlling for no factors among the ones listed previously (i.e. seasonality, technological change, macroeconomic fluctuations, etc) we would end up with a percentage change in voice volumes of 112,7% and a percentage change in data volumes of 489,2%.

The comparison between the counterfactual and the simple before/after percentages is very informative. As for voice, it tells us that the change in volumes that we observe after the RLAH is almost entirely due to the new roaming rules. The voice market seems to be a consolidated one, where technological changes or other factors play little role. The one thing that made consumers change their consumption habits was indeed RLAH.

The data market looks pretty different. The counterfactual estimates accounts for around 50% of the before/after scenario, suggesting that RLAH had a significant effect in increasing data volumes whilst also other factors contributed.

Confronting the results on data volumes is useful to understand what a counterfactual estimation does. If we had evaluated the impact of RLAH solely on the basis of the before/after comparison, we would have ended up overestimating the impact of the regulation because we would have given the same weight to all the factors that influence the roaming volumes dynamics. On the contrary, having found a proper control group, we are able to disentangle how much of the observed increase can be credited to the regulation.

Finally, Table 14 presents the estimates related to the Consumer Surplus change.

Table 14: Change in Consumer Surplus

The quantification proposed above informs that the Consumer Surplus of European consumers increased by 2.4 Billion Euro because of RLAH. The growth has been particularly strong in the data market, given the unprecedented high data volumes consumed in the market and the fact that the price charged on roaming data (i.e. for 1 GB used) was very high in most European countries.

The calculation refers to the period 15th June 2017 – 15th June 2018. However, it is possible to roughly extrapolate the Consumer Surplus change for the two-year period after RLAH, by exploiting the relationship between the volumes produced in the second year (Vol2) and those generated in the first year (Vol1).

Basically, it is assumed that:

Upon this hypothesis, the total Consumer Surplus change is quantified in the range of 5.4 Billion Euro.

Evaluating waterbed effect in the domestic market

The quantification of the Consumer Surplus we have presented so far is valid as long as we assume that RLAH did not affect the domestic market, and in particular, that the price for domestic services has not increased because of RLAH (what is called “waterbed effect”). Indeed, it could have been that mobile operators, in order to counteract the loss in revenues suffered from not being able to apply high surcharges to roaming traffic, decided to increase prices at the domestic level. In this case, consumers would have been confronted with a significant reduction of prices in the roaming market and at the same time an increase in tariffs for their domestic consumption. Overall, their change in surplus – which we have previously computed – could have been negatively influenced.

To evaluate the assumption of no waterbed effects, we rely upon quarterly data on domestic mobile retail prices for representative consumption baskets[1] collected by Teligen. The data cover a total of 36 OECD countries, among which 24 are EEA countries. The data refer to the period 2016 Q3 – 2017 Q4.

The empirical strategy we implement is similar to the one adopted before. In this case, the DiD model is built such that the treated group is represented by the 24 EEA countries, whilst the remaining 12 non-EEA countries form the control group.

The outcome variables are the prices of the 2006 consumption baskets defined by the OECD for high, medium and low voice volume users.

In order to be sure to capture only the impact of the roaming regulation, and to get rid of any other confounding factor, in the estimation we take into account specific characteristics that do not vary over time at the country-basket level (i.e. consumption patterns of users in a country with respect to patterns of users from other countries), as well as for specificity at the basket-quarter level.

The resulting coefficient of interest informs about the impact of the regulation on domestic prices in the EEA with respect to the rest of the world. In particular, a not statistically significant coefficient would mean that domestic prices in the EEA have not increased nor decrease significantly with respect to prices in the rest of the world, suggesting that no waterbed effects has materialized[2].

Table 15: Impact of RLAH on domestic prices

Note: Robust standard errors in parentheses. Errors clustered at the country-basket level.

***p<0.01, **p<0.05, *p<0.1

The results of the analysis presented in Table 15 supports the hypothesis of a null waterbed effect. The domestic prices of the three consumption baskets show small but insignificant increases in the EEA countries after the regulation, relative to the other OECD countries.

Overall, these findings suggest that the regulation did not imply statistically significant waterbed effects on the domestic market. One possible explanation is that domestic and international roaming services are only very weak substitutes or complements, so that operators have no incentives to adjust their domestic prices. Another explanation is that the roaming market is only a relatively small part of the overall business of the mobile operators, so that any adjustments in price strategies would be small and difficult to detect.

