Submission by the Commissioner of Competition before the Canadian Radio-television and Telecommunications Commission—Telecom Notice of Consultation CRTC 2014-76—Review of wholesale mobile wireless services

August 20, 2014


I. Introduction

  1. This is the second intervention of the Commissioner of Competition (the "Commissioner") in response to Telecom Notice of Consultation CRTC 2014‑76.
  2. As set out in the Commissioner's first intervention, when spectrum access is limited and network costs are high, there may be low incentives for incumbent mobile wireless carriers ("incumbents") to offer competitive terms for wholesale mobile wireless arrangements (such as roaming agreements and tower sharing arrangements) to carriers that are also rivals in retail mobile wireless services markets in Canada.  This artificially and unnecessarily limits the ability of new mobile wireless carriers ("new entrants") to offer competitive choices to retail consumers and apply competitive pressure to incumbents.
  3. There is strong evidence that incumbents may not be offering competitive terms for wholesale wireless arrangements to new entrants. In particular the Canadian Radio‑television and Telecommunications Commission (the "CRTC") recently concluded in Telecom Decision CRTC 2014‑398 ("CRTC 2014‑398") that it was "clear" that Rogers Communications Partnership ("Rogers") has engaged in unjust discrimination and undue preferences, contrary to subsection 27(2) of the Telecommunications Act,Footnote 1 in its wholesale roaming arrangements with certain new entrants – both in respect to price and non‑price terms.Footnote 2
  4. While section 27.1 of the Telecommunications Act establishes the maximum amount that an incumbent can charge other carriers for wholesale roaming, it does not address non‑price terms. It is therefore necessary for the CRTC to consider non‑price terms and conditions in wholesale roaming agreements in this proceeding.
  5. The economic incentive incumbents have to engage in vertical foreclosureFootnote 3 resulting in the unjust discrimination and undue preference in respect of wholesale roaming remains, notwithstanding the fact that Parliament has capped the amount an incumbent may charge a new entrant for wholesale roaming. The capping of these rates will likely increase the incumbents' incentive to constrain the non‑price terms of wholesale roaming agreements, or otherwise degrade the quality of the service they offer to carriers that are their rivals in retail mobile wireless services markets.
  6. Thus, it remains important to directly address, and ideally remove, the incumbents' incentive to engage in vertical foreclosure. Such steps will remove, among other things, the incumbents' incentives to constrain the non‑price terms of wholesale roaming agreements, or otherwise degrade the quality of the services they provide to new entrants.  Ultimately, the development of wholesale mobile wireless services markets relying on strong competitive market forces should be the long‑run objective of this proceeding.
  7. In the remainder of this intervention, the Commissioner provides further comments on the issues discussed in his first intervention in light of additional evidence that has been put on the record in this proceeding since his first intervention. The Commissioner also provides comments on regulatory measures available to the CRTC.