[1] Consumption baskets are built following OECD guidelines. See for more details: OECD (2017), “Revised OECD Telecommunication Price Baskets”. Working paper, OECD, Directorate for Science, Technology and Innovation Committee on Digital Economy Policy.

[2] The definition of baskets has changed since 2018 making price data from 2018 on not comparable with those before. This is the reason why our analysis covers only the time period up to 2017 Q4. Nonetheless, we believe that there is no reason to believe that operators would have waited to increase domestic prices, had they planned to do so once the date of implementation of the regulation was known.

Annex 4C Cost model for wholesale roaming costs

The cost model used for estimating wholesale roaming costs

In order to assess the costs of providing wholesale roaming services in the 31 EU/EEA countries for the purposes of the roaming review, an independent study was commissioned 28 . The study’s dual objective was to assess the cost of providing both wholesale roaming services and voice call termination rates. For the purposes of this SWD, the main focus of this summary of the study is on the estimated costs related to roaming, whilst taking into account the outcomes of the estimated costs of providing wholesale voice termination services.

The cost study was conducted by Axon Partners Group from mid-March 2018 to mid-July 2019. At the start of the study, a first workshop was organised with all relevant stakeholders on 10 April 2018 in order to collect feedback on the methodology proposed by Axon. A comprehensive data gathering aimed at the European mobile operators was then conducted in June-July 2018, via the NRAs, in order to obtain, from mobile operators, the relevant information and inputs needed to populate a country-specific model for each Member State. The aim was to build 31 models with a similar skeleton, based on country-specific input, facilitating as best possible the estimation of the relevant mobile wholesale costs in each of the 31 Member States. 29 The models would rely on country-specific input where relevant and, where not, on averages/common values across the EU/EEA. On 29 October 2018 the first draft cost model was shared with stakeholders for consultation, followed by a second version of the draft model shared on 15 February 2019 for a second round of consultation. The relevant comments and suggestions received during both consultations were implemented in the draft final model, which was presented to stakeholders at a second workshop on 28 May 2019. The cost model was then finalised and published on 24 July 2019.

To ensure transparency throughout the project, several steps were taken to associate the NRAs, operators and other stakeholders. First, two workshops were held and two rounds of consultation were organised over the period going from October 2018 to March 2019. 30 Stakeholders were also consulted on the structure and content of the data gathering exercise of June-July 2018 itself. Also, a steering committee composed by experts from NRAs was established and regular meetings between the Commission, Axon and the Steering Committee were held throughout the project. The steering committee consisted of representatives from 8 NRAs and was composed of members of the two BEREC Expert Working Groups dealing with roaming and with termination rates.

The final cost model estimates the costs of providing wholesale mobile roaming services as well as voice termination in 28 EU/EEA countries.

A cost model enables the user to evaluate costs of all elements required to offer mobile services. For the purposes of the roaming review, the relevant costs to be estimated are those related to wholesale roaming services, specifically the estimated costs for:

·Data roaming

·Outgoing voice calls

·Outgoing SMS

The Axon cost model allocates costs related to wholesale roaming services following the so-called Long-Run Incremental Cost (LRIC) plus some allocation of joint and common costs (LRIC+). This cost standard can be divided in two parts, essentially the LRIC and the plus (+). LRIC is a way to derive the cost of producing an additional increment of a given output, when assessed over a long period of time (long-run). In economic theory, the long-run means that all inputs relevant to the production of the output are considered variable. The LRIC+ cost standard allows for including joint and common costs which are relevant for other services as well. Accordingly, costs estimated under the LRIC+ cost standard are higher than costs derived under LRIC. However, as these shared elements are also necessary for the relevant service, and in line with the regulatory obligation for the wholesale roaming cap to cover such costs 31 , the Axon Cost model deploys the LRIC+ standard for all services related to roaming. This approach ensures that for shared equipment needed for e.g. data and voice the costs are captured in the estimation. As companies need to recover joint and common costs to ensure long-term sustainability, joint and common costs are shared among the services that generate them and accordingly recovered by any price cap set above the estimated costs for those services.