II. Incumbents possess market power in retail mobile wireless services markets

  1. The Commissioner submits that the incumbents possess market power in retail mobile wireless services markets in Canada. The appropriate framework for assessing market power is set out in the Abuse of Dominance Guidelines and the Merger Enforcement Guidelines.Footnote 4 The CRTC has endorsed this approach and in Telecom Decision CRTC 94‑19 highlighted "a general consensus that the concept of 'market power', commonly used in competition law, should be the standard by which to determine whether a market is competitive"Footnote 5 in telecommunications proceedings.
  2. The evidence put forward in the Brattle ReportFootnote 6 demonstrates that Canadian retail mobile wireless services markets are characterized by above‑normal profits and comparatively low service penetration levels. These are both indicators of market power.
  3. In addition, Canadian retail mobile wireless services markets are characterized by high concentration and very high barriers to entry and expansion.  Furthermore, such markets are characterized by other factors that, when combined with high concentration and high barriers to entry and expansion, create a risk of coordinated conduct in these markets.Footnote 7
  4. The mere exercise of market power, whether on a unilateral or coordinated basis, may not by itself raise concerns under the Competition Act.  However, as noted in his first intervention, the Commissioner submits that, due to the vertically‑integrated nature of Canadian mobile wireless services markets, the CRTC should take the incumbents' retail market power into account when considering both incentives and opportunities for anti‑competitive conduct in upstream wholesale mobile wireless services markets.
  5. The incumbents assert that retail mobile wireless services markets are competitive.Footnote 8 The incumbents' experts are however more qualified in their conclusions. The incumbents' experts make statements such as: "Canadian markets are workably competitive"Footnote 9 and "[t]here is no reliable evidence that competition in the provision of wireless services in Canada is insufficient or weak".Footnote 10
  6. The CRTC should reject the incumbents' assertions about the competitiveness of retail mobile wireless services markets for the following reasons:
    • As demonstrated in the expert report commissioned by the Commissioner and prepared by Doctors James D. Reitzes, Kevin Hearle, Giulia McHenry, Jeremy Verlinda and Coleman Bazelon of the Brattle Group that is filed together with the Commissioner's second intervention (the "Supplementary Brattle Report"), there are a number of weaknesses in the analysis submitted by the incumbents' experts. The Supplementary Brattle Report shows that there are credible scenarios under which a vertically integrated wireless carrier would have the incentive and ability to raise its wholesale price to rival carriers, so as to increase downstream rivals' input costs and resulting downstream prices. It concludes:
      • Evidence suggests that there are geographic areas in Canada where small wireless carriers and MVNOs [Mobile Virtual Network Operators] have only one effective roaming service provider. In these cases, vertically integrated wholesale providers appear to have strong incentives to use price and non‑price mechanisms to raise rivals’ costs and sabotage the quality of rivals’ retail wireless service.
      • Competition among two or more upstream suppliers to offer wholesale wireless services to a downstream rival provides no guarantee that input prices and service quality will be offered on competitive terms. Wholesale suppliers’ incentives to foreclose downstream rivals will remain regardless of the number of upstream suppliers. Moreover, only under certain circumstances will the presence of multiple upstream suppliers limit their ability to raise rivals’ costs or sabotage quality, even when the bid non‑cooperatively against each other to secure wholesale agreements.
      • The report submitted by Baziliauskas and Mathewson (B&M) on behalf of Rogers ignores the possibility of imperfect information. B&M argue that there is no evidence to indicate any existing incumbent coordination, and that vertically integrated carriers have little incentive to increase costs to downstream rivals. However the model presented by B&M depends on an assumption of perfect information, both on suppliers’ costs and of consumers’ valuations. As explained in the Supplemental Brattle Report, relaxing this critical assumption reintroduces the possibility of incentives and ability to vertically foreclose.
    • The expert economic submissions of all three incumbents either rely on, or make reference to, the profitability analysis that forms the centerpiece of the C‑W Report.Footnote 11 In his submissions in Telecom Notice of Consultation CRTC 2013‑685 (the “Roaming Proceeding”), the Commissioner identified several material deficiencies in the profitability analysis that forms the centerpiece of the C‑W Report and filed in this proceeding his own expert report, which demonstrates that TELUS’ and Rogers’ wireless businesses are generally earning above‑normal returns on their investments, consistent with the exercise of market power.
    • The incumbents' experts who opine for them on the competitiveness of retail mobile wireless services markets have not reviewed or considered the incumbents' confidential information or the incumbents' internal company documents.Footnote 12 The incumbents' experts have instead prepared reports that consider only public information. The incumbents are large and sophisticated enterprises that create and record masses of information relevant to assessing market power in the ordinary course of their businesses.  The incumbents' experts have chosen to express their opinions based on limited public information notwithstanding that relevant confidential information and internal company documents are uniquely available to the incumbents. This choice to prepare expert reports without information that is uniquely available to the incumbents makes the incumbents’ experts’ approach suspect, and their experts’ conclusions inherently unreliable.
  7. In light of additional evidence that has been put on the record in this proceeding since his first intervention, the Commissioner affirms his conclusion that the incumbents possess market power in retail mobile wireless services markets in Canada.