In contrast, incoming voice (voice termination) is calculated purely on the basis of the LRIC cost standard (pure LRIC), in accordance with the Commission's 2009 Recommendation on Termination Rates 32 , which recommends the estimation of termination rates based on a bottom-up pure LRIC approach. For incoming SMS, the Axon cost model follows the approach adopted in the previous cost model 33 , where no costs are allocated to termination of incoming roaming SMS. 34 To ensure cost recovery for these services, the cost of incoming roaming SMS is re-allocated to outgoing roaming SMS.

The cost model takes into account a wide selection of relevant parameters including radii (coverage) of the mobile sites, the different geo-types in each Member-State, whether or not a Member State exerts seasonal consumption-spikes and many other elements. As an example, seasonality is taken into account for Croatia, France, Greece, Malta and Spain who were able during the data collection exercise to prove that seasonality had an effect on the dimensioning of their networks. Essentially, any network must support the peak demand and the seasonality assessment determines whether traffic is distributed evenly across months or whether it peaks at specific months (e.g. summer period or winter period). Therefore, seasonality is assessed on a country-by-country basis to ensure that the relevant peak-time of the country is considered. For a full description of seasonality and the other elements considered, please see the full set of published materials.

Relevant results of the cost model

The Axon cost model estimates network costs incurred by an efficient operator. Accordingly, any additional (non-network) costs incurred by the visited network when providing wholesale roaming services must be considered in order to ensure full cost recovery. In order to apply the estimated costs in the exercise of setting adequate caps for roaming wholesale services, costs for transit (data and voice) and termination (voice) must be added to ensure cost recovery for providing these services. The results presented in this section therefore includes these additions.

For roaming data services, transit rates must be added to the estimated network costs. For voice roaming, transit and call termination rates must also be taken into account. This is because the visited network is paying the network operator where the call placed by the roaming customer will terminate. For example, a Spanish customer visiting Germany makes a call back to Spain. To complete the call, the German (visited) operator must first originate the call on the German operator’s network and then transit the call through a number of countries back to Spain where the call is finally terminated at the receiver. The German operator must cover these transit and termination costs, therefore these costs must be considered to ensure cost-recovery by the German (visited) operator.

As roaming SMS are transited without extra costs incurred by the visited network, for the purpose of comparison, no further costs needs to be added to these estimates. For this purpose and to illustrate a more complete estimate of the cost of providing roaming voice and data services, this section describes these results of the model including transit and call termination costs.

The Commission services acknowledge the need to consider call termination and transit costs when assessing the appropriate level of any potential wholesale roaming price cap. Transit costs added here are estimates performed outside of the Axon cost model and not subject to the same modelling exercise. The estimated transit costs used are based on a data collection performed jointly by BEREC and the Commission, where operators provided transit costs incurred in year 2018. These calculations are subject to some uncertainty, as not all operators (or even countries) were able to reply to this specific data collection. However, from the reported data available to the Commission services, the following average transit costs have been estimated:

·Transit price for roaming data service, 2018:    0.20 €/GB

·Transit price for roaming voice service, 2018:    0.0050 €/minute

As these estimates are only available for year 2018, these costs are applied to each year under evaluation here. The above transit costs were consulted with operators during the first consultation round in November 2018, where 75% of NRAs and 48% of MNOs agreed with these estimates. For MNOs it must be noted, that although less than half agreed with the estimates, those who disagreed had contradictory views and considered the estimate to be either too high or too low 35 .

Adding the transit costs to the results of the Axon cost model, the estimated costs are illustrated in Figure 2 below.

Figure 2: Estimated costs for providing data roaming services including transit

Source: Axon Consulting, SMART 2017/0091 and Commission service estimates

The cost model estimates that costs are declining across all countries, with estimates between 0.7 and 4 €/GB in 2017 and converging downwards to a range of 0.7 to 1.6 €/GB in 2025. The convergence is most apparent for Spain and Malta where the largest decreases are observed, followed by Belgium and Hungary. A number of countries, e.g. in Poland, have very limited developments in costs throughout the period. For 2025, the highest estimated costs for roaming data services are found in Hungary.

For voice roaming services, in addition to transit, termination rates shall also be a part of costs to be recovered by the host operator, meaning that one must add the costs of terminating the call in another European network. For termination rates, these will under a Delegated act be set by the European Commission, from 2021 and until 2025 in accordance with article 75 of the Code 36 . The delegated act proposes the following maximum mobile termination rates, which are added to the costs of the Axon cost model for wholesale roaming voice services:

2022: 0.0055 €/min

2023: 0.004 €/min

2024: 0.002 €/min

2025: 0.002 €/min

Adding the termination rates and transit to the costs estimated in the Axon cost model, the costs are derived as shown in Figure 3 below.

Figure 3: Estimated costs for providing voice roaming services, including transit and termination

Source: Axon Consulting, SMART 2017/0091, Commission service estimates and input from the Delegated act setting Union-wide termination rates.

In 2017, voice roaming costs are estimated between 1.4 and 3.3 €cents/minute, when including costs of transit and termination. Poland displays the lowest costs whereas Belgium and Malta are the two highest cost countries in 2017. The estimated costs for 2025 converge to a range of 0.7 to 1.2 €cents/minute for almost all countries, with Malta being the only country remaining above 1.2 €cents per minute throughout the period, with 1.84 €cents/minute estimated for 2025. Looking at the other high-cost country in 2017, Belgium, convergence with the other Member States is observed already in 2018 and continues downwards until 2025. For Malta, these high estimates are driven by the thickness of walls in Maltese buildings, requiring operators to build and maintain a comparatively larger number of sites than elsewhere seen 37 .

As roaming SMS are transited without extra costs incurred by the visited network, no further costs needs to be added to these estimates and the estimates can be evaluated directly from the Axon cost model, as seen in Figure 4 .

Figure 4: Estimated costs for providing SMS roaming services, including transit and termination

Source: Axon Consulting, SMART 2017/0091

For roaming wholesale SMS, the estimated costs are presented in Figure 4 showing a more constant level across countries than for roaming voice and data. Unlike voice and data, the cost estimates for roaming SMS derived from the cost model are not subject to additional costs, such as termination or transit, due to the construction of the SMS wholesale regime. This follows from the roaming regulation, attributing any termination costs for incoming SMS to the equivalent outgoing service to ensure cost recovery.  38 Incoming roaming SMS are not charged at neither retail nor wholesale level, which means that the costs generated are reallocated to roaming SMS outgoing (origination of roaming-SMS). This approach towards roaming SMS considered in the cost model is consistent with the previous approach taken in the TERA Consultants cost model. 39  

As shown in Figure 29, the cost model estimates unit costs for roaming SMS between 0.05 and 0.4 €cents/SMS in 2017, i.e. a factor 8 between the lowest and highest costs. A couple of countries display some downwards convergence, but not to the extent seen for roaming voice and data. By 2025, the costs are estimated between 0.05 and 0.3 €cents/SMS, indicating a six-fold difference between high- and low- cost countries. 40

Proposed caps resulting from the cost model

From the results derived from the Axon model, a proposed cap for each relevant roaming wholesale service is needed for the impact assessment. For this purpose, the Commission takes into account the applicable caps in 2022, costs derived from the Axon cost model (including transit and termination), interest of ensuring cost recovery and current wholesale prices charged by operators. Option 3 of the impact assessment proposes a reduction of the applicable wholesale caps and this section outlines the proposed caps for that option.

Wholesale roaming data services

Until 1 July 2022, applicable glide path rate defining the maximum wholesale rate for data service stands at 2.5 EUR/GB. The highest estimated cost from the Axon cost model including transit in 2022 is 1.74 EUR/GB and average cost estimated in the model standing at approximately 1 EUR/GB. According to the latest available BEREC benchmark report (25th edition), the average wholesale cost per GB paid in Q1 2020 is 1.53 EUR/GB. In 2025, the Axon cost model estimates a highest cost of 1.63 and average cost around 0.9 EUR/GB.

Considering the above observations, the Commission proposes a continuation of the decreasing cap observed since 2017. Specifically, the Commission proposes to set a two-step glide path accommodating both the decreasing costs observed in the cost model whilst gradually reducing the caps to minimize disruptions for the operators. Therefore, from 1 July 2022, the Commission proposes a cap of 2 EUR/GB for data services, decreasing in 1 January 2025 to 1.5 EUR/GB. These values will be the foundation for the sustainability analysis performed in the impact assessment under option 3.

The cap of 2 EUR/GB proposed in 2022 is slightly above the maximum efficient cost of 1.74 EUR/GB estimated for the same year. This cap is proposed to balance the transition to the cost proposed in 2025 of 1.5 EUR/GB and ensure that the operators have sufficient time to negotiate wholesale agreements reflecting the decreasing caps.

To this regard, the Commission acknowledges that the proposed cap of 1.5 EUR/GB for 2025 is slightly below the observed highest estimated costs (including transit of 0.2 EUR/GB) in two Member States, namely Hungary (1.63 EUR/GB) and Belgium (1.58 EUR/GB). Current actual charged wholesale costs (25th BEREC report, Q1 2020) in Hungary and Belgium reveals charges of 1.37 EUR/GB and 1.56 EUR/GB respectively. This indicated to the Commission, that a cap of 1.5 EUR/GB in should also ensure cost recovery in these two Member States, especially taking into account the efficiency gains and accordingly decreasing costs observed in the past. As the cap of 1.5 EUR/GB proposed for 2025 is either above (Hungary) or very close (Belgium) to the actual cost charged in 2020, cost recovery should also be ensured in these Member States.

Further, the Commission has projected the traffic for data, voice and SMS in the two Member States and analysed the total impact of the proposed caps for option 3 (including for voice and SMS as described below), finding that a total cost recovery will be ensured under the full proposal.

Therefore, option 3 propose for data services the following two-step glide path:

·From 1 July 2022:        2 EUR/GB

·From 1 January 2025:    1.5 EUR/GB

Wholesale roaming voice services

In the current roaming regulation, voice services have since 15 June 2017 seen an applicable cap of 0.032 EUR/min. The Axon cost model estimates a decreasing cost from 2022-2025, enhanced when accounting for the outcome of the Delegated act setting Union wide termination 41 . The highest estimated cost decrease from 0.022 EUR/min in 2022 to 0.0184 EUR/min in 2025. In 2022, the average cost estimated for wholesale roaming voice service is 0.0135 EUR/min, decreasing to 0.0096 EUR/min in 2025.

From the 25th BEREC report on roaming, the average wholesale price per minute was 0.0166 in Q1 2020. The Commission is therefore certain, that the costs as estimated from the Axon cost model, including transit and termination, will ensure cost recovery for operators providing roaming wholesale voice services.

The cap for roaming voice services has remained constant since 2017 and in the current proposal the Commission proposes under option 3 a two-step decreasing glide path, based on both the decreasing costs from the Axon model and the decreasing termination rates resulting from the delegated act. As caps will apply from 1 July 2022, the proposed two-step glide path follows the same dates as observed for data caps above. As such, the Commission proposes the steps to follow the cost-decreases seen from the Axon model and accordingly proposed the caps seen below:

·From 1 July 2022:        0.022 EUR/min

·From 1 January 2025:    0.019 EUR/min

Wholesale roaming SMS services

The wholesale cap setting the maximum charge for one SMS while roaming has been constant since the introduction of RLAH, at 0.01 EUR/SMS as of 15 July 2017. The Axon cost model estimates a fairly constant cost for delivering this service in the modelled period of just under 0.003 EUR/SMS from 2022-2025. The average estimated cost for an SMS in the relevant period is 0.0011 EUR/SMS. No additional costs for transit and/or termination is needed when setting the cap for this service. From the BEREC report, an average price of 0.0021 EUR/SMS is presented for Q1 2020. For reference, in 2018 the average price paid for a SMS was 0.0031 EUR/SMS.

The efficient unit cost for an SMS is by the Axon cost model estimated to just under 0.003 EUR/SMS. Taking the current cap of 0.01 EUR/SMS and average prices paid just over 0.0031 EUR/SMS as well as the efficient cost, a two-step reduction of the SMS cap is proposed, to mitigate the transition to the efficient cost.. As such, option 3 involves the below cap for wholesale roaming SMS services:

·From 1 July 2022:        0.004 EUR/SMS

·From 1 January 2025:        0.003 EUR/SMS

Annex 5: Baseline

Baseline A. Sustainable provision of RLAH

A1. Sustainability of RLAH and cost recovery at wholesale level preserving incentives to invest in visited networks and avoiding distortion of domestic competition in visited markets

Measures in place

The Roaming Regulation provides a number of measures to enable providing of RLAH in a sustainable manner throughout the Union:

At wholesale level, where operators have a wholesale access obligation to ensure provision of roaming services, in order to make wholesale costs sustainable in a “Rome-Like-At-Home” regime, there has been a substantial reduction of wholesale roaming price caps applicable to wholesale agreements between operators, while ensuring that operators providing the wholesale service can recover their costs.

Specifically, Regulation (EU) 2017/920 set price caps at 3.2 €c/min for voice (reduced by 36%), at 1 €c/SMS (reduced by 50%) and at 7.7 €/GB for data (reduced by 85%) It also defined a glide path for further reducing price caps for data, with a last step at 2.5 €/GB in 2022. These wholesale roaming price caps ensured that wholesale costs could be fully recovered by the operator providing the wholesale roaming service.

At retail level, where operators have the obligation to provide roaming services at the same conditions as domestically for periodic travelling, they have the possibility to apply:

(a) a fair use policy to prevent abusive or anomalous use of roaming services at domestic prices (such as permanent roaming); and (b) exceptional and temporary derogations to forestall any risk of domestic price increases.

The Commission Implementing Regulation (EU) 2016/2286 (CIR) stipulates detailed rules a) on the application of fair use policy and b) on the methodology for assessing the sustainability of the abolition of retail roaming surcharges and the application to be submitted by a roaming provider for the purposes of that assessment.

Baseline data

The report on the roaming review of 29 November 2019 concludes that both safeguard rules at retail level (fair use policy and sustainability derogation) have worked adequately. Therefore, the Commission does not intend to amend the rules laid down in the Implementing Regulation (EU) 2016/2286.

According to data collected by BEREC for the International Roaming Benchmark Report, the use of fair use policies has been stable in general, not exceeding 4% of total roaming traffic for voice and 6% for data.

At the same time, the number of sustainability derogations exhibit a broadly declining trend. Furthermore, as the 2019 review report concludes, operators that have obtained sustainability derogation have been using it in general with parsimony. As shown in the 2019 SWD, voice and data traffic subject to derogation in the EU does not exceed in average 3% and 1.5% of total roaming traffic respectively. Furthermore, the only country, where voice or data traffic subject to derogation exceeds 12% of total roaming traffic is Lithuania.

The following table illustrates the number of derogations granted per member state in the period 31 August 2018 to 31 August 2019 42

Table 16: Sustainability derogations granted 31 August 2018 to 31 August 2019

While wholesale caps have remained stable for both voice and SMS, actual wholesale rates paid have declined moderately for voice but more substantially for SMS. Specifically, the average wholesale rate for voice calls has reduced between Q3 2017 and Q3 2019 by 14% (from 0.022 to 0.0189 €/min). In the same period, the average wholesale price for SMS has reduced by 54% (from 0.52 to 0.24 €/SMS). Similarly to SMS, the average wholesale price for data has reduced between Q3 2017 and Q3 2019 by 56% (from 3.6 €/GB to 1.59 €/GB). This reduction has been sharper than the corresponding reduction in wholesale caps (42%, from 7.7 €/GB to 4.5 €/GB). However, between Q1 2019 and Q1 2020, we observe an increase in average wholesale rate for voice (3.5%) and SMS (11%). For data we observe a reduction which is however lower than the reduction in the wholesale caps in the same interval (13% versus 22%).

Two factors mainly determine this decline. Firstly, new (and, for data, annually decreasing) maximum wholesale roaming prices laid down in the Roaming Regulation have acted as much lower ceilings on prices, triggering competitive market dynamics between operators offering wholesale roaming access below those ceilings. Secondly, the introduction of RLAH has resulted in significant increases in roaming volumes, thereby fuelling further competition in wholesale roaming prices.

The 2019 review report concludes that, while there is some evidence of economic space between the wholesale price caps currently programmed until 2022 and the level of costs of all operators, the case for further reductions in order for the RLAH regime to function better while maintaining domestic competitive dynamics needs to be further analysed. To reach this conclusion, the review has taken into account the findings of the external study, commissioned for the needs of the roaming review 43 , and the BEREC Supplementary cost analysis 44 , published on 20 September 2019.