III. The vertical foreclosure incentive

  1. In his initial intervention, the Commissioner articulated his view that incumbents may have an incentive and ability to raise the rates that they charge new entrants for wholesale roaming services, and potentially other services, because such price increases raise downstream rivals' costs and cause retail customers to switch from the new entrants to the incumbents' own retail operations.Footnote 13 This incentive arises because incumbents are vertically‑integrated.
  2. In CRTC 2014‑398, the CRTC concluded that it was "clear" that Rogers had engaged in unjust discrimination and undue preferences, contrary to subsection 27(2) of the Telecommunications Act, in its wholesale roaming arrangements with certain new entrants – both in respect of price and non‑price terms.Footnote 14  In reaching its conclusion, the CRTC noted that Rogers' ability to engage in these practices was attributable to its "stronger bargaining position".Footnote 15  The evidence in this proceeding suggests that Rogers' incentive, and the incentive of other incumbents, to engage in these and other practices that raise downstream rivals' costs is attributable to the incumbents' vertical foreclosure incentive.
  3. In discussing the appropriate remedies in CRTC 2014‑398 the CRTC noted the new entrants' submissions that "an interim remedy [should] be put in place, until the conclusion of" the present proceeding.Footnote 16  CRTC 2014‑398 indicates that the CRTC may have accepted the new entrants' invitation for an "interim remedy" as the CRTC found for the purpose of that proceeding that it would not put in place a remedy for unjust discrimination with respect to wholesale mobile roaming rates given the price caps now set out in section 27.1 of the Telecommunications Act.Footnote 17  The CRTC further noted that its decision did not address other remedial terms and conditions, apart from roaming rates and exclusivity in wholesale roaming arrangements, such as seamless roaming or the regulation of wholesale roaming as these "issues are under consideration" in the present proceeding.Footnote 18
  4. The Commissioner's position is that this proceeding provides the CRTC with an appropriate forum to address generally the incentives that cause the incumbents to raise downstream rivals' costs and engage in vertical foreclosure.
  5. While section 27.1 of the Telecommunications Act reduces the ability of incumbents to act on their incentive to raise downstream rivals' costs and engage in vertical foreclosure, the Supplementary Brattle Report demonstrates that this wholesale price cap will likely result in incumbents constraining the non‑price terms of wholesale roaming agreements.  Incumbents may act on these incentives in a variety of ways.  For example, incumbents may deny or delay seamless roaming, impose restrictive terms and conditions, or otherwise degrade the quality of service they offer to carriers that are potential rivals in retail mobile wireless services markets.  Given the difficulty of identifying all potential forms of unjust non‑price discrimination and non‑price undue preferences, it will likely be difficult for the CRTC to curtail vertical foreclosure by prohibiting specific conduct. It is therefore preferable to address the foreclosure incentive directly by encouraging facilities based competition and thereby eliminating the foreclosure incentive.

Evaluating the incentives for vertical foreclosure

  1. To evaluate the incentives for a vertically‑integrated provider of roaming services to raise downstream rivals' costs and engage in vertical foreclosure, the relevant considerations for the CRTC in assessing the vertical foreclosure incentive are noted in the Commissioner's first intervention (at paragraph 31):
    • Is switching difficult for wholesale customers?
    • Do wholesale customers pass on wholesale price increases to retail customers?
    • Do customers switch from new entrants to incumbents?
    • What is the profit gain when customers switch away from new entrants?
    Below, the Commissioner comments on the evidence, based on the record to date, that incumbents have had, and have acted upon, an incentive to raise downstream rivals' costs and engage in vertical foreclosure.

Switching is difficult for wholesale customers

  1. When determining whether wholesale customers could switch to alternative services, it is important that the CRTC consider the extent to which it is feasible for a wholesale customer or third party to duplicate an incumbent's facilities or services.  The high costs of building mobile wireless towers and acquiring sufficient spectrum on a national basis means that, in many cases, new entrants will need to reach roaming agreements to provide national coverage to their customers.
  2. Across Canada, there are generally no more than two mobile wireless carriers with extensive coverage in any given region. Rogers operates its proprietary nationwide network, while Bell and TELUS achieve a national network by way of network sharing agreements. Thus, there are two network providers in much of Canada; in many rural regions, however, there is only one.Footnote 19
  3. It is possible that even a modest number of regions where there is only one provider could give a carrier significant leverage in negotiations for roaming agreements. The relevant question is how much less attractive would the new entrants' retail offering be to retail consumers if this regional service gap existed?
  4. The incumbents' interventions do not provide any evidence on the extent to which there are areas with only one suitable wholesale provider. More fundamentally, they offer no evidence regarding the influence that these single facility areas have over the ability for wholesale customers to negotiate competitive roaming rates, both in these areas and more generally.
  5. Additionally, even in those regions where there are two networks, these networks are imperfect substitutes, and there can be impediments to wholesale customers switching networks.  For example, several types of legacy handsets, including widely used models such as BlackBerry, cannot operate on the Bell/TELUS network.Footnote 20 In the Commissioner's view, this may be an ongoing source of switching costs.  As wireless technology advances in future years, handsets that are currently supported by both the Bell/TELUS and Rogers networks may be left behind by one incumbent network operator, thereby compounding these problems.
  6. In sum, in light of the record to date, the Commissioner's view is that switching providers of wholesale roaming services is difficult in some cases for new entrants requiring roaming services.

Wholesale customers are passing on wholesale price increases to retail customers

  1. When prices for wholesale services increase, wholesale customers may pass the extra costs along to their retail customers by raising their own prices.  This may significantly weaken these carriers' positions in downstream retail markets, reducing competitive pressure on the incumbents' own retail offerings.
  2. It is important to note that the pass‑through of the extra costs need not be entirely through higher prices, but may also occur through reductions in non‑price dimensions of service.  An example of this is the choice by new entrants to not offer 3G connections to their customers when they are roaming (the "throttling problem").  For example:
    • Bragg Communications Inc. ("Eastlink") states that it does not request the incumbents' LTE roaming services because of the cost to Eastlink for data usage by customers using this technology, given the wholesale data rates.Footnote 21
    • Globalive Wireless Management Corp. ("WIND") states that, as of July 2014, it was offering only 2G connections outside its own network area, and was making this choice because of the high prices it was being charged to send data on other carriers' networks.Footnote 22
  3. This throttling problem appears to be particularly demonstrative of pass‑through that can be traced to the incentive to raise downstream rivals' costs and engage in vertical foreclosure.  An incumbent that was pricing to maximize wholesale profits would not necessarily raise wholesale prices so high that demand by the wholesale customer for a particular service is completely choked off.Footnote 23 Rather, the incumbent could impose non‑price restrictions or raise the wholesale price to a point that it forces its wholesale customers to compete less vigorously. This potentially represents a deadweight loss for the Canadian economy.

Customers are switching from new entrants to incumbents

  1. In their submissions in the Roaming Proceeding, WIND and Eastlink indicated that using "pricing zones" has caused them to lose a substantial number of customers.
    • WIND indicated that because it had to pass its high roaming prices on to its customers, its services were less attractive to a large segment of potential customers.  As a result, its initial rate of customer acquisition slowed and WIND has not been able to achieve reasonable scale.Footnote 24
    • Eastlink had to abandon the zone approach to retain its customers, at the cost of absorbing a financial loss every time its customers travel outside of Eastlink's home network. Footnote 25

Incumbents gain profit when customers switch away from new entrants

  1. In his first intervention, the Commissioner noted that to assess the incumbents' ability and incentive to raise downstream rivals costs and engage in vertical foreclosure, the CRTC should consider the margins that the incumbents earn from their retail offerings.  The reason is that it is not the number of customers gained, per se, that matters. Rather, it is the profit gained that matters.
  2. The Supplementary Brattle Report quantifies this profit gain. In particular, it shows:
    • Vertically integrated MNOs [Mobile Network operators] have incentives to pursue non‑trivial markups on wholesale wireless services. In the Supplemental Brattle Report, it is found that the average raising rivals’ cost upward pricing pressure on wholesale wireless services could be as much as $9 across Canada nationally, and much higher for the dominant carriers in certain provinces.
  3. In sum, in light of the record to date, the Commissioner submits that incumbents have an incentive to raise roaming rates charged to new entrants above the levels that would exist but for vertical integration.  Further, the Commissioner submits that these higher roaming rates have substantially lessened and prevented competition in retail mobile wireless services markets in Canada, and reduced consumers' surplus, total economic welfare and the efficiency of the Canadian economy.

IV. The CRTC should remove the vertical foreclosure incentive

  1. In the Commissioner's view, the CRTC should consider measures to address the incumbents' incentive to raise therates that they charge new entrants for wholesale roaming services, and potentially other services, because such price increases cause retail customers to switch from the new entrants to their own retail operations.
  2. The implementation of section 27.1 of the Telecommunications Act reduces the ability to act on this incentive.  However, as the Supplementary Brattle Report demonstrates, this price cap will likely result in incumbents constraining the non‑price terms of wholesale roaming agreements (e.g., denying or delaying seamless roaming), or otherwise degrading the quality of the service they offer to carriers that are their rivals in retail mobile wireless services markets. The Supplementary Brattle Report demonstrates the link to these incentives.  It states:
    • Implementing price caps alone does not remove the incentive and ability of large vertically integrated Canadian wireless carriers to employ non‑price foreclosure mechanisms. The CRTC may wish ultimately to consider remedies that would limit the ability of large MNOs to engage in non‑price mechanisms that reduce downstream rivals’ quantities or skew such rivals’ offerings in ways that protect the MNO’s own market positions.
  3. To remove the vertical foreclosure incentive, the CRTC should consider regulatory measures to enhance the competitiveness of wholesale tower sharing services.  Facilities‑based competition is the most direct way to remove the vertical foreclosure incentive.  Eastlink submits that it will take many years to build out its networks sufficiently that it could compete with Bell/TELUS or Rogers for the provision of wholesale roaming services.Footnote 26 However, given that it will take several years, in the interim, regulatory measures that enhance the competitiveness of wholesale tower sharing services could improve existing wireless facilities and make switching between wholesale roaming suppliers a more realistic alternative for new entrants
  4. Additionally, the CRTC could consider implementing vertical separation between incumbents' wholesale and retail operations.  The information available on the public record of this proceeding is not sufficient for the Commissioner to offer a view on the overall merits of this alternative.  From an economic perspective, any such remedy involves both gains (i.e., the elimination of the vertical foreclosure incentive) and costs of lost vertical efficiencies.
  5. The Further Revised and Restated Amended Consent Agreement between the Commissioner, major Canadian banks and Interac provides a vertical separation model for the CRTC to consider.  This Consent Agreement, approved by and registered with the Competition Tribunal, contains various safeguards to discourage vertical foreclosure, in addition to caps on the rates that an upstream entity can charge to downstream rivals of its shareholders.Footnote 27  The additional safeguards include specified independence requirements for decision‑makers and the imposition of a legal requirement on those decision‑makers prohibiting acts in furtherance of vertical foreclosure.Footnote 28 The CRTC could also consider prohibiting the communication of confidential information between the wholesale and retail operations of the incumbents' business as it has in other CRTC proceedings.Footnote 29

V. Conclusion

  1. As noted above, in the Commissioner's view, the CRTC should directly address the vertical foreclosure incentive. The implementation of section 27.1 of the Telecommunications Act addresses the vertical integration incentive in the short term. But the end game for the CRTC should be to establish a market structure that is both free of vertical foreclosure incentives and does not require a price cap.
  2. To achieve this, the CRTC should act aggressively to assure that wholesale tower sharing services are available. Additionally, vertical separation should be considered.
  3. As he did in his interventions in the Roaming Proceeding, the Commissioner again states his position that an appropriate remedy for vertical foreclosure incentives is one that is sufficiently effective to ensure that competitive rates are reached.Footnote 30 If, in deciding upon an appropriate remedy, the CRTC is faced with a choice between a remedy that goes further than is strictly necessary to eliminate the vertical foreclosure incentives and a remedy that does not go far enough, the Commissioner would prefer the former remedy.Footnote 31
  4. Ultimately, the development of wholesale mobile wireless services markets relying on strong competitive market forces should be the long‑run objective of this proceeding. At an appropriate point in the future, the CRTC should re‑examine the need for regulatory measures in wholesale mobile wireless services markets. Such a review should focus on the factors discussed above in section III of the Commissioner’s intervention.
Date modified: