Telecom Notice of Consultation CRTC 2019-57 — Further Comments of the Competition Bureau

I. Competition Bureau proposes wireless policy to lower prices for Canadians

  1. This proceeding is an important opportunity to ensure that Canadians can access affordable and high-quality wireless services. The Canadian Radio-television and Telecommunications Commission (CRTC) initiated this proceeding to ensure their regulatory framework facilitates “sustainable competition that provides reasonable prices and innovative services, as well as continued investment in high-quality wireless networks in all regions of the country.” As a key driver of these outcomes, competition is at the forefront of this proceeding.
  2. In this submission, the Competition Bureau (Bureau) aims to provide the CRTC with advice on enhancing wireless competition based on a principled and evidence-based assessment of the state of wireless competition in Canada. To this end, the Bureau and its economic expert have conducted in-depth analyses of Canadian wireless competition using extensive confidential data from Canadian wireless carriers. Additionally, the Bureau has completed an extensive scan of international wireless markets in order to inform its recommendations.
  3. The findings of the Bureau and its economic expert can be summarized as follows:
  4. Bell, Telus and Rogers (the Big 3) possess market power at both the retail and wholesale level in most regions in Canada. The Canadian wireless market is highly concentrated in the hands of three players who enjoy high levels of profitability compared to both their international and domestic peers.
  5. Where the Big 3 face a wireless disruptor, prices are significantly lower. Facilities-based regional competitors who operate their own wireless networks, such as Sasktel, Videotron and Freedom Mobile, are increasingly disrupting the Canadian wireless landscape. Prices are generally in the range of 35-40% lower in the parts of Canada where wireless disruptors have achieved a market share above 5.5%.
  6. Wireless disruptors offer the most promising path forward. They drive lower prices, greater choice and increased levels of innovation in Canada over the long term. As opposed to service-based Mobile Virtual Network Operators (MVNOs), who are necessarily beholden to incumbents’ networks, these regional disruptors have the required independence and infrastructure to act aggressively to fight for their share of the pie.
  7. Canadians could save substantially through more competition from wireless disruptors. While many Canadians are increasingly benefitting from competition driven by wireless disruptors, the full effects of a more competitive wireless industry have not yet been experienced. The Bureau’s research finds that promoting the growth and expansion of wireless disruptors, even to reach only 5.5% market share, has the potential to deliver significant price reductions. The benefits are much higher if regional carriers can attain a market share of 20%.
  8. MVNOs can drive lower prices and greater choice, but they also could threaten the demonstrated progress in enhancing competition in this industry to date. While MVNOs can have positive effects on pricing in the marketplace, they are unlikely to deliver the benefits of sustained and vigorous competition that facilities-based wireless disruptors are capable of providing. The Bureau is concerned that the introduction of MVNOs would disproportionately affect these wireless disruptors, putting at risk the positive effects that they have had on pricing, and may impact long-term incentives to invest in high-quality networks in Canada.
  9. The CRTC should adopt an MVNO policy that is temporary and focused on incentivizing and accelerating facilities-based competition from disruptors. The CRTC should allow wireless disruptors to act as MVNOs as a transitional step to becoming full-fledged facilities-based providers as they continue their expansion. This would ensure that the progress made by these providers to date will continue to pay dividends to Canadians. An investment-based MVNO policy achieves the goal of spurring additional price competition from wireless disruptors in the short term, while avoiding the risk of declining network quality in the long term.
  10. Additional measures can also improve the level of competitive intensity in the Canadian wireless market. Measures aimed at lowering barriers to entry and reducing switching costs would support greater competition in the Canadian wireless marketplace. These measures include mandated seamless handoff, more effective tower sharing and site access rules, and updated roaming rates that better represent the current marketplace.

II. Introduction and information relied upon

  1. The Commissioner of Competition (Commissioner) and his economic expert are pleased to provide the CRTC with an analysis of competition in the wireless marketplace and a recommended path forward to promote competition in this industry.Footnote 1 As an independent agency at arms length from the industry, the Bureau brings a unique perspective, as an intervener without any commercial interest in the outcome of the Proceeding.
  2. The Bureau’s recommendations and findings herein are based on the following evidence.
    1. The Bureau commissioned an expert economic study (Matrix Study) conducted by Matrix Economics. Matrix Economics employs a team of highly trained economists with expertise in assessing competition issues, including in telecommunications markets. The Matrix Study analyzes the state of competition in the retail wireless marketplace and evaluates the benefits of incentivizing additional competition among wireless service providers.
    2. The Bureau and its expert based their analyses on confidential data and information provided by 19 Canadian wireless service providers in response to the CRTC’s disclosure order,Footnote 2 supplemented by publicly available information.
    3. The Bureau also conducted a study of international jurisdictions, including those that have adopted mandated MVNO access. The findings of the study are based on a review of publicly available information as well as interviews with industry stakeholders in the countries studied, including telecommunications regulators.
  3. The Bureau believes in the importance of transparency. Transparency in our agency’s analysis helps to ensure that our recommendations are fully informed with a clear understanding of the evidence, minimizing the potential for ineffective policy and unintended consequences. The Bureau and its expert have therefore sought to disclose their findings and underlying methodology wherever possible without divulging competitively sensitive information.

III. Guiding premise: understanding current competitive dynamics is critical to designing an effective remedy

  1. In a perfectly competitive market, social welfare is maximized, and outcomes are efficient. However, in some situations referred to as “market failures”, the conditions required to achieve the efficient outcome are not present. Market failures can be an important reason why the government may need to intervene through regulation.Footnote 3 Before pursuing regulatory intervention, it is critical to understand whether or not there has been a market failure, and, if so, to understand the extent and scope of the market failure. This allows for an appropriately designed remedy, which is the least intrusive course of action to reach the desired effect.
  2. An exercise of market power can be an example of a market failure since it can produce inefficiencies. When market power is exercised, it has a negative impact on consumers by increasing the price firms can charge above what can be justified based upon their costs. When consumers are faced with these higher prices, they, in turn, lower the amount they consume. This exercise of market power forces some consumers, who would have otherwise purchased a good or service at a competitive price, to reduce or forego the purchase altogether, because prices are too high. These foregone purchases create what economists refer to as deadweight loss, or allocative inefficiency.Footnote 4 Market power can also manifest itself in other ways referred to as “non-price effects” which can take the form of a reduction in service, quality, product choice, incentives to innovate or other dimensions of competition that customers value.
  3. The Bureau’s report is organized to follow this logical progression where Part 1 starts by determining whether there is a market failure, in the form of market power, which may merit regulatory intervention by the CRTC and seeks to “diagnose” the extent of the competition issue, including what markets are affected and to what degree.
  4. With a complete understanding of the problem at hand, in Part 2 of this report the Bureau assesses the benefits and challenges associated with potential remedies and recommends a focused remedy that seeks to stimulate competition in the short run but preserves the benefits of a market-oriented solution in the long run.

Part 1: The current state of wireless competition in Canada

IV. The national wireless carriers continue to possess retail market power

  1. An assessment of wireless competition in Canada following the CRTC and Bureau frameworks finds the Big 3 possess market power at the retail level. In this section, the Bureau first discusses the appropriate product and geographic market definition, as well as the applicability or relevance of precisely delineating markets. Then, the Bureau tests for the presence of indicators that are typically present where firms have market power such as high market shares and barriers to entry.
  2. A significant benefit of the wireless carrier data made available to the Bureau by the CRTC is that beyond these indicators, the Bureau’s expert was able to test directly for market power by conducting an analysis which tests for “the ability, within a relevant market, of [the Big 3] to raise or maintain prices above those that would prevail in a competitive market,” which follows the CRTC’s definition of market power.Footnote 5 The Matrix analysis is conservative in the sense that they compare markets with less competition to markets with more competition rather than to a perfectly competitive benchmark. The Matrix Study findings are also applicable to the CRTC’s assessment of rivalry among, and substitutability of, the wireless providers. This is followed by a discussion of qualitative evidence that appears to reaffirm the Matrix Study’s finding that these high prices have resulted in pent up demand for more data in the Canadian wireless market. The Bureau also responds to industry commentary that Canadians are willing to pay higher prices for high quality networks and that prices have declined to competitive levels.

a. Indicators suggest the Big 3 have market power in many regions in Canada

  1. The keystone to regulatory intervention in the wireless market is an assessment of market power. Market power can lead to higher prices, reduced output and loss of economic efficiency. The Bureau evaluated a number of indicators consistent with the framework established by the CRTC in Telecom Decision 94-19Footnote 6 and found the features of the wireless market are consistent with a finding of market power.

Product market

  1. Wireless service offerings can vary, with each provider selling a particular range of service plans. There may be groupings of wireless service plans that themselves qualify as relevant product markets. A candidate product market may be plans with large amounts of data tailored to customers who consistently use large amounts of data each month. These customers would be unlikely to switch to a plan with a low data allowance if faced with a price increase.Footnote 7 However, it was unnecessary to draw bright lines across the continuum of wireless offerings since competitive conditions are similar across different candidate relevant product markets.Footnote 8 Precisely delineating each market is therefore not necessary. For example, in Quebec, regardless of whether a customer wants to purchase a plan with a low or high data allowance, their competitive options are either one of the Big 3 or Videotron.
  2. There are some minor variations in offerings across carriers; however, these differences are unlikely to affect the competitive effects analysis for three reasons.
    1. Differences in plan offerings are minor. For example, in Quebec, Videotron does not offer prepaid plans. Prepaid plans account for 12% of subscriptions in Canada and there likely are comparable postpaid offerings that compete with prepaid plans.Footnote 9
    2. The strengths and weaknesses of each carrier’s product offering is accounted for in the competitive effects analysis. For example, if a regional carrier is not adequately catering to the demand in the market where they operate, the competitive effects analysis is unlikely to show they are having an impact on prices.
    3. Even if a competitor currently does not offer a particular plan, the introduction of a new plan likely does not involve a significant sunk investment so a competitor may be able to rapidly begin supplying the market in response to an attempted price increase for a specific product.Footnote 10
  3. The competitive effects analysis in the Matrix Study allows the CRTC to identify where the Big 3 possess market power. Those conclusions apply regardless of the product market being examined as the same competitive conditions exist regardless of product segmentation. Competitive conditions vary from a geographic rather than product market perspective. Therefore, for the purposes of the Proceeding the relevant market is the provision of wireless telecommunications services as studied.

Geographic market

  1. A wireless service provider can only sell to customers living within its network’s coverage area, since its customers cannot permanently roam on another service provider’s network.Footnote 11 Practically speaking, this means that a consumer’s available substitutes are dictated by where they live in Canada. Carriers currently operate in regions as narrow as a city (e.g. Eastlink operates in Sudbury and Timmins) or as broad as a province (e.g. SaskTel operates across the entire province of Saskatchewan). Relative price levels between different areas are functional indicators that can be used to determine what geographic areas are substitutes.Footnote 12 The Matrix Study shows that pricing varies in regions as narrow as a city.Footnote 13 Therefore, the relevant geographic market is likely either as narrow as a city, or as broad as a province, depending on the region.
  2. More specific guidance on the relevant markets wherein the Big 3 likely possess market power will follow in discussing the competitive effects analysis conducted in the Matrix Study.

Market share

  1. The Canadian wireless market is highly concentrated by any measure.
    1. Market shares- The Bureau has established market share and concentration thresholds in merger reviews below which effective competition is generally likely to constrain increased market power due to the merger. For a unilateral exercise of market power the threshold is 35% for the merged firm. The market share held by the largest firm exceeds the 35% safe harbour thresholdFootnote 14 in all markets except Quebec.
    2. Four-firm concentration ratio- The threshold established by the Bureau for a coordinated exercise of market power, is 65% for the four largest firms (known as the four-firm concentration ratio or CR4) and 10% for the merged firm. The four-firm concentration ratio is 100% in every region in Canada except Ottawa,Footnote 15 well in excess of the 65% safe harbour threshold.Footnote 16 Even a three-firm concentration ratio would exceed the 65% threshold, with the Big 3 accounting for 88.3% of the total number of subscribers nationally.
    3. Herfindahl-Hirshman Index (HHI)- the United States Department of Justice (US DOJ) characterizes markets in which the HHI is in excess of 2,500 points as highly concentrated.Footnote 17 The HHI for wireless services is higher than 2,500 points in every Canadian province and territory, ranging from 2,660 in Quebec to 8,158 in the Northwest Territories.Footnote 18
  2. As the Matrix Study shows, this pattern holds at the national, provincial and city levels.Footnote 19

Barriers to entry

  1. In its decision in CRTC 2015-177, the CRTC found that “barriers to entry into the retail market are very high.”Footnote 20 It is highly unlikely that industry conditions have changed to a degree sufficient to warrant the CRTC coming to a different conclusion in this Proceeding. High sunk capital expenditure investments continue to be required. Entry is also dependent on access to spectrum, a finite public resource, a large proportion of which has already been allocated in Canada. The national wireless carriers and incumbent regional carriers have taken decades to construct their existing infrastructures, and thereby possess a considerable advantage over a new company attempting to establish or grow their presence in the wireless industry. Further, the large sunk capital investments already committed by the Big 3 make them very difficult to displace. The regional new entrants in the wireless market have been building their networks over a period of 11 years since they first purchased spectrum in 2008 but are generally only having a significant impact on pricing in recent years.Footnote 21

Profitability

  1. In the 2015 study commissioned by the Bureau, the Brattle Group found that Telus and RogersFootnote 22 were generally earning above-normal returns on their investments,Footnote 23 consistent with an exercise of market power (Brattle Report).Footnote 24 It is likely that this continues to be the case.
  2. As illustrated in Figure 1 below, the Big 3 continue to maintain large profit margins relative to their global peers. This was the case in 2015 when the Brattle Report was commissioned and continues to be the case.

Figure 1 - Average EBITDA as a Percent of Reported Revenue for Wireless Operators (G7 + Australia)Footnote 25

Figure 1 – Average EBITDA as a Percent of Reported Revenue for Wireless Operators (G7 + Australia)
Details of the Average EBITDA as a Percent of Reported Revenue for Wireless Operators (G7 + Australia)
Country 2010 2011 2012 2013 2014 2015 2016 2017 2018* 2019*
Australia 24.9% 26.1% 30.1% 33.7% 35.9% 35.1% 37.4% 34.0% 34.9% 33.7%
Canada 42.2% 40.8% 42.2% 43.6% 42.6% 40.6% 41.2% 42.3% 42.1% 42.6%
France 33.3% 32.1% 28.5% 28.8% 29.0% 32.8% 32.5% 32.8% 35.2% 35.8%
Germany 41.3% 41.0% 39.4% 36.8% 31.2% 30.7% 32.2% 33.9% 34.9% 35.2%
Italy 44.9% 44.6% 43.4% 42.3% 36.7% 33.6% 36.7% 37.3% 39.8% 40.9%
Japan 33.9% 32.7% 32.0% 33.1% 32.9% 13.0% 13.1% 12.3% 17.8% 19.6%
UK 20.2% 21.4% 20.1% 23.0% 21.2% 25.0% 23.8% 24.3% 25.4% 26.7%
US 33.4% 31.0% 31.7% 33.2% 32.4% 36.0% 38.2% 38.6% 41.0% 40.8%
*Estimated (partially for 2018)

  1. The Big 3 have argued that these figures are not comparable given the large fixed costs associated with building a network in a country like Canada, with a large land mass and low population density. Figure 2 below compares capital intensities across countries. Capital intensity is the ratio of capital expenditures to revenues.Footnote 26 A high capital intensity would mean the company reinvests a significant proportion of the revenues they obtain. The Big 3 do not appear to have a higher capital intensity than their international peers, ranking near the middle of the pack. The exception is Shaw, whose capital intensity is higher than any of the other carriers measured, likely due to their recent network upgrade and expansion.

Figure 2 – International carriers’ total wireless capital intensityFootnote 27

Figure 2 – International carriers’ total wireless capital intensity
Details of the International carriers’ total wireless capital intensity
Country Company Capex ($B) Revenue ($B) Capex Intensity
Canada Bell $0.74 $8.32 8.92%
Rogers $1.05 $9.02 11.59%
Telus $0.88 $8.07 10.86%
Freedom $0.34 $0.95 36.13%
United States Verizon $9.53 $91.09 10.46%
AT&T $8.86 $71.74 12.35%
Sprint $10.19 $31.45 32.41%
T-Mobile $5.28 $42.62 12.38%
Australia Telstra $1.76 $10.25 17.1%
Optus $1.16 $5.70 20.44%
Vodafone $0.73 $3.46 21.17%
*Figures are the sum of quarters ending Dec 17, Mar 18, Jun 18, Sep 18
**Revenue reflects total wireless revenue

  1. The Big 3 also maintain high levels of profit relative to their domestic peers. Figure 3 shows that Shaw’s wireless margin is less than half that of the Big 3.

Figure 3 – Canadian carriers’ wireless EBITDA margins, 2018Footnote 28

Figure 3 – Canadian carriers’ wireless EBITDA margins, 2018
Details of Canadian carriers’ wireless EBITDA margins, 2018
Source: Company Annual Reports
Shaw 18.5%
Bell 42%
Telus 43%
Rogers 45%

  1. In spite of lower margins, the new entrants also maintain higher capital intensity than the Big 3, highlighting the major investments currently being made and still required to enter and expand in the wireless industry.

Figure 4 – Canadian carriers’ capital intensity, 2018Footnote 29

Figure 4 – Canadian carriers’ capital intensity, 2018
Details of the Canadian carriers’ capital intensity, 2018
  Bell Rogers Telus Shaw/Freedom Quebecor/Videotron Cogeco
Capital Intensity (Total) 16.9 18.5 16.6 26.1 16.8 18.5
Capital Intensity (Mobile) 7.8 11.8 11.0 36.1 16.5  
Capital Intensity (Wireline) 25.3 36.3 31.3 23.9    

  1. The wireless carriers continue to maintain high levels of profitability relative to both their international and domestic peers, despite maintaining lower capital intensity. It is likely that Brattle Report’s previous finding that the Big 3 are earning above-normal returns on their investments continues to be the case.

b. The Matrix Study provides direct evidence of market power

  1. While several indicators are suggestive of market power, the Matrix Study provides direct evidence that Bell and Rogers - and, by inference, Telus have market power in many regions of Canada. It finds that the Big 3 are able to charge higher prices and offer lower plan limits when competitive discipline from regional competitors is low.
  2. To reach this finding, the Matrix Study employed two widely used econometric techniquesFootnote 30 to assess the competitive impact of regional carriers on the Big 3:
    • Cross-sectional regression analysis (regression analysis): Regression analysis is also used to provide an assessment of the effects of market sharesFootnote 31 of regional competitors on the Big 3 at a city levelFootnote 32 in terms of monthly billingsFootnote 33, plan limits, data usage, and the per GB priceFootnote 34. This regression technique is also used to provide an assessment of the impact of regional competitors’ presence on the following quality metrics for the Big 3 at a city level:
      1. average download speed,
      2. percent of downloads over 5 Mbps, and
      3. speed score (a combination of quality related metrics listed, as well as others like ping time and time on Long Term Evolution (LTE)).
      A number of sensitivity analyses are then undertaken in order to test the robustness of the regression results. The regression analysis forms the primary economic analysis underpinning the overall findings.
    • Event study difference-in-differences analysis (diff-in-diff analysis): The diff-in-diff analysis seeks to understand changes in prices, plan limits and the price per GB before and after entry events or the introduction of new offerings by regional competitors at a city level. This is undertaken by grouping cities into those that experienced a unique event and those that did not. The unique events studied in the Matrix Study include:
      1. Freedom’s introduction of its Big Gig plan in October 2017Footnote 35 and
      2. Eastlink’s entry into 5 cities from 2016 to 2018.
  3. Through these analyses, the Matrix Study finds that competition from the regional carriers has a significant effect in disciplining the pricing of the Big 3. This effect is illustrated by the following key results:
    • Lower prices: The Big 3 charge lower prices, controlling for other factors, in geographies where they face competition from regional carriers.
    • Higher plan limits: The Big 3 offer higher plan limits, controlling for other factors, in geographies with greater regional competition. Further, the diff-in-diff analysis showed that the Big 3 increased their plan limits in response to Freedom’s introduction of their Big Gig plan.
    • Lower per GB prices: The price per GB was lower in areas where regional competitors were present. Further, the diff-in-diff analysis of Freedom’s introduction of their Big Gig plan showed a significant decrease in the per GB price, while there was no measurable effect on monthly billings when considered without adjusting for changes in plan limits.
    • Greater data usage: The Big 3’s consumers use more data, controlling for other factors, in geographies with greater regional competition.
    • No effect on network quality: Regional competition does not have a measurable effect on the Big 3’s network quality,Footnote 36 as measured by download speeds or PC Magazine’s speed score.Footnote 37
  4. However, the Matrix Study notes that the competitive effects of regional competition on the Big 3 are most significant for the mid-sized regional competitors. In particular:
    1. There is no measurable effect of competition from regional competitors on the Big 3’s per GB price in areas where the regional competitor has a market share below about 5.5%.
    2. A regional carrier with a market share above 5.5% and below 20% places increasing downward pressure on the Big 3’s prices as their market share increases. On average, for regional competitors with penetration rates within this range, each additional percentage point of market share gained decreases per GB prices of the Big 3 by about [#######] percent.Footnote 38
    3. There is no measurable incremental benefit of increasing the presence of a regional competitor beyond approximately 20% market share. It is important to note the difference between regional incumbents and regional entrants. Regional carriers who exceed this threshold are incumbent regional carriers like SaskTel so their competitive influence on their markets pre-dates the data period (2014-2018). Prices in their markets were already lower in 2014, so prices in Saskatchewan have declined at a slower rate than in markets with regional entrants. Overall price per GB levels are lower in these markets than in those with regional entrants.
  5. With respect to geographic differences in market power, the Matrix Study tells us the following:
    1. The Matrix Study results indicate that markets with no regional carrier or a regional carrier with a market share below 20% are experiencing the effects of an exercise of market power since the Big 3 can charge significantly higher prices in these areas. This would include all markets except a few cities in Quebec, Saskatchewan and Northern Ontario.
    2. Markets where the regional carrier’s share exceeds 20% are not necessarily competitive, but are experiencing materially greater levels of competition.
    3. It is difficult to make firm conclusions about levels of competition in Manitoba for the following reasons.
      1. The 2016 acquisition of MTS by Bell and the related divestiture of assets to Xplornet and TelusFootnote 39 mark significant structural changes in the market, the effects of which one would expect to take time to fully materialize.
      2. During the data period, following the acquisition of MTS, Bell’s CEO announced they would not raise prices for a period of one year.Footnote 40
      3. The data available following Xplornet’s entry only spans 3 months.Footnote 41
      4. There has been speculation that one of the Big 3 may be planning to purchase SaskTelFootnote 42 and therefore, may be moderating their conduct.Footnote 43
      5. The Big 3 have been under significant scrutiny, including high Canadian wireless pricing being raised in election platformsFootnote 44 and therefore, may be moderating their conduct.
    4. The Bureau’s analysis of data collected by the CRTC shows that pricing in Manitoba followed a similar trend to that of Saskatchewan. It did not increase, but did not decrease as significantly over the period as it did in other markets like Quebec. That said, the Matrix Study would predict that this market may be at risk of an exercise of market power since Xplornet is below the 5.5% threshold with a market share of 2.5%. Further, Xplornet’s network in Manitoba only covers city centersFootnote 45 so outside of major population centers in Manitoba, the Big 3 are the only option for consumers.
  6. An important implication of these findings from a policy perspective is that achieving these economic benefits requires the promotion of competitors that can attain the scale of mid-sized facilities-based providers.

Canadians are likely consuming less than they would in a competitive market

  1. Where market power is exercised, prices are increased above competitive levels, resulting in lower data usage and market penetration than would be seen in a competitive market. This is because consumers are forced to either reduce their consumption in response to a lack of competitive offerings that meet their needs, or remove themselves from the market entirely. The information gathered in this Proceeding suggests that Canadians are likely lowering their consumption in the face of high prices in certain regions in Canada.
  2. The Matrix Study finds that data usage is significantly higher in markets with a regional competitor. This suggests that in markets with less competition, high prices have been restricting Canadians use of wireless services resulting in pent up demand for additional data.
  3. Canadians’ pent up demand for data has also been illustrated by the consumer response whenever a more competitive offering has been introduced in the market.
  4. In 2017, Freedom introduced their Big Gig plan, offering 10 GB of data for $50.Footnote 46 The Big 3 responded by introducing similar promotional plans which led to an intense but brief period of competition (4 days). There was a significant surge in demand for these plans such that Big 3 customer service staff and systems were overwhelmed.Footnote 47
  5. This has also been borne out in the context of high Canadian overage fees and a lack of competitive pricing for high usage data plans. In 2019, Rogers made higher data plans their “flagship” offering and did away with usage overage charges in favour of data throttling. Since then, Rogers has seen triple their anticipated uptake of their “Infinite” plans.Footnote 48 This suggests there was pent up demand for higher data usage but Canadians were choosing to consume less due to a lack of competitive offerings and high penalties if they were to use more than their plan limit.
  6. Rogers CEO Joseph Natale offered commentary on how the prevailing market dynamic had impacted consumer behaviour:
    • “The approach to overage in Canada had seriously decelerated data growth rates. Canadians had become increasingly afraid to use data given the evolution of overage rates in our industry. On a comparative basis, average data consumption in Canada had fallen to one-third the U.S. average and the bottom quartile of the most advanced global markets in the world.”Footnote 49
  7. Rogers explained that customers who have signed up for their “Infinite” plans are using on average 50% more data.Footnote 50
  8. These surges in demand when more competitive offerings are available suggest Canadians’ usage of wireless service has been artificially restricted by supra-competitive pricing, both for plan pricing and overage penalties. This is further illustrated by the Matrix Study finding of higher data usage in regions with more competitive pricing due to wireless disruptors. This is direct evidence of an exercise of market power and the associated allocative inefficiency resulting from Canadians consuming less due to high prices.

The Big 3 have high-quality wireless networks, though that may be in part due to market power

  1. The Big 3 often point to their high-quality networks when opposition to elevated Canadian pricing is raised.Footnote 51 International comparisons do show that Canada has fast speeds and reliable networks. Download speeds obtained through Speedtest by Ookla show that in 2019, Canada had the fastest wireless network in the G7 and the sixth fastest speeds in the world.Footnote 52 High-quality networks are a significant benefit enjoyed by Canadians.
  2. Wireless service providers have suggested that high-quality networks are the reason prices are higher in Canada since Canadians are willing to pay this premium for higher quality networks.
  3. The Bureau cautions there may be explanations other than an inherent demand for fast networks being uniquely present in Canada. In particular, the Bureau’s expert has shown that Canadians are consuming less due to a lack of competition from regional carriers, which results in supra-competitive pricing among the Big 3. As Rogers has described “average data consumption in Canada had fallen to one-third the U.S. average and the bottom quartile of the most advanced global markets in the world.”Footnote 53 This necessarily has an impact on network quality comparisons as there is less traffic on Big 3 networks than their international peers. Faster networks typically result in higher usage levels, not lower, which suggests that the Canadian experience is not demand driven, but rather, due at least in part, to supra-competitive pricing and restricted usage.
  4. Therefore, while the Bureau agrees that Canadians enjoy networks with fast speeds, and that those speeds are valued by consumers, it cautions that high speeds should not be used as a justification for supra-competitive pricing in Canada.

Prices have been decreasing but not necessarily due to competition

  1. In their initial submissions, a number of carriers state that wireless prices in Canada are declining as a result of increased competition.Footnote 54 If this were the case, it could suggest that any market power the Big 3 may have held in the past is being reduced through competition.
  2. Canadians are not necessarily seeing their monthly bill decrease but are receiving larger data buckets. As shown below, in Figure 5 the price a consumer pays for a GB of data has been steadily decreasing. However, their monthly bill has not been decreasing significantly, in fact it has actually slightly increased.

Figure 5 – Comparison of ARPU and $/GB, excluding device subsidiesFootnote 55

[## ##]

  1. The Big 3 acknowledge this trend in their investor calls. For example, Bell described this trend as follows: “[t]he migration of customers to LTE has happened and now the buckets have gotten larger, and as a result, you know, you see the type of ARPU we have got, which, frankly, underlying those two things, is still positive growth from an ARPU perspective.”Footnote 56
  2. Overtime, the market has also moved towards fewer restrictions on call allowances, and more recently, the elimination or reduction of overage charges.
  3. It is unclear how much this downward trend in price per GB can be attributed to an increase in competition. In the alternative, the trend could be attributable to a decrease in costs. Technological advances have likely reduced the cost of delivering a GB of data to consumers, for example, due to increased efficiency in spectrum use.Footnote 57 This is consistent with the downward trend in the price per GB that can be found in many countries across the world. However, Canadian prices are declining at a lower rate. The figure below is a simple average of international pricing from the Wall/Nordicity Report.Footnote 58 Though there is a decline, Canadian prices have decreased over this period less rapidly than those of our peers.

Figure 6 – Comparison of ARPU and $/GB, excluding device subsidies

Figure 6 – Comparison of ARPU and $/GB, excluding device subsidies
Details of the Comparison of ARPU and $/GB, excluding device subsidies
 2018201720162015
Canada$77.6785.60$84.09$91.93
US$73.67$65.48$79.19$103.00
UK$30.00$34.04$35.97$66.33
France$32.50$36.18$56.60$56.90
Germany$47.00$72.88$70.85$82.07
Italy$25.50$38.32$48.41$64.95
Australia$23.67$54.47$47.45$81.73

  1. Regardless, the Matrix Study accounts for any trend in price, including a downward trend, and still finds regional carrier competitive effects. This result indicates that irrespective of whether there has been a general decline in prices, prices would be lower with greater competition.

c. Conclusion- the Big 3 possess retail market power

  1. The wireless industry exhibits market power indicia – high concentration, high profitability, and high barriers to entry. Beyond that, the Matrix Study provides direct evidence that the Big 3 are able to charge higher prices where they do not compete against a wireless disruptor, and in turn, consumers can only afford to use lower amounts of data.
  2. Prices are decreasing on a per GB basis, however:
    1. customers’ actual monthly bills have actually been increasing slightly as they purchase larger data buckets; and
    2. this decline in per GB price is likely attributable, at least in part, to a decline in costs with improvements in technology going from 3G to 4G and soon to 5G.
  3. However, irrespective of a decline in price, the Matrix Study shows that prices would likely be lower if there was more competition so this trend does not alleviate the Bureau’s concerns of market power. This is consistent with international pricing trends which show the prices in Canada are declining at a slower rate than they are internationally. This leads the Bureau and the Matrix Study to the conclusion that the Big 3 possess market power at the retail level in most of the regions in which they operate.

V. Market power likely exists at both the retail and wholesale levels

  1. While the CRTC indicated that the Proceeding would focus on the question of competition in the retail market,Footnote 59 it is still important to consider whether market power has been maintained at the wholesale level following the CRTC’s 2015 finding.Footnote 60 It follows that due to the vertically integrated nature of the market, a retail market power finding implies a wholesale market finding.
  2. A finding of market power at the retail level implies there is also likely market power at the wholesale level given the vertically integrated nature of this industry. Retail pricing is supracompetitive and there has been insufficient wholesale competition to undermine those high retail prices. Therefore, the findings previously described apply to both the retail and wholesale levels of the market.
  3. That said, as described in the subsequent section, the fact that minimal to no wholesale market has developed, except through regulatory intervention, further indicates that the Big 3 possess market power at the wholesale level of the market.

a. A lack of naturally occurring MVNOs in Canada may itself be a sign of market power

  1. That said, in addition to the reasons outlined in section III above, the fact that a wholesale market has not developed without regulatory intervention (i.e., a naturally occurring MVNO or roaming market) is likely a sign of insufficient competition and market power at the wholesale level in and of itself. Based on international experience, MVNO markets occur naturally when one or more of the following conditions exist:
    1. there is excess capacity at the wholesale level;
    2. there is differentiation between the MNOs’ offerings; or
    3. there is a willingness on the part of MNOs to negotiate wholesale agreements with MVNOs.
    These conditions are most likely to exist in a market with strong competition and do not exist in Canada.

Excess capacity

  1. Robust MVNO markets have emerged in several jurisdictions that do not mandate MVNO access, such as in the US, the UK, the Netherlands and Australia. In these markets, there are or were, four MNOs operating in the market when MVNOs emerged. A greater number of networks may contribute to higher levels of excess network capacity in these jurisdictions. Excess capacity creates an economic incentive to earn additional profit by selling that capacity to MVNOs. In the Canadian market, in which there are three national MNOs operating, the network sharing agreement between Bell and Telus means that there are in reality only two national networks. This likely leads to less excess capacity, and a lower profit incentive to negotiate with MVNOs.

Differentiation

  1. Another competitive factor which may facilitate commercially negotiated MVNO agreements is the degree to which MNOs are differentiated. Differentiation can disrupt coordination, including coordinated efforts to keep MVNOs from entering the market.Footnote 61
  2. In contrast, in Australia, for example, the service offerings, network coverage and market share of its three MNOs are highly differentiated. In Australia’s large and sparsely populated outback region, Telstra is often the only service option, conversely in Australia’s urban areas, Vodafone offers high volumes of data relatively cheaply.Footnote 62
  3. The effects of differentiation in wireless markets in the UK are also noted by the Organisation for Economic Co-operation and Development (OECD):
    • “Indeed, a degree of asymmetry [among MNOs] may actually be beneficial. The fact that some of them have more spectrum of one type and less of another, as compared with their competitors, may have the potential to increase competition, as it is likely to require the different MNOs to adopt different commercial strategies to compete in the market, creating the potential for greater innovation and differentiated competition (and perhaps reducing the risk of tacit collusion [another term used to describe coordination]).”Footnote 63
  4. In contrast, Canada’s 3 national MNOs offer similar network coverage and service offerings and have symmetric market shares nationally.

Willingness of MNOs

  1. According to a recent study of nearly 1,500 MVNOs globally prepared for the New Zealand Commerce Commission (NZCC Study), the willingness of MNOs to open capacity to third parties has been the biggest commercial driver for a successful MVNO market. They suggest it is often sparked by a fully-established operator leading the way, and others following suit.Footnote 64
  2. In a number of jurisdictions that the Bureau has studied, mandated MVNO access was imposed on one or more MNOs, which were previously unwilling to negotiate with MVNOs, due to a coordinated effort to thwart wholesale competition. The introduction of a mandated MVNO regime successfully sparked competition in the wholesale market by disrupting coordination amongst the MNOs.
  3. For example, prior to mandating MVNO access, Spain’s competition authority observed collective dominance in denying access to MVNOs on the part of the country’s three MNOs. Collective dominance is a term used to describe coordinated behaviour.Footnote 65 However, once MVNO access was mandated, it opened the door to competition in the wholesale market and by 2017 there were over 30 MVNOs operating in the wireless market in Spain.Footnote 66 Similarly, while mandated access was only imposed on the acquiring firm Hutchison in Austria since it was in the context of a merger remedy, the agreement drove other MNOs in the market to enter into MVNO agreements of their own in order to capture wholesale revenue.Footnote 67
  4. In their initial submissions, a number of carriers suggest there has been no refusal on the part of MNOs to negotiate agreements with MVNOs. For example, SaskTel submits that:
    • “We note that various forms of MVNOs have appeared in Canada in the time before this NC. Many have launched, some have failed, some have been purchased by facilities-based providers, and some continue to operate. Such providers include: 7-Eleven Speak Out Wireless, DCI Wireless, Execulink Mobility, Onstar, Petro Canada Mobility, Zoomer Wireless, PC Mobile and Virgin Mobile (now owned by Bell Canada). These types of arrangements will continue to arise and meet varying levels of success. Commission intervention was not required to entice their entrance or foster their continued existence.”
  5. However, commercially negotiated MVNO agreements have not materialized in the Canadian market to a significant degree. As described in the Notice of Consultation the CRTC notes that it has twice had to intervene in the wireless service market to address disputes between a carrier and a potential MVNO when the prospective MVNO was unable to successfully negotiate an agreement with a wireless carrier and that this is “symptomatic” of a broader problem of the lack of an organic retail MVNO market.Footnote 68
  6. A number of submissions made in this Proceeding by potential MVNOs suggest that, while there is demand for wholesale wireless services in Canada, there is an unwillingness on the part of MNOs to negotiate with MVNOs. For instance, in its initial submission to this Proceeding, Cogeco submits:
    • “Cogeco for one has met considerable resistance in its efforts to secure wholesale RAN [Radio Access Network] services over the years. Some of those unfruitful discussions were summarized in documents filed by Cogeco on September 4, 2014, in responses to undertaking no. 4, in the proceeding leading to the 2015 Wireless Decision. Additional discussions have been held in 2016, 2017 and 2018 between Cogeco and various wireless providers and have not led to the conclusion of any wholesale RAN arrangements.”Footnote 69
  7. In response to Q207 of the request for information (RFI) of 5 April 2019, Bell, Rogers and Telus collectively list almost 30 potential MVNOs that have approached them since August 2018, or with whom they are currently negotiating, to access their wireless network. Of those, none are reported to have resulted in an agreement and 13 of the negotiations are listed as ongoing. Given the short timeframe covered by the request, it is likely that this understates the number of MVNOs who would enter the market were there to be a willingness from carriers to engage with them.
  8. Further, the Bureau’s review of the few MVNOs that have been able to obtain an agreement in Canada found that their contracts are highly restrictive. [## ##].
  9. [## ##].Footnote 70
  10. The Bureau’s review found that current MVNO agreements can limit the ability of MVNOs to differentiate themselves by restricting the types of wireless plans they are able to offer.
  11. [## ##].
  12. [## ##].
  13. [## ##].Footnote 71
  14. [## ##].
  15. [## ##].Footnote 72
  16. It is therefore unsurprising that MVNOs in Canada have had a limited ability to attract subscribers.Footnote 73
  17. Apart from these limited MVNOs who are significantly restricted by their agreements in their ability to compete, the only sales at the wholesale level are roaming agreements which have been mandated by the CRTC and ISED in response to concerns of market power.Footnote 74

b. Conclusion

  1. In sum, the finding of retail market power also applies to the wholesale market given the highly vertically integrated nature of retail competition. Regardless, the only wholesale activity that has been achieved organically, and not through mandated regulatory intervention, relates to very few MVNOs who face highly restrictive agreements. It is therefore likely that there is market power at the wholesale level of the wireless market

VI. The Canadian wireless market is susceptible to coordination and evidence of it persists among the Big 3

  1. In its 5 July 2019 RFI, the CRTC asked whether any of the factors associated with coordinated behaviours are present in the Canadian wireless market.Footnote 74 The Bureau has found that the wireless industry generally, and in particular the Canadian wireless industry, is highly susceptible to coordinationFootnote 76 as the necessary conditions and facilitating factors for coordination are present. Wireless markets have features that make them susceptible to coordination generally; this is why coordination is prevalent in wireless markets internationally.
  2. The CRTC also asked whether there is evidence of coordinated behaviour in the Canadian wireless market.Footnote 77 The Bureau has found that it is likely that coordinated behaviour among the Big 3 exists. While there is some recent evidence suggesting a possible break or disruption of coordination, it is too early to say whether this is a durable and meaningful shift in the Big 3’s competitive behaviour. It is also unclear, given the timing of this activity, whether its cause is increased competition in the marketplace or the threat of regulatory intervention.

a. Coordination is prevalent in wireless markets internationally

  1. Carriers have argued that coordination is “extremely unlikely in the wireless industry.”Footnote 78 However, a review of wireless markets in OECD countries demonstrates the prevalence of coordinated behaviour in wireless markets. International regulators have found, or identified the potential for, coordination among MNOs within the markets for retail and wholesale wireless services in many jurisdictions.
  2. In 2002, the Netherlands Competition Authority (NMa) fined 5 wireless operators for coordination – a decision which was upheld in 2011.Footnote 79 In addition, in 2013, while the NMa could not find evidence of illegal price agreements, it did find that through unilateral public statements in the press or at conferences there was a risk that operators coordinated their behaviour.Footnote 80
  3. In 2005, Ireland’s Commission for Communications Regulation (ComReg) conducted a comprehensive review of the wireless market, and also identified a number of structural characteristics conducive to coordinated behaviour between telecom providers Vodafone and O2, most notably the structure of the market, the incentive to coordinate, the ability to coordinate, the ability to detect cheating, the enforceability of compliance, and actual and/or potential market constraints. ComReg found that Vodafone and O2 coordinated on both price and the denial of network access to independent service providers or MVNOs.Footnote 81 Following agreement from the European Commission (EC) on its finding of joint dominance, ComReg mandated that Vodafone and O2 open their networks to MVNOs.Footnote 82 Joint dominance is a term used to describe coordinated behaviour.Footnote 83
  4. Similarly, as part of their finding that Spain’s three MNOs held significant market power in the wireless market, the Comisión del Mercado de las Telecomunicaciones (CMT) found a number of structural characteristics conducive to coordinated behaviour, most notably the structure of the market, transparency of the wholesale market, the symmetry of the cost structure of the MNOs and lack of actual and/or potential market constraints.Footnote 84
  5. CMT found that, despite the evidence of service providers seeking wholesale access to MNOs’ networks, Telefónica, Vodafone and Amena had a tacit agreement to persistently refuse network access to new competitors in order to maintain high levels of profitability:
    • “The incentive to coordinate is in CMT’s view strong. Given the high levels of profitability already achieved in the mobile market by all three MNOs and the very significant barriers to entry, the latter have a strong incentive not to provide access to MVNOs that would threaten their profitable business model. By denying such access, they reserve to themselves the retail market at high levels of prices and profitability, without the threat of one or more new competitors.”Footnote 85
  6. In 2016, the EC determined that the proposed joint venture between Wind and Hutchison in Italy would significantly impede competition by enabling both coordinated and non-coordinated anti-competitive effects in the retail market. In a subsequent review of the joint venture remedy, it states:
    • “Indeed, the 2016 Transaction would have increased the likelihood that the three remaining MNOs would be able to coordinate their behaviour and raise prices in a sustainable way, even without entering into an agreement or resorting to a concerted practice within the meaning of Article 101 of the TFEU. In particular, the 2016 Transaction would have led to an overall alignment of incentives of the JV, TIM and Vodafone to coordinate their competitive behaviour. Reaching terms of coordination would have been possible and coordination would have been likely to be sustainable after the 2016 Transaction.”Footnote 86
  7. In 2008, the French competition authority found the three MNOs in their country were coordinating to refrain from offering wholesale access to MVNOs on favourable terms.Footnote 87 The incumbents successfully coordinated on restricting their product variety to avoid cannibalization. To stimulate competition in the mobile sector, the French regulatory authority granted the fourth 3G radio spectrum license to Free Mobile in 2010.Footnote 88 Within a few months prior to the anticipated entry, all three incumbents introduced subsidiary brands to provide low-cost services.Footnote 89 The price and quality levels of the subsidiary brands were closely matched to Free Mobile’s once it started offering its service to consumers.Footnote 90 A study of the French mobile market after this event found that the incumbents’ launch of the fighting brands was due to a breakdown in coordination spurred by entry.Footnote 91 Consumers gained considerably from the added variety of the new entrant and the fighting brands, and from the incumbents’ price response to the entry.
  8. Concerns related to coordination have also been raised in the US, where the US DOJ has characterized the wireless market as susceptible to coordination:
    • “Certain aspects of mobile wireless telecommunications services markets, including transparent pricing, little buyer side market power, and high barriers to entry and expansion, make them particularly conducive to coordination.”Footnote 92
  9. More recently, coordination concerns were also expressed in relation to the proposed merger between T-Mobile and Sprint alleging the merger would make it easier for the three remaining national facilities-based mobile wireless carriers to coordinate their pricing, promotions, and service offerings leading to “increased prices and less attractive service offerings for American consumers, who collectively would pay billions of dollars more each year for mobile wireless service.”Footnote 93
  10. There are several other competition agencies in OECD countries that have alleged MNO coordination, including agencies in the Czech RepublicFootnote 94, SwitzerlandFootnote 95, GermanyFootnote 96, MaltaFootnote 97 and SloveniaFootnote 98.
  11. That multiple OECD jurisdictions have found evidence of coordinated behaviour in wireless markets suggests that this industry has the necessary conditions and facilitating factors for coordination.

b. The Canadian wireless market is susceptible to coordination

  1. The Canadian wireless market possesses the hallmark features of a market that is susceptible to coordination. Coordination occurs when firms that repeatedly compete in the same market adopt a strategy to minimize competitive responses from other firms. They can develop a tacit understanding that each firm will adopt an accommodating reaction to their conduct. This happens when firms can signal – although not explicitly communicate – with each other through repeated interactions such as setting or announcing prices. This understanding facilitates maintaining higher prices.
  2. Put simply, in a competitive market, if a firm raises its price, it risks losing customers who will switch to their competitors. However, where there is coordination, a firm can raise its price with sufficient comfort that its competitors will follow suit. This reduces firms’ disincentive to increase price by reducing the associated risk of losing customers.
  3. If coordination is to be sustainable, the following features of the market are likely to be present:
    1. firms individually recognize mutually beneficial terms of coordination;
    2. firms monitor one another's conduct and can detect deviations (price reductions) from the terms of coordination;
    3. firms respond to any deviations from the terms of coordination through credible deterrent mechanisms; and
    4. coordination will not be threatened by external factors, such as the reactions of existing and potential competitors not part of the coordinating group of firm.Footnote 99
  4. In the Canadian wireless services market firms appear to recognize that coordination is mutually beneficial, for example, by describing the need for competitive restraint through “price discipline”Footnote 100, and referring to competitive activity as uneconomic during investor calls.Footnote 101
  5. Firms are able to communicate using pricing signals and public announcements and posted prices are available publicly and are closely monitored and quickly reacted to. Finally, the threat of retaliation from competitors in a Big 3 “home market” is a significant factor in carriers’ pricing decisions due to multi-market contact. This will be discussed further in the subsequent sections.
  6. There are also many more specific characteristics that can affect the sustainability of coordination. These factors make it easier for competitors to communicate effectively through signalling, determine whether a competitor has broken the coordinated agreement, and effectively punish those they are coordinating with when there is a deviation. Below is an analysis of these factors as they relate to the Canadian wireless market.

Concentrated market, few competitors

  1. Coordination is easier to sustain when there are fewer players in a market that can act as a disruptor, and the coordinated profits are shared amongst fewer firms. Irrespective of the measure of market concentration, Canada’s wireless market is highly concentrated with few competitors. As discussed earlier, the US DOJ characterizes markets in which the HHI is in excess of 2,500 points as highly concentrated.Footnote 102 The HHI is higher than 2,500 points in every Canadian province and territory ranging from 2660 in Quebec to 8158 in the Northwest Territories.Footnote 103 The four-firm concentration ratio is 100% in every region in CanadaFootnote 104 and the Big 3 possess 88.3% of the total number of subscribers nationally.

Market symmetry and stability

  1. More symmetric and stable market shares facilitate coordination. With asymmetry, the firm with the lowest market share has more to gain from a deviation, and less to lose from retaliation. Where shares are stable over time, it can reflect that the agreement is stable and make detecting deviations easier. As illustrated in Figure 7, market shares have been remarkably stable at the national level over the last decade and are symmetric with each of the Big 3 holding close to a third of subscribers. There is also symmetry in brand offerings, with each carrier having flanker brands (Virgin, Koodo, Fido) and discount brands (Chatr, Public Mobile, Lucky).

Figure 7 – National wireless market shares by subscribers, 2008-2018 Footnote 105

Figure 7 – National wireless market shares by subscribers, 2008-2018
Details of the National wireless market shares by subscribers, 2008-2018
Figure 7. National mobile wireless market shares by subscribers, 2008–2018
Firm 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018
Other 8% 5% 7% 9% 10% 10% 10% 10% 11% 10% 10%
Rogers 38% 37% 37% 36% 34% 34% 33% 33% 33% 33% 33%
Telus 27% 28% 27% 27% 28% 28% 28% 29% 28% 28% 28%
Bell 27% 30% 29% 28% 28% 28% 29% 28% 28% 29% 29%
Total 100% 100% 100% 100% 100% 100% 100% 100% 28% 29% 29%
Source: CRTC, Communications Monitoring Report 2008-2018, available at http://www.crtc.gc.ca/

Frequency of interaction

  1. Coordination is easier when firms interact more frequently since firms are able to more predictably interpret signals through repeated interaction and firms can react more quickly to a deviation. The Big 3 interact consistently across many product lines and geographies and have done so for decades.

Barriers to entry

  1. The threat of entry can impede sustainable coordination. However, barriers to entry are universally considered to be high in the wireless market. In its decision in CRTC 2015-177, the CRTC found that “barriers to entry into the retail market are very high.”Footnote 106 High capital expenditure investments are required, and entry is dependent on access to spectrum- a finite public resource, a large proportion of which has already been allocated in Canada. The Big 3 and incumbent regional carriers, have taken decades to construct their existing infrastructures, and thereby possess a considerable advantage over a new company attempting to establish a presence in the wireless industry. Even the new entrants in the wireless market have been building their networks over a period of 11 years since they first purchased spectrum in 2008.Footnote 107

Small transactions relative to market demand

  1. Coordination is more effective when firms can quickly react to a deviation. If the gains from undercutting are short lived, the incentive to undercut is reduced.Footnote 108 In industries with few customers and large contracts, which make up a large portion of overall market demand, coordination is more difficult to establish and sustain since each firm would want to cheat and win the large contract. This does not apply in the Canadian wireless market which is characterized by a large number of small-sized transactions. Individual customers are not large enough to incent the Big 3 to deviate since one customer’s cell phone bill accounts for a negligible amount of total market demand.

Network sharing

  1. The CRTC has also asked whether network sharing adds to the risk of coordinated behaviour.Footnote 109 In relation to network sharing agreements, the OECD suggests there is a risk for tacit collusion (another term used to describe coordination) between the MNOs on the quality of, or access to, their networks, and in doing so the risk of the establishment of collective dominance is increased:
    • “A market with four MNOs sharing two RANs with high barriers to entry would give rise to coordination concerns, as it is much easier to establish a focal point and to monitor and punish each other’s behaviour, and there is limited outside pressure. A particular concern could be the extent and quality of coverage, which would be subject to competition between sharing entities rather than (the greater number of) MNOs. Coverage is easy to monitor and any investment to improve coverage could be punished by the competitors matching that investment to neutralise any advantage…
    • ... [t]here is also a similar potential concern that such agreements could increase the risk of industry-wide co-ordination to deny access to sites or RAN to new entrants. The reduction in the number of RANs might make such coordination easier and if it did occur there would be a risk of significant harm to competition and for consumers. The likelihood of this issue arising is subject to the same drivers as of potential unilateral incentives to deny access.”Footnote 110

Evidence of past coordination

  1. A history of coordination in the market is relevant because previous and sustained coordinated behaviour indicates that firms have successfully overcome the hurdles to effective coordinated behaviour in the past.Footnote 111 The Bureau has previously concluded that the Big 3 were coordinating in markets where they did not face sufficient competition. This finding was based on an extensive review of a large number of internal company documents from the Big 3 and a detailed pricing analysis using company data.Footnote 112
  2. The Big 3 typically cite opaque pricing, innovation, and increasing demand as preventing coordination from being established; however, the three sections below explain why this characterization of the market is not well founded.

Price transparency

  1. Coordination can be difficult to sustain when individual prices are not readily observable and cannot be easily inferred from readily available market data. In the absence of market transparency it is difficult to monitor and detect deviations from coordination. Transparent pricing and offers can also be used as a communication mechanism. As will be elaborated upon further in the subsequent section addressing price signalling, prices in the Canadian wireless market are transparent – pricing is publicly available and firms closely monitor one another’s pricing and plan offers by interacting with competitor web sites, reviewing price announcements, and using “mystery shoppers.”

Innovation

  1. Coordination can be more difficult in innovative markets. The reason being if there is a significant risk that one’s competitor will introduce a drastic innovation which will result in a significant erosion to one’s market share, then carriers would anticipate that their market position is short-lived so they would have the incentive to deviate from the coordinated agreement to profit in the short term. Essentially in innovative markets, the threat of future retaliation may not be credible, or the profit penalties may be decreased.
  2. However, this factor is not applicable in the wireless market where innovation is often externally driven by equipment manufacturers and standard-setting organizations. Innovations in this market are largely not dictated by the carriers and can be characterized as predictable innovations. Handset manufacturers introduce new features and functionalities with respect to devices which are available to each of the Big 3. Networks are developed according to global standards which do not originate in Canada. Wireless providers like the Big 3 contribute to the research and development, but do not drive it, and there is advanced notice and long lead times associated with innovations. Therefore, generally no significant advantage is gained by a particular national carrier through innovation, as they have advance warning and equal access to innovative new products and network technologies.

Demand growth

  1. Demand for wireless services is growing, with data usage being the driver of that growth. Markets in a period of rapid growth can be less conducive to coordinated behaviour. This is because entry may be more likely in growing markets and the prospect of future entry then hinders the ability to coordinate. However, in markets in a period of growth where entry barriers are high, market growth may actually facilitate coordination, as it is easier to sustain coordination when short-term gains from a deviation are small relative to the costs of future retaliation. This implies that for a fixed number of market participants, coordination is actually easier to sustain in growing markets since today’s profits are small compared with tomorrow’s profits.

Big 3 can signal their pricing intentions and monitor one another’s conduct to detect deviations

  1. The Big 3 appear to communicate using publicly posted prices with pre-specified end dates to signal what they intend to do with their prices in the near future. The Bureau has identified several ways in which it appears the Big 3 communicate their pricing intentions with one another.
  2. First is through forward looking public announcements regarding prices and promotional pricing with pre-specified end dates. For example, the illustrative timeline below shows a series of pricing announcements by the carriers resulting in a $5 increase in wireless prices on all plans from Rogers, Bell and Telus.

Figure 8 – Timeline of pricing announcements related to $5 increase in plan prices

Figure 8 – Timeline of pricing announcements related to $5 increase in plan prices
Details of Timeline of pricing announcements related to $5 increase in plan prices
  1. Telus announces a price increase for March 1st.
  2. Bell announces a price increase for March 12.
  3. Rogers announces a revised end date for their promotion of March 12.
  4. Telus revises its price increase for March 12.
  5. Bell imposes a $5 price increase on all plans on March 12.
  6. Rogers imposes a $5 price increase on all plans on March 12.
  7. Telus imposes a $5 price increase on all plans on March 12.
  8. Telus revises their price increase announcement for March 14.
  9. Rogers announces a price cut that will end on March 31.

  1. One carrier first signals their pricing intention, consisting of the price change and the date it will occur, and then waits to see whether or not the other carriers respond. The other carriers may respond by proposing differing changes to their own pricing or timing of the price change and then see whether the other carriers follow their proposed change. Finally, they implement the price change. This signalling is completed with no risk to the carriers as these prices are not yet available for sale, they are only to be implemented at a future date.
  2. As a result, the Big 3’s pricing and product suite ends up mirroring one another’s quite closely over time. For example, the figures below illustrate price movements on three plans, demonstrating how the carriers’ offerings mirror one another in terms of price as well as when plans with particular data buckets are available. Notably, the gaps in the graphs indicate that a plan of that data bucket was not available during that period.

Figure 9 – Price for a 4GB plan by service providerFootnote 113

Figure 9 – Price for a 4GB plan by service provider
Details of Price for a 4GB plan by service provider
4GB October 2017 November 2017 December 2017 January 2018 February 2018 March 2018 April 2018 May 2018 June 2018 July 2018 August 2018 September 2018 October 2018 November 2018 December 2018 January 2019 February 2019 March 2019 April 2019 May 2019 June 2019 July 2019
Bell $100 $100   $85 $105     $90 $90 $90 $90 $90 $90 $90     $85          
Rogers $100 $100 $100 $85 $105     $90 $90 $90 $90 $90 $90 $90                
Telus $95 $95 $95 $85 $100 $105   $90 $90 $90 $90 $90 $90 $90     $90          

Figure 10 – Price for an 8GB plan by service providerFootnote 114

Figure 10 – Price for an 8GB plan by service provider
Details of Price for an 8GB plan by service provider
8GB October 2017 November 2017 December 2017 January 2018 February 2018 March 2018 April 2018 May 2018 June 2018 July 2018 August 2018 September 2018 October 2018 November 2018 December 2018 January 2019 February 2019 March 2019 April 2019 May 2019 June 2019 July 2019
Bell         $110   $120 $120         $105 $105 $105         $100    
Rogers         $110   $120 $120     $105 $105 $105 $105 $105         $100    
Telus         $110   $120 $120     $105 $105 $105 $105 $105         $100    

Figure 11 – Price for a 10GB plan by service providerFootnote 115

Figure 11 – Price for a 10GB plan by service provider
Details of Price for a 10GB plan by service provider
10GB October 2017 November 2017 December 2017 January 2018 February 2018 March 2018 April 2018 May 2018 June 2018 July 2018 August 2018 September 2018 October 2018 November 2018 December 2018 January 2019 February 2019 March 2019 April 2019 May 2019 June 2019 July 2019
Virgin $105 $115   $115   $115 $115 $115                            
Fido $105 $115 $115 $115   $115 $115 $115                            
Koodo $105 $105 $115     $115         $90                 $85    

  1. In addition to using forward-looking public announcements to signal price changes, the Big 3 speak about pricing activity and their future pricing intentions in investor calls signalling their pricing intentions for the subsequent quarter. This is demonstrated in the below excerpt from a 2014 Rogers investor call.Footnote 116
    • Glen Campbell - BofA Merrill Lynch - Analyst
      Just a quick follow-up on that. I mean the pricing my sense was fairly aggressive through most of the quarter. And now sort of late in the quarter and into Q2 it is much, much more disciplined so would it be fair to say that you are making the trade-off for ARPU at the expense of sub growth for the balance of the year?
    • Guy Laurence - Rogers Communications Inc. - President & CEO
      Glen, it's Guy, if I may answer that. I think that is true to say. What I would say is that Rob and I have decided to change tactics slightly. I'm not saying it is a structural change, but we are experimenting with changing tactics. Taking some of the noise out of the market wasn't necessarily caused by us in the first place by the way, but never the less we have reduced the amount of promotional activity and we are continuing to monitor that to see how it plays in the marketplace more widely and how it obviously affects our own trajectory going forward.
  2. Similar statements were recently made after Bell did not match Telus and Rogers reductions in device subsidies in a 2019 Telus investor call.Footnote 117
    • Maher Yaghi - Quantitative Analyst at Desjardins
      And just to follow that is, do you believe that the industry can sustain the healthy profits or the margins that they have right now if these unlimited plans which create pressure on ARPU are not offset by the reduction in subsidies as one of your competitors continue to offer subsidies on these plans?
    • Darren Entwistle – Telus Communications President & CEO
      I think the industry in terms of that last point will discipline itself and in that regard. Secondly, if that happens, I don't think it's going to be sustainable. I think the pressure will drive the right economic decisions within the industry.
  3. A potential strategy among the Big 3 of using selective price cuts to enforce pricing discipline was noted in an April 2015 report in Canadian Telecom and Cable Weekly entitled “Flanker Brand pricing: Enforcing the discipline.” The article noted new pricing on Bell’s Virgin brand, which matched an “aggressive promo” by Koodo, which itself was launched shortly after Fido’s “relaunch” with reduced pricing. The article notes that “the good news in this situation is that we believe the Telus’ motivation with the Koodo promotion is to push Rogers to reverse the price cut on its 1GB/2GB plans (i.e., an effort to enforce discipline), not to grab market share.”Footnote 118
  4. These types of forward looking pricing statements at investor relations events have been used as a coordination mechanism in other jurisdictions including wireless markets in ItalyFootnote 119 and the Netherlands.Footnote 120

Multi-market contact and the establishment of “home markets”

  1. It is well recognised that firms can sustain coordination more easily when they compete against one another in several markets.Footnote 121 This concept is referred to in the economic literature as “multi-market contact”. Multi market contact can lead to coordination when a firm chooses not to compete aggressively in one market, for fear of their competitor’s retaliation in other markets. Multi market contact among the Big 3 is significant, and encompasses a number of geographies and business lines at both the wholesale and retail level.
  2. Bell, Telus and Rogers have “home markets” where they have historically held the lion’s share of the market. These are typically also where they have extensive wireline infrastructure. When the Big 3 compete against one another, they are putting their own base of subscribers in their “home market” at risk since their competitor could retaliate by driving down prices where they have their base of subscribers. Instead, they choose to compete aggressively against one another, and compete more aggressively against strong regional carriers since those carriers cannot retaliate in the Big 3’s “core” markets.
  3. This concept was well explained by a telecom analyst in describing MTS’ strategic position prior to their acquisition by Bell.
    • MTS remains in an unenviable strategic position – While new entrants have not built in Manitoba, what MTS has [is] far worse - 3 strong national incumbent competitors. In addition 1) national incumbents can price aggressively in Manitoba knowing full well that MTS cannot retaliate in the national incumbents’ core markets in Toronto, Montreal, Calgary, Vancouver etc.”Footnote 122
  4. Information collected during the Bell/MTS inquiry supported the likelihood that Bell, Rogers and Telus weigh the advantages from vigorous competition in one area against the danger of retaliation in their “home market”.Footnote 123 If, however, firms with multi market exposure encounter a firm that does not possess these same incentives in a particular market, coordination can break down in that space. The regional carrier does not have an ability to retaliate in the home markets of the Big 3, so the Big 3 do not open themselves up to risk by competing against a regional carrier to steal share. Regional carriers, therefore, can play a disruptive role in the market by spurring the Big 3 to compete vigorously where they are present. Ultimately, this multi market exposure softens competition among the Big 3 through coordination.

c. Concentration and network quality do not explain regional price differences

  1. Differences in market concentration and network quality do not consistently explain regional price differences, while the presence of a wireless disruptor does.
  2. As explained previously, HHI is a commonly accepted measure of market concentration. The HHI increases both as the number of firms in the market decreases and as the disparity in size between those firms increases. The higher the HHI, the more concentrated the market.
  3. Under unilateral competition, economic theory predicts that areas with a lower HHI are more competitive and would therefore have lower prices. However, as Figure 12 illustrates, there does not appear to be a relationship between HHI and average prices. For example, prices in Saskatchewan are significantly lower than prices in PEI, despite Saskatchewan having a higher HHI index.

Figure 12 – Price per GB-HHI correlation, 2018 Footnote 124

[## ##]

  1. The alternative explanation typically cited by the national carriers for the lack of correlation between concentration and pricing is that variation in network quality is what explains price differences.
  2. For example, the Bell commissioned report by CRA for this Proceeding finds that:
    • The regressions using 2018 data generate similar results to those found using 2016 data. Plan characteristics, data allowances, and provincial population characteristics explain much of the observed variation in 2018 wireless plan prices across provinces. Network quality measures are important explanations for differences in wireless plan prices that are not due to plan characteristics, data allowances, and population characteristics. As we found before, the relationship between network quality and wireless plan prices is non-linear; hence, a flexible approach to the estimation is undertaken. Using this flexible approach and the resulting coefficients on the network quality variables, we find that, for example, most of the difference in wireless plan prices between Saskatchewan and Ontario in 2018 is explained by network quality, plan characteristics, data allowances, and population characteristics.Footnote 125
  3. However, the effect of Average Download Speed (in logs) on price in 2018 follows an inverted U pattern (see CRA report Figure 1A, page 39). In effect, this means that faster average download speeds sometimes explain higher prices but also sometimes explain lower prices. This highlights that the question of interest is not merely explaining variation in prices but studying the impact of competition on outcomes including price and quality. To that end, the Matrix Study’s analysis of the confidential and public data found no effect from the presence of regional competitors on network quality indicators for the Big 3. This suggests that pricing differences correlated with regional carrier presence are not similarly linked to variations in network quality.
  4. Instead of quality differences, what does appear to consistently explain pricing is the existence of a regional competitor who is able to disrupt coordination between the national carriers as evidenced by their affect on pricing.

d. Recent pricing activity suggestive of possible reprieve in coordination but timing is suspect

  1. The Matrix Study demonstrates that prices in regions with a wireless disruptor are lower. However, the data underlying its study also shows that prices in regions with and without a strong regional competitor are converging.
  2. Figure 13 illustrates that prices in regions with incumbent regional carriers, such as SaskTel, have decreased at a slower rate than prices in regions with regional entrants, such as Videotron. As such, while the prices in regions without incumbent regional carriers started at higher price levels, they are declining more quickly than prices in regions with incumbent regional carriers.

Figure 13 - $/GB by province, 2014-2018 Footnote 126

[## ##]

  1. In addition, recent events indicate a possible break from coordinated behaviour between the Big 3. As previously discussed, the Big 3 can use forward looking public announcements to signal their pricing intentions and monitor one another’s conduct to detect deviations.
  2. Figure 14 illustrates recent pricing announcements related to the introduction of Rogers’ “Infinite” plans. It shows that, when Rogers announced the launch of their “Infinite” plans, Bell and Telus announced similar plans but only on a promotional basis with a specified end date. Rogers did not respond to the promotional price signal and maintained their plans on a permanent basis, resulting in Bell and Telus deciding to subsequently follow suit.

Figure 14 – Timeline of announcements related to unlimited data plans

Figure 14 – Timeline of announcements related to unlimited data plans
Details of the Timeline of announcements related to unlimited data plans
  1. Bell matches Rogers plan on promotional basis until June 30. - June 12, 2019
  2. Rogers announces "unlimited" Infinite plans for 10GB, 20GB, and 50GB. - June 12, 2019
  3. Telus announces a 10GB plan with bonus 5GB promotional offer until July 2. - June 12, 2019
  4. Telus matches Rogers 10GB and 20GB "unlimited" plans as promotion until July 2. - June 18, 2019
  5. Bell confirms it will keep its "unlimited" plans in the market. - July 3, 2019
  6. Telus makes 10GB "unlimited" plan permanent. - July 3, 2019
  7. Rogers introduces "unlimited" plans that span both Canada and U.S. - July 10, 2019
  8. Rogers announces 365K customers are on 10GB+ "unlimited" plans and are using 50% more data. - July 23, 2019
  9. Telus reduces price of 10GB "unlimited" plan in Saskatchewan, Manitoba and Quebec. - July 25, 2019
  10. Bell drops "unlimited" plan prices in Saskatchewan, Manitoba and Quebec. - August 14, 2019
  11. Rogers announces 750K customers are on its "unlimited" data plans. - September 12, 2019

  1. While there do appear to be some signs of a reprieve in coordination, the Bureau cautions that the timing of these events makes it difficult to know whether this competitive activity is attributable to a competitive disruption of coordination or if this action is a response to something else, such as the current regulatory scrutiny and proposed MVNO policy. As the CRTC has described, “the mere possibility of competition does not provide sufficient grounds to forbear.”Footnote 127

VII. Conclusion

  1. The findings of Part 1 provide the CRTC with an understanding of the scope and extent of the market power in the Canadian wireless market. This is characterised by high market shares, high barriers to entry and high levels of profitability. Whether or not effects are due to a unilateral or coordinated exercise of market power, the Matrix Study provides direct evidence of market power. It shows that the Big 3 have the ability to raise prices above competitive levels in most regions in Canada. As a result, many Canadians are paying higher prices for wireless services and using less data.
  2. However, the Bureau also found that competition brought about by regional carriers is showing promising signs of delivering lower prices. Where regional carriers have achieved sufficient scale, Canadians are paying significantly less for their wireless services and consuming more data.
  3. With this knowledge, in Part 2 of this report the Bureau will outline considerations for designing a focused remedy to stimulate competition while preserving investment incentives.

Part 2: An assessment of possible remedies to promote competition

  1. In this section, the Bureau first describes the lessons learned from a scan of international jurisdictions including those with a mandated MVNO policy. With an understanding of both the competitive dynamics in Canada and lessons from abroad, the Bureau then examines remedy options available to the CRTC and discusses the merits of each option. Finally, the Bureau recommends a targeted remedy that it believes is most likely to be effective in stimulating competition in the short run while maintaining the viability of wireless disruptors and ensuring investment in high quality networks will continue.

VIII. International experience with mandating MVNO access

  1. The Bureau’s international study focuses on jurisdictions that have adopted mandated MVNO access. The international experience with mandating MVNO access can help inform Canada’s path forward. While international comparisons are rarely apples to apples, and involve some limitations, they demonstrate the different ways in which MVNO access can be implemented, and the degree to which each model was successful in fostering competition and lowering prices.
  2. The conclusions herein are based on a review of publicly available information as well as interviews with various industry stakeholders in the countries studied, including telecommunications regulators.
  3. This study revealed that in cases where MVNO access has been mandated, regulators have mostly opted to incentivize carriers to commercially negotiate rates rather than set a wholesale rate or reserving network capacity.
  4. MVNOs typically capture a share of less than 10% of subscribers in the countries where they operate and tend to target “niche” markets. However, despite a relatively low market share, MVNOs have placed some competitive pressure on MNOs.
  5. While in theory mandated MVNO access may contribute to decreased investment incentives, several examples exist of jurisdictions where MVNO access has been mandated and no significant decrease in investment has occurred.
  6. The Bureau acknowledges the limitations associated with international comparison studies of mandated MVNO access, including those associated with this study. In particular:
    1. there are few countries that have introduced mandated MVNO access and therefore a limited sample size from which to draw inferences;
    2. among the countries with mandated MVNO access there is little information on outcomes like pricing available over the necessary time period;
    3. outcomes reflect the impact of numerous concurrent and consecutive policies and market events, so strong causal effects related to mandated MVNO access are seldom solely attributable to one particular policy such as MVNO access; and
    4. there are significant differences between the countries studied in terms of population density, land mass and other related features that may mean policies that were effective elsewhere may not always be appropriate in Canada.
  7. The international findings herein are informative, but not definitive in answering the policy questions posed by the CRTC in this Proceeding. One must keep in mind the specific characteristics of the Canadian market in applying these findings.

  1. Mandated MVNO access has been imposed differently across jurisdictions and very rarely involves setting a wholesale rate. There are two ways through which mandated access has been imposed internationally: as a merger remedy or through regulatory intervention.

Merger remedies

  1. The intended outcome of mandated MVNO access as a merger remedy is typically to prevent merger-induced anticompetitive effects, such as price increases, rather than to place downward pricing pressure on the market. This outcome has been achieved with mixed success.
  2. The EC has imposed mandated MVNO access as a merger remedy in Austria (2012), Germany (2014), Ireland (2014) and Italy (2016). These cases are discussed below. In all cases, MVNO access was imposed only on the acquiring firm and wholesale rates were set based on commercial negotiations.
  3. In Austria, mandated access was imposed on the acquiring firm, Hutchison. As part of the merger condition, Hutchison was required to enter into an upfront MVNO agreement subject to EC approval; this initial agreement was then used as a ‘reference offer’ to establish the maximum amount Hutchison could charge other MVNOs. Hutchison was required to grant access to up to 16 MVNOs for up to 30% of its network capacity for 10 years.Footnote 128
  4. In Germany, before it could acquire E-Plus, Telefónica was required to sell up to 30% of its network capacity at a fixed price.Footnote 129 Unlike the Austrian model in which MVNOs face a marginal cost for each customer addition,Footnote 130 the German model eliminates the marginal cost associated with adding a customer in order to incentivize MVNOs to maximize their network utilization.Footnote 131 Drillisch, a small carrier with no network, agreed to buy 20% of Telefónica’s network capacity.Footnote 132
  5. In 2014, Irish MVNOs had a combined market share of 7.4%.Footnote 133 Following a 4-to-3 merger, the acquiring firm Three Ireland was required to sell up to 30% of its network capacity to two MVNOs (15% each) as a condition of its merger with O2.
  6. A study by the Body of European Regulators for Electronic Communications (BEREC) of these merger remedies found that in all three cases (Austria, Germany and Ireland) there is at least some evidence that retail prices for new customers increased due to the merger compared to the situation without the merger, at least in the short run. In Austria, the merger led to significant price increases in 2014 and 2015, after which it appears that competitive pressure from MVNOs resulted in price decreases from elevated pricing levels. Price increases also followed the mergers in Germany and Ireland, but as available data only extended for a year and a half following the merger, only short to medium run impacts were estimated.Footnote 134
  7. In 2016, the EC determined that the proposed joint venture between Wind and Hutchison in Italy would significantly impede competition by enabling both coordinated and non-coordinated anti-competitive effects in the retail market and eliminating competitive constraints in the wholesale market.Footnote 135 The EC took a different approach in remedying this merger. As part of the remedy, the merged entity, Wind Tre, was required to enter into a transitional roaming agreement with the new entrant, Iliad, until it transitioned to an MNO. Iliad was also provided with assets like spectrum to assist it in transitioning to an MNO.Footnote 136 The merger remedy appears to have been successful. Iliad has shown promise by capturing 4% of “human” SIMs (3.2% of total SIMs) in under a year while still in its transitional phase - market share which was captured from the 3 national incumbents.Footnote 137 In 2018 the EC approved Hutchison taking full control of the joint venture provided it continued to meet the obligations of the remedy. The reasons for EC approval included the effectiveness of the 2016 remedy, wherein prices continued to decline and competition among mobile operators intensified.Footnote 138
  8. Similarly, as part of the remedy for the merger of T-Mobile and Sprint in the U.S. earlier this year, in addition to divesting prepaid assets, spectrum and retail and cell cites, T-Mobile/Sprint must provide the new entrant DISH full MVNO access based on commercially reasonable terms for a period of 7 years while DISH builds out its network.Footnote 139 As part of this agreement, DISH has committed to deploying a nationwide 5G network that covers at least 70% of the U.S. population by 2023, or face a fine of $2.2 billion.Footnote 140 The remedy remains somewhat uncertain given that a number of states filed a lawsuit to stop the merger, despite the remedy.Footnote 141
  9. While there are very few instances to study, the Italian remedy, in which temporary MVNO access was granted to allow a firm to become an MNO, appears to have been most successful by contributing to both a decrease in prices and the entry of a facilities-based competitor. On the other hand, results from instances of pure MVNO remedies with no requirement to transition to an MNO have been mixed. The Austrian remedy appears to have been successful in driving down prices; the German and Irish remedies have not.

Regulatory interventions

  1. In contrast to merger remedies, which typically aim to address a likely lessening of competition, regulatory interventions are designed to increase competition in the market by, among other things, expanding service offerings and generating downward pricing pressure on the market. Mandated MVNO access by way of regulatory intervention has been imposed through obligations on spectrum licensing, regulatory backstops and wholesale rate regulation.
  2. In Hong Kong (2001), Ireland (2002) and the Czech Republic (2013), MVNO access has been imposed as a condition of spectrum licensing.
  3. In Ireland, MNOs were required to offer MVNOs a minimum 35% discount off of retail rates and the regulator could intervene if agreements could not be reached within three months of the request.Footnote 142
  4. In Hong Kong and the Czech Republic, wholesale rates were set through commercial negotiations. In Hong Kong, the spectrum condition mandating MVNO access was never relied on and has not been used in spectrum auctions since the 3G auction in 2001.Footnote 143 In the Czech Republic, MNOs are still required to offer MVNO access based on fair and reasonable terms as a condition of their spectrum licenses.Footnote 144 Under the Czech Republic’s Electronic Communications Act, an MNO may lose the right to use assigned frequencies if it fails to comply with conditions of the spectrum auction.Footnote 145
  5. Instances of regulatory intervention involving wholesale rate regulation are very rare, with most regulators opting for a ‘lighter touch’ approach by mandating fair and reasonable commercial negotiations, with a regulatory backstop in place in case negotiations fail.
  6. In Spain, MVNO access was mandated in 2006 due to what the Comisión del Mercado de las Telecomunicaciones (CMT) observed as persistent excess profits and collective dominance in denying access to MVNOs on the part of the country’s three MNOs which were found to jointly hold significant market power in the wholesale market for access and call origination to the mobile networks.Footnote 146
  7. CMT opted for light touch regulation, requiring the three MNOs to meet reasonable requests for access and set reasonable access prices. Following the decision, MNOs were able to reach agreements that best fit their business models, with CMT intervention occurring only in cases of disputes where negotiations fail.Footnote 147
  8. Mandated MVNO access was deregulated in 2017 as the Spanish competition authority Comisión Nacional de los Mercados y de la Competencia (CNMC),Footnote 148 found the market had achieved “effective competition”.Footnote 149 However, if operators unreasonably removed access or sharply raised wholesale rates, CNMC had the authority to intervene under Spanish competition law.Footnote 150
  9. In 2006, the Norwegian Communications Authority (Nkom) determined that Norway’s largest wireless carrier Telenor, with a subscriber share well above 50%, held significant market power. In an effort to encourage the development of a third facilities-based carrier and promote retail competition, Nkom imposed mandated MVNO access and national roaming on Telenor. The access obligation was based on the ladder of investment theory, which aims at promoting incentives to invest by reducing barriers to entry.Footnote 151 Nkom considered full MVNOs, meaning an MVNO that has its own core network, as an important step to in becoming a facilities-based carrier.
  10. In 2016, Nkom extended the access obligation to also include light MVNOs, in Norway referred to as service providers or resellers, which do not have their own core network. Prior to the extension of the access obligation, the new entrant Tele2 had lost its spectrum and was bought by Norway’s second carrier Telia.Footnote 152 While Nkom initially concluded Telenor was only required to provide MVNO access on non-discriminatory terms, Nkom imposed price controls on Telenor in the form of a margin squeeze test in 2016.Footnote 153
  11. Both Japan (2002) and Israel (2009) also required MNOs to offer access based on fair and reasonably negotiated terms, with a regulatory backstop in the form of arbitration if an agreement could not be reached.Footnote 154 In Japan, MNOs typically offer MVNOs access to their networks at the same rate as the regulated interconnection rate, which is set at reasonable costs plus reasonable profits.Footnote 155 However, there is currently some question in Japan as to the fairness of the interconnection rate and Japan’s Ministry of Internal Affairs and Communications has launched a review of the interconnection rate calculation.Footnote 156
  12. In 2013, the Peruvian telecommunications regulator, OSIPTEL, issued a working paper advising that, given the impending launch of a fourth MNO in the Peruvian market, the Peruvian congress should not mandate MVNO access at that time.Footnote 157 However, later that year, against the advice of OSPITEL, the Peruvian congress passed a law mandating MNOs to provide access to MVNOs on the basis of commercial negotiations; should commercial negotiations fail after 60 days OSPITEL will step in to set the conditions.Footnote 158
  13. The Bureau is aware of other jurisdictions in which MVNO access has been mandated. For example, South Korea imposed mandated MVNO access in 2010 wherein wholesale prices for MVNOs are on a retail-minus basis;Footnote 159 however, whether MVNOs use this provision is unknown to the Bureau. Similarly, in Colombia, while the Bureau is unaware of the specifics of the policy wherein MVNO access was mandated in 2015, if no agreement is achieved after a set period of time the regulator has established a calculation methodology to determine prices in order to establish price predictability.Footnote 160
  14. All in all, the preferred approach internationally in designing mandated MVNO access appears to be a regulatory “back stop” where commercial negotiations fail. Were the CRTC to adopt such an approach, a more focused review of the relative strengths and weaknesses in the policy design of these remedies could be a useful guide.

b. Mandated MVNO access is often accompanied by other regulatory measures

  1. Where MVNO access has been mandated, it is often accompanied by other regulatory measures to facilitate entry by decreasing barriers to entry and switching costs. This can make it difficult to assess the success of mandated access regulation in isolation.
  2. Perhaps the most sweeping regulatory measures aimed at supporting competition in the wireless market have been in Israel. In addition to mandating MVNO access in 2009, in 2011 the Israeli Ministry of Communications (MOC) issued spectrum licenses for infrastructure-based operators wherein the company would be completely repaid for the cost of the spectrum if they captured 7% of the market within 5 years. Coverage requirements were included with gradual threshold increases each year. In addition, license rebates were offered corresponding to market share gained. These new operators were granted the right to use national roaming facilities of existing operators for a limited period.Footnote 161
  3. Two licence winners — Golan and Hot mobile (who both began operating in 2012) — met those thresholds and were repaid for the cost of the spectrum they acquired. However, while the repayment incentive associated with capturing market share may have been successful in driving down prices,Footnote 162 it may have also caused one of the entrants, Golan, to price under operating cost in order to meet the share threshold for repayment. Golan also did not meet the network build requirements relying solely on their temporary mandated access and was sold in 2017.Footnote 163 Hot mobile and Golan (under new ownership) both continue to operate in the Israeli market.
  4. In addition to the spectrum incentives, the MOC had previously significantly lowered roaming rates and decoupled the cost of the device from wireless plans to allow customers to more easily switch their service plan to a new provider while continuing to pay off the cost of their device with their previous provider.Footnote 164
  5. Colombia also banned linking equipment acquisition to mobile service provision in 2014 and later this year Japan’s Ministry of Internal Affairs and Communications will introduce a new rule to ban operators providing a device subsidy to delink the purchase of equipment and mobile service provision.Footnote 165 In addition, Colombia banned permanence clausesFootnote 166, which have been found to lower product variety and increase switching costs,Footnote 167 instead allowing customers to pay off their phones in installments, independent of their plan. Colombia’s Comisión de Regulación de Comunicaciones (CRC) has stated that banning permanence clauses has allowed a larger number of stores, supermarkets and other retail outlets to operate in the MVNO business, and has also spurred international mobile firms to offer more affordable smartphones.Footnote 168
  6. In Germany, Austria and Ireland, where MVNO access was mandated as part of a merger remedy, mandated MVNO access was accompanied by structural remedies in the form of spectrum divestments or spectrum set asides in the next spectrum auction. In all of these cases, however, the spectrum divestments or set asides required of the acquiring companies were never exercised as no new entrant acquired the spectrum.Footnote 169
  7. As part of the merger remedy in Italy, Wind Tre was required to divest spectrum to Iliad and provide it with access to sites through divestiture or co-location.Footnote 170 At the same time Norway mandated MVNO access in 2006, it also mandated national roaming in order to stimulate the development of a third facilities-based carrier.
  8. In Hong Kong, MNOs are subject to spectrum utilization requirements to ensure they maximize their network capacity and to prevent spectrum hoarding. This may explain why mandated access conditions on 3G spectrum licensing were never relied on for MVNO agreements, as carriers already had an incentive to fill their network.Footnote 171
  9. At the same time as Peru introduced regulation aimed at facilitating the entry of MVNOs, they also reduced mobile termination ratesFootnote 172 and regulated interconnection fees. From 2010-2014 interconnection fees for new entrants were set at a higher rate than the rates set for incumbent carriers in order to incentivise investment by entrants.Footnote 173 MVNOs in Peru are also required to contribute to the country’s telecommunications investment fund aimed at expanding access across the country.Footnote 174
  10. Mandated MVNO access internationally has been accompanied by a number of policies aimed at spurring competition or incentivizing infrastructure investment. While providing insight into a wide array of policies, this can make drawing specific causal inferences about the success of a particular policy difficult, given the multitude of policies introduced. The Bureau therefore cautions that the CRTC should carefully assess any definitive conclusions drawn by interveners based on an analysis of a single country, or small number of countries.

c. MVNOs tend to target “niche” segments of the market

  1. The Bureau found that MVNOs tend to target “niche” segments in the markets they serve. In some cases these are small segments that the MNOs choose not to cater to, for example, business customers or customers needing to make a high-volume of long distance calls. In other cases, these are larger segments which the MNOs may choose to cater to through a flanker brand, such as value-conscious customers.
  2. The NZCC Study found that MVNOs cater to eight key segments. The study of nearly 1,500 MVNOs globally found that “discount” (22%), “retail” (17%), and “ethnic” (12%) are the most prevalent segments (Figure 15).
  3. “Discount” MVNOs rely on low prices as a scaling opportunity to minimize distribution costs, while ”retail” MVNOs are commonly owned by supermarket chains with strong brand strength and are typically leveraged as part of a loyalty program. “Ethnic” MVNOs target immigrant communities and often offer low-cost international roaming to those customers’ country of origin, similar to international and roaming MVNOs.Footnote 175

Figure 15 – MVNOs by segment type Footnote 176

Figure 15 – MVNOs by segment type
Details of the MVNOs by segment type
Discount 22%
Retail 17%
Ethnic 12%
Business 11%
Specialist data 11%
International roaming 8%
Youth/media 5%
Bundled 4%
Other 10%

  1. The Bureau also found a number of mischaracterizations of the market segments targeted by MVNOs in the initial interventions of this Proceeding.
  2. First, the expert report of Richard Feasey, prepared for Rogers, suggests that MVNOs rarely focus on value-conscious consumers and typically focus on more profitable segments of the market. It states:
    • “The CRTC should not assume that mandating wholesale access for MVNOs will meet the needs of low income groups. The evidence from Spain, Norway and other markets I have researched is that the majority of MVNOs focus their attention on profitable customer groups such as business users and frequent international callers. Only a relatively small proportion of MVNOs in the rest of the world are ‘low cost’ providers of the kind which the CRTC wants to promote.”Footnote 177
  3. However, this statement does not align with the NZCC Study of over 1,500 MVNOs worldwide that shows that the largest proportion (22%) target discount customer segments.
  4. Second, in reference to Nkom’s 2016 decision to mandate MVNO access for service-based providers, Mr. Feasey writes:
    • “This noted, two features of the Norwegian experience are, I think, of interest. First, the Norwegian regulator considers that the benefits of having MVNOs in the Norwegian market are expected to accrue principally to business users, which reflects the fact that MVNOs have acquired a significant share of this market segment in Norway. In contrast, competition from MVNOs in the consumer segment is very modest, and lower than in many other European countries in which there has been no wholesale regulation. No mention is made of low income consumers in the Norwegian regulator’s analysis and there appears to be no expectation that its regulatory measures will affect that group.”Footnote 178
  5. Mr. Feasey’s statement confuses Nkom’s terminology. What Nkom refers to as MVNOs, which own their own core network and for which access was mandated in 2006, tend to focus on the business market. However, what Nkom refers to as service providers (also known as resellers), which do not have their own core network and for which access was mandated in 2016, do in fact focus on value-conscious consumers. For instance, independent service providers Telipol, Chili Mobil and Sponz offer the most competitive price for a 5 GB plan in Norway.Footnote 179
  6. Nkom’s 2016 decision to mandate service provider access states that the intended goal of its remedy includes improving affordability:
    • “The choice of remedies is largely based on Principle 3 in Nkom's remedies document, i.e. that the remedies must, to the greatest possible extent, facilitate long-term, infrastructure-based competition. However, Nkom also wishes to stimulate service competition and innovation at product level in order to ensure that users throughout the country have access to good quality, affordable and future-oriented mobile services, which is an expressed objective of the Ecom Regulations. In order to achieve the two aforementioned objectives, sufficient access must be available to input factors at wholesale level at the right price.”Footnote 180
  7. Similarly, a recent EC decision on the proposed merger of T-Mobile and Tele2 further illustrates that discount MVNOs appeal to, and can capture, a non-trivial share of the market:
    • “Mass-market MVNOs, such as Simpel or Youfone, offer no-frills subscriptions for the low and mid budget segment and do not target particular niche customers… Mass-market MVNOs are the strongest type of MNVO in the Netherlands. Simpel and Youfone have been the only MVNOs which have been able to capture a non-marginal share of subscribers in the postpaid private segment.”Footnote 181
  8. In conclusion, the Bureau found that MVNOs tend to target “niche” segments in the markets they serve which can be either small segments that the MNOs choose not to focus on, or larger segments like discount customers.

d. Evidence on the competitive role of MVNOs is mixed

  1. Subscriber share of MVNOs, in aggregate, does not typically exceed 10% in the markets where they operate. However, low subscriber share does not necessarily mean that MVNOs are ineffective in placing competitive pressure on MNOs. MVNO access has been mandated with varying degrees of success in delivering on policy objectives. However, despite relatively low subscriber shares, in some cases MVNOs have placed competitive pressure on the market in some jurisdictions.
  2. Following the 2012 Hutchison-Orange merger in Austria, wireless prices increased until MVNOs began entering the market in late 2014. The Austrian Regulatory Authority for Broadcasting and Telecommunications (RTR) conducted an ex post study and found that prices for new customers increased between 20% and 90% following the merger.Footnote 182 Those results were confirmed in a study by BEREC, which also found that prices decreased in 2016 following MVNO entry in 2014.Footnote 183 Prior to the merger remedy, there was only one independently owned Austrian MVNO, Vectone, with a market share of less than 1%.Footnote 184 Independently owned MVNOs currently have a market share of roughly 7% in Austria.Footnote 185
  3. Dr. Georg Serentschy, former CEO of the RTR directly links MVNO entry to the decrease in prices in an open letter on mobile pricing in Austria: “Even that measure [RTR monthly bill index] is now below 2011 levels, with MVNO entry driving four consecutive quarters of tariff reductions” (emphasis added).Footnote 186
  4. Further, though mandated access was only imposed on the acquiring firm Hutchison, the agreement sparked competition in the wholesale market leading other MNOs in Austria to voluntarily enter into MVNO agreements in order to capture wholesale revenue.Footnote 187
  5. Similarly, mandated MVNO access in Spain spurred competition in both the retail and wholesale markets. In their initial submissions in this Proceeding, several carriers highlight Spain’s decision to deregulate mandated MVNO access, but fail to acknowledge that the decision to do so followed a finding by CNMC that the wireless market had achieved “effective competition”.Footnote 188
  6. For example, in its initial submission, Telus states that there are “…several countries (Israel and Spain in particular) that have specifically rejected MVNO regulation”, citing Dr. Dippon’s report on regulatory frameworks of MVNOs. In his report, Dr. Dippon highlights the deregulation of mandated MVNO access in Spain, however he does not expand on why that decision was made, and simply states:
    • “Spain, which was one of the rare exceptions of MVNO-related regulation in 2014, decided to amend its regulatory regime and now relies on market forces.”Footnote 189
  7. In 2006, Spain imposed mandated MVNO access due to what regulators observed as collective dominance of the country’s three MNOs in denying access to MVNOs. However, following a subsequent review of the market in 2017, the CNMC found that mandated MVNO access had allowed entry into the retail market for a large number of MVNOs and resulted in a significant decrease in retail prices. In its decision letter, the EC describes the results of CNMC’s market analysis and its proposal to withdraw mandated MVNO access:
    • “CNMC explained that wholesale regulation allowed the entrance in the retail market of a large number of MVNOs. Currently, there are 33 MVNOs (13 of which are Full MVNOs) active in the market which have a joint market share of 10.7%. Traditional MNOs have the following retail market shares: Telefónica 30.2%, Vodafone 25.5% and Orange 27.2%. Since the last market review, a fourth MNO, Yoigo, which is part of the Masmóvil group, has been established. Yoigo holds a retail market share of approximately 6.4%.CNMC observed that, between 2011 and 2016, retail prices significantly decreased.
    • Between 2011 and 2016, CNMC observed a significant decrease of wholesale prices. The average revenue per minute of the wholesale voice call origination service dropped by more than 75 % between 2011 and 2016. The commercial agreements between MNOs and MVNOs have, according to CNMC, resulted in prices near the termination prices which are subject to BU pure-LRIC [Business Unit pure Long Run Incremental Cost] price regulation.”Footnote 190(References omitted)
  8. The growth of the MVNO market also led to the top 3 MNOs in Spain launching their own flanker brands to target low-budget and youth markets.Footnote 191
  9. Following mandated MVNO access in 2009 and the previously described spectrum tender in 2011, prices in Israel decreased dramatically. In 2011, the OECD Communications Outlook ranked Israel among the most expensive countries in the OECD for high usage baskets, however in 2013 it was among the least expensive.Footnote 192 However, as previously mentioned, it is difficult to isolate the effects of mandated MVNO access from the other regulatory measures imposed in Israel. In particular, the decreases in price seem to occur after the 2011 spectrum tender which led to two new entrants intended to transition into MNOs. These new entrants priced aggressively upon entering the market to gain the required market share for spectrum repayment.
  10. There are also several instances in which mandated MVNO access did not appear to successfully apply competitive pressure on the wireless market. For example, despite having mandated MVNO access since 2013 (against the advice of OSIPTEL), the four firms that currently have licenses to operate MVNOs in Peru are not yet properly operating in the countryFootnote 193 and Peru’s first MVNO, Virgin Mobile Peru, has already left the country.Footnote 194 In addition, as previously discussed, prices appear to have increased in Ireland and Germany following mandated MVNO policies in response to mergers. Further, following the 2014 merger remedy in Ireland in which Three Ireland was required to sell up to 30% of its network capacity to two MVNOs (15% each), one of the entrants, ID mobile, has since gone bankrupt as it was unable to fill its capacity.
  11. The Bureau also found a number of mischaracterizations with respect to the competitive impact of MVNOs in the initial interventions of this Proceeding. For instance, in the expert report prepared for Bell, Robert Madelin states:
    “...the [European] Commission, which has moved away from accepting MVNO access remedies, is not expected to revert back to accepting such remedies in merger cases”Footnote 195
  12. In his report, Madelin explains that the EC has changed its policy towards imposing MVNO access as a result of findings of significant market power or as a spectrum condition. Madelin identifies MVNOs lack of ability to compete with MNOs as the reason for the EC’s reluctance to accept MVNO merger remedies.
  13. However, in the EC’s recent decision on the proposed merger of T-Mobile and Tele2 in the Netherlands, while it does find that MVNOs “are not able to exercise the same degree of competitive pressure that is exercised by MNOs”, it explains that this may be the case because MVNOs are often constrained in their ability to compete.Footnote 196
  14. In his expert report prepared for Rogers, Mr. Feasey submits:
    “I explained in my 2014 Report how the ‘ladder of investment’ theory had initially been pursued but subsequently abandoned by regulators in Europe. This theory offered the promise that regulation could be introduced on a temporary basis to help entrants establish themselves in the downstream retail market before ‘climbing the ladder’ and investing in their own network facilities. But no MVNO, so far as I was aware, had ever ceased to rely upon regulated wholesale services because it had subsequently deployed its own mobile wireless network. Nor am I aware of any regulator in the world that has intervened in mobile wireless wholesale markets in the expectation that their intervention would be time limited because those MVNOs would subsequently transform themselves into facilities-based carriers.”Footnote 197
  15. The Bureau found that, while rare, there have been instances of MVNOs transitioning to MNOs. However, in the few instances that an MVNO has transitioned into an MNO, they are often acquired by another MNO.Footnote 198
  16. Further, the Bureau also found examples of regulators that have intervened in wireless wholesale markets with the expectation that their intervention would be time-limited and that MVNOs would subsequently transition into facilities-based carriers. For example, in the decisions of 2006 and 2010, Nkom facilitated gradual investments in infrastructure through access to established infrastructure at different levels (ladder of investment). Tele2 used gradual investment in the process of becoming a network owner in the Norwegian market.Footnote 199
  17. Similarly, in Italy, the EC and the Italian regulator and competition authority for the communications industry (AGCOM) accepted a remedy which included a combination of transitional roaming and divestitures, allowing the new entrant to transition from an MVNO to an MNO.Footnote 200 Iliad, the new entrant, has shown promise by capturing 4% of “human” SIMs (3.2% of total SIMs) in under a year while still in its transitional phase - market share which was captured from the three national incumbents.Footnote 201
  18. To conclude, evidence on MVNO policies’ ability to stimulate competition is mixed and dependent on the specific competitive conditions of each market.

e. Evidence on the degree to which MVNOs can reduce investment incentives is unclear

  1. In their initial submissions, many carriers cite a study by Kim et al. in support of the argument that mandated MVNO access will threaten investment and Canada’s 5G future.Footnote 202 The study examined data from 21 OECD countries from 2000-2008 and found that mandated MVNO access was associated with lower investments by MNOs. It is important to note that a key caveat of the paper is that lower investment does not mean under-investment.Footnote 203 That is, the authors do not seek to link any measured decrease in investment to worse outcomes for consumers or society overall. Additionally, the study analyzes just six instances of introducing or repealing mandated MVNO access regulations and in all cases these regulatory changes were from, or to, non-mandatory MVNO access regulations.Footnote 204 Furthermore, the authors acknowledge the difficulty in compiling data on regulations.Footnote 205
  2. Through its research, the Bureau did not find conclusive evidence from other jurisdictions of reduced investment incentives resulting from mandated MVNO access. In a number of jurisdictions where mandated MVNO access was effective in driving down prices, the Bureau did not find evidence that a significant decrease in investment had occurred.
  3. In Austria, Spain and Japan, where mandated MVNO access has contributed to placing downward pricing pressure on the wireless market, there does not appear to have been any significant decrease in investment.Footnote 205
  4. Austria ranked 13th in the world based on Ookla’s July 2019 global mobile speedtest,Footnote 207 was one of the first in the EU to auction 5G licenses, which was held in March 2019,Footnote 208 and is one of the top 5 countries in terms of mobile data usage per mobile broadband subscription.Footnote 209 Similarly, while Japan has mandated regulated MVNO access since 2002, it is one of the top 5 countries for 5G readiness along with the United States, China and South Korea,Footnote 210 suggesting network investment remains healthy, and is one of the top three countries for mobile broadband subscriptions per 100 inhabitants.Footnote 210 In 2018, Spain’s 4G coverage was 94%, on par with the EU average, and it had assigned 30% of its 5G spectrum which was higher than the EU average of 14%.Footnote 212
  5. Prior to Nkom’s decision to mandate MVNO access for service providers in 2016, Telenor argued that such a requirement would reduce its incentive to invest. In its 2016 decision, Nkom provided its assessment:
    • “In Nkom's opinion, regulated service provider access should not become such a good alternative to the construction of own infrastructure that it reduces the incentives to invest…However, Nkom is of the view that this is more an issue concerning the regulatory requirements relating to service provider access compared to other forms of access rather than an issue of whether access should be imposed.
    • Based on the above, Nkom is of the view that there is a need to impose an obligation on Telenor to accommodate any reasonable request for service provider access with the products and services that are included in the relevant market…In Nkom's opinion, the benefits to competition of such an obligation outweigh any potential disadvantages for Telenor.”Footnote 213
  6. In 2018, Telenor was named the world’s fastest network based on the Ookla speed test. In their press release, acting CEO of Telenor Norway Bjørn Ivar Moen said: “For more than a century, Telenor has been building a solid foundation in order to deliver the very best services possible. However, our country continues to challenge us with its vast mountains and deep fjords. That is why we’re especially proud of these results. We have ambitious expansion plans moving forward and these will continue throughout 2019.”Footnote 214
  7. The Bureau also found a number of mischaracterizations of the impact MVNOs have had on investment in other jurisdictions in the initial interventions of this Proceeding.
  8. In his report, submitted on behalf of Bell, Mr. Yossi Abadi suggests that fierce competition brought about by MVNOs, while serving to reduce prices for wireless services, also led to a decrease in capital expenditure and layoffs in the wireless industry in Israel.
  9. While there are indications that increased competition has led to reduced investment in Israel,Footnote 215 it is unclear how the policies implemented by the MOC may have played a part. As previously described, the MOC imposed a number of regulatory measures to promote new entrants and reduce barriers to switching in addition to mandating MVNO access. This makes it difficult to assess to what degree each policy may have impacted prices and investment. For instance, the spectrum repayment incentive which may have caused new entrants to price under cost in order to gain share could have been more instrumental in driving down prices to an unsustainable level and negatively impacting investment than the MVNO requirement.
  10. Additionally, in support of this claim, Mr. Abadi points to data from an article in an Israeli newspaper, Haaretz, and OECD data on telecommunications investment in the OECD area. The Bureau cautions against placing much weight on the inferences drawn by Mr. Abadi regarding the OECD data for two reasons:
    1. The data represents all telecommunications investment and should therefore not be used as a proxy for wireless investment, particularly given that wireline investment often exceeds that of wireless investment. For example, based on public financial statements, capital expenditure of Bell, Rogers and Telus from 2007-2017 for wireline ($52 billion) was nearly double that of wireless ($25 billion).
    2. In demonstrating the decrease in telecommunications investment, Mr. Abadi compares the investment per person in only the individual years 2010 and 2015. However, this does not offer an accurate picture of investment over that time period. Figure 16 shows investment per person between 2009, the year in which MVNO access was mandated, and 2015. It shows that, in the two years following mandated MVNO access, investment per person actually increased, and only in 2012 and 2015 was the per person investment below 2009 levels.

Figure 16 – Telecommunications investment per person in Israel Footnote 216

Figure 16 – Telecommunications investment per person in Israel
Details of the Telecommunications investment per person in Israel
2009 2010 2011 2012 2013 2014 2015
$123.23 $127.41 $130.55 $118.98 $124.58 $123.30 $111.25

  1. It is also important to take into account that countries without mandated wholesale access also experience ebbs and flows in investment from year-to-year. For example, Figure 17 illustrates the combined wireless capital investment for Bell, Rogers and Telus. It demonstrates the fluctuations in capital expenditure from one year to the next, for example, between 2009 and 2010 wireless capital investment decreased by 18.3%.

Figure 17 – Combined wireless capital expenditure of Bell, Rogers and Telus Footnote 217

Figure 17 – Combined wireless capital expenditure of Bell, Rogers and Telus
Details of the Combined wireless capital expenditure of Bell, Rogers and Telus
2009 2010 2011 2012 2013 2014 2015
$123.23 $127.41 $130.55 $118.98 $124.58 $123.30 $111.25
  1. The expert report, An Analysis of the Performance of the Canadian Mobile Wireless Industry, submitted on behalf of Telus, suggests that capital expenditure in Canada and the US, which rely on facilities-based competition, is greater than in the EU, which has relied more heavily on service-based competition.Footnote 218
  2. In the same way comparing wireless prices across countries can be difficult, so too can comparing capital expenditure. For instance, differences in costs, characteristics of networks and population density can drive differences in capital expenditure. While the use of econometric techniques that control for observable factors (e.g., network quality) attempt to remove such effects, direct comparisons without such analysis should be observed with caution.
  3. While the Telus expert report attributes greater capital expenditure in Canada and the US to a reliance on facilities-based competition, it is also possible that variations in population density and land mass could be contributing to lower capital expenditure per subscriber in the EU. It is also important to note, that while the report groups together all EU countries, very few continue to impose mandated MVNO access. Similar to the Abadi report, the report also cites figures which represent all telecommunications investment and should therefore not be used as a proxy for wireless investment.

IX. Designing a mandated MVNO policy

a. Summary of Bureau findings relevant to assessing the merits of mandated MVNO access

  1. In order to introduce additional competitive pressure into Canada’s wholesale wireless and retail wireless markets the CRTC is considering an MVNO model where the national wireless carriers would be required to provide wholesale access to their networks.
  2. While MVNOs have emerged organically in other markets, this has not occurred in the Canadian market. There are a handful of MVNO-like offerings in Canada (e.g. 7-11 Speak Out, Petro Canada Mobility, PC Mobile), however the terms on which these entities access wholesale wireless services from the national wireless carriers do not suggest they are providing a competitive constraint to the Big 3 or regional carriers.
  3. While the Bureau continues to have concerns regarding the degree of market power in the Canadian wireless market, there are promising signs that policies aimed at promoting facilities-based competition are paying dividends. The Matrix Study found significant price impacts from regional carrier facilities-based competition, particularly for carriers who are able to reach between 5.5% and 20% market share.
  4. Although facilities-based competition has begun to yield competitive benefits for Canadian consumers in the form of increased choice and lower prices, the growth of regional new entrants has been slow and uneven across the country. While portions of the country have begun to benefit from the presence of a wireless disruptor, there remains a substantial population with no material alternative to the national wireless carriers.
  5. For example, the Matrix Study shows the growth of Freedom is having a price-reducing impact on the Big 3, but not yet at the level of Videotron in their respective markets. The story is not consistent among all regional new entrant carriers, as Eastlink does not appear to have had a meaningful impact on pricing in its service area, thus far.
  6. In principle, a mandated MVNO regime has the potential to achieve a higher competitive intensity in the wireless market; however, to state the obvious, the design of the regime is critical to its success.
  7. International experience with mandating MVNO access is mixed, with some jurisdictions experiencing increased competition at the wholesale and retail levels of the market while others do not observe meaningful increases in competition. The Matrix Study findings imply that an MVNO would have to obtain above 5.5% market share to have an impact on the market, but ideally an MVNO would achieve a 20% market share to have a comparable impact to that of wireless disruptors. International experience suggests that is an ambitious goal particularly in the context of a mandated access regime. For example, with mandated MVNO access, MVNOs in Spain collectively obtained a 10% market share, and in Austria, the MVNO share was 7%.
  8. When considering the investment impact of MVNOs, there is little evidence of decreased investment in response to these policies internationally. However, many jurisdictions introduced this policy in part to incent a fourth MNO entrant. In the Canadian context, most regions in Canada have a fourth MNO in the process of expanding to reach their full competitive potential. The investment implications are likely to be different in Canada for the facilities-based entrants who may be particularly at risk in competing with an MVNO with access to the incumbents’ relatively higher-quality networks. This could be detrimental to their future investment prospects, particularly given their already narrower margins and high capital intensity.

b. The design of a mandated MVNO policy

  1. There are two key institutional design choices from a competition perspective:
    1. the criteria for access to MVNO wholesale; and
    2. the approach to rate setting. Design choices are in part dependent on the specific policy goals of the CRTC in the wireless market.

c. Access criteria

  1. Although any form of mandated MVNO involves imposing a requirement on MNOs to provide access to their wireless networks, one element of differentiation is the criteria a company must meet in order to be granted access to wholesale wireless services. The type of access criteria selected by the CRTC hinges on whether the CRTC sees MVNOs as a method of accelerating and incentivizing the expansion of facilities-based competition or a way to introduce services-based competition into a market in which it has not developed.

MVNO towards a facilities focus

  1. An MVNO model with a facilities focus would provide wholesale access to wireless disruptors that could reasonably implement or expand an existing network of physical wireless infrastructure. This would provide temporary mandated access to companies willing and able to build a network and transition customers after the mandated access period.
  2. The goal of a facilities-focused MVNO regime is to allow an existing regional MNO to begin offering wireless service in an area in which it intends to build but has not yet established the corresponding infrastructure. Under a temporary MVNO regime the revenue earned from wireless competition would support infrastructure investments for use once the MVNO regime has expired. This could particularly benefit those carriers who wish to expand, but are limited by lower margins and higher capital intensity. For instance, Freedom suggests:
    • “It is true that new regional competitors, such as Freedom, have coverage limits on our network. And, it is also true that Freedom operates at half the margin and over 2 times the CAPEX intensity as the Big 3, making it more difficult and higher risk for Shaw to close those coverage limits.
    • If regulatory action were required and focused on bridging these network coverage limits, then new wireless competitors would be able to sign up subscribers and build a revenue base in these areas, enhancing the quality and reach of their services, and accelerating competition immediately. The revenues would allow new competitors to gain scale quickly, driving investments in facilities-based competition in those areas, establishing strong medium and long term competition to challenge the national dominance of the Big 3.”Footnote 219
  3. Along with spurring infrastructure investment for regional players, this approach also allows for price competition in the short term in areas where a regional player is not currently present. This allows for an extension of some of the pricing benefits identified in the Matrix Study in a much shorter timeframe than would be achieved under current investment time horizons.
  4. A source of complexity in this approach is reaching a wholesale rate that both incents regional players to shorten the time horizon to full network deployment, and expand the potential reach of investments, without discouraging them from making those investments in the first place.Footnote 220 It is possible that a regional player would be willing to acquire customer relationships through the temporary MVNO period without establishing the underlying infrastructure, but the result would be significant consumer disruption within the area of its spectrum holdings as the MVNO regime expires. The CRTC would need to be confident it can avoid this possibility if it wishes to pursue this facilities-focused path.
  5. This risk could be mitigated by attaching build out requirements to the MVNO regime within a regional players’ spectrum holdings, similar to the ones found in spectrum set-aside auctions.
  6. While in the context of merger remedies, recent experience of international jurisdictions, such as ItalyFootnote 221 and the U.S.Footnote 221, as described above, may aid the CRTC in designing an MVNO model with a facilities focus.
  7. A facilities-focused MVNO model has strong potential to disrupt the current market dynamic at the retail level as expert analysis by the Bureau has shown, and has the potential to affect the level of competition in the wholesale market in the long term, as additional network capacity is added to the market through entrant network builds. This may, in turn, spawn MVNO entry organically once a sufficiently competitive market is attained.

MVNO towards a facilities-services mix

  1. The alternative to an MVNO regime with a facilities focus is one that encourages a mix of facilities-based competition and service-based competition by selecting broad access criteria. Criteria would be established for access to wholesale wireless markets but it would necessarily include a larger pool of potential entrants than a facilities-focused approach.
  2. In terms of stimulating new entry in the retail wireless market an MVNO model that allows broad market access shows the most promise.
  3. Although wireline and wireless represent different markets with unique characteristics, the broad MVNO approach employed in the wireline market might result in a similar market structure of a high number of relatively small participants with a select few more established players. However, in contrast to the wireline market, there are currently facilities-based entrants in the process of building network infrastructure.
  4. The most obvious trade off involved in this approach is that while a wider pool of entrants is possible, these entrants will have limited ability to invest in establishing a wireless network given their lack of spectrum assets.
  5. This approach may also exacerbate existing network gaps for rural and remote communities, reducing the business case for investing in already less commercially attractive regions.Footnote 223 However, the CRTC appears to have appropriately recognized that other policy remedies might be required to extend infrastructure investments to currently unconnected or under-connected communities.
  6. There is also an increased risk that facilities-based regional entrants with lower margins resulting from ongoing network infrastructure investments to catch up to the Big 3 are disadvantaged against MVNO entrants, relative to higher margin incumbent players. It is possible that this risk could be mitigated through a less generous wholesale rate. However, this may in turn defeat the purpose of the policy.
  7. Although experience in jurisdictions such as Spain suggests that a temporary MVNO regime that encourages a mix of facility and service competition is possible, it is more likely that this regime would extend in perpetuity.

Bureau recommendation

  1. The Bureau recommends that the CRTC adopt a facilities focus in designing an MVNO remedy.
  2. All else equal, facilities-based competition is the most sustainable and effective form of competition. Regulated service-based competition is a justifiable approach where market-based competition will be insufficiently vigorous to promote consumer interests absent regulation.
  3. More specifically, a weakness of a broad mandated MVNO model to generate service-based competition is that the new entrant MVNO is reliant on an adversarial supplier likely incented to raise the MVNO’s costs and use whatever levers are available to make them a less effective competitor downstream.Footnote 224 In response to these incentives, the regulator is able to set terms and conditions of access, as well as a wholesale rate.
  4. This approach has the potential to generate greater competition, as it has in the broadband market, but represents a second best outcome since the MVNO is jointly reliant on the network operator and regulator to set the bounds within which they can compete.
  5. A key distinction between broadband and wireless markets is that multiple vertically integrated regional wireless providers are building their own networks. This was not the case in the broadband market where there is little business rationale for an additional player to build a standalone wireline network.Footnote 225
  6. Regional network operators like Videotron have full control over their pricing, product offerings and network quality, allowing them to compete more aggressively given their greater latitude. Therefore, it may not be necessary to move to the second best option in wireless if regional competitors are willing and able to expand their reach and compete aggressively. In essence, a broader MVNO policy may result in more entry but likely less effective entry.
  7. Expanded regional MNO competition may also pave the way for organic MVNO entry in the future. As regional operators expand their networks, they create more viable wholesale alternatives to the Big 3, increasing the likelihood that a market will develop for wholesale wireless service. This outcome could occur due to the increased quality and reach of regional networks, and additional capacity being created in the market.
  8. The competitive impact that the wireless disruptors are having is encouraging and a sign that a competitive market may be developing. However, 11 years into pursuing policies to promote facilities-based competition, not all Canadians are seeing those benefits. An MVNO policy with a facilities-based focus is therefore fitting as it is itself a middle ground. If properly designed, it could deliver more competition from wireless disruptors in the short term, while maintaining their incentive to continue to invest and expand.
  9. A broad MVNO access criteria may also deliver competition, but at a cost. The prospects for success, and the risks associated with adopting a more broad MVNO access policy, are highly dependent on rate setting. If the rate is too high, the policy will have minimal effect on competition.Footnote 226 If the rate is too low, wireless disruptors are put at risk, as the economics of being an MVNO are more appealing than investing. This would likely result in a regulated solution in perpetuity.
  10. Since MVNOs would be purchasing access to the Big 3’s networks, which test favourably against most entrant networks, an MVNO with little investment, can obtain a significant advantage over wireless disruptors who have invested for more than a decade. Further, international experience suggests that it is likely MVNOs will target a similar customer base to that of the wireless disruptors who are particularly vulnerable given their lower margins and higher capital intensity. Were it to be the case that wireless disruptors were still not showing promise, it might be justifiable to pursue a broader MVNO policy; however, in the Bureau’s view at this time the risks associated with such a policy are too high for it to be warranted.
  11. A facilities-focused MVNO policy is intended as a temporary measure with the goal of deregulating the market after a 5 year period. The 5 year period was selected to align with the CRTC’s next wireless review. The Bureau recommends that the CRTC limit access only to those who are likely to be able to build a network within a 5 year period and transition the customers they won as an MVNO to their own network. To determine access, the Bureau suggests that the criteria outlined in the Bureau’s remedy guidelines may be an appropriate reference point. Specifically, a potential MVNO should have the “managerial, operational, and financial capability” to compete effectively in the relevant market(s) following the mandated MVNO access period.Footnote 227
  12. The most likely entrants to be able to prove that they have the “operational capability” to build out in 5 years are those who are already operating wireless networks in Canada. A facilities-focused MVNO policy does inherently mean fewer entrants will be eligible to enter the market; however, given the importance of economies of scale in wireless network operation, this may be a logical outcome.
  13. Practically speaking, only wireless providers who have a sufficient quantity and mix of spectrum in an area, but have not yet built out, are likely to have the operational capability to build out in 5 years.Footnote 228 Under this policy, Saskatchewan and Quebec are therefore unlikely to see further entry, as those regional carriers have already built an extensive network in the places they where hold spectrum. That said, this outcome is consistent with the competition assessment that shows those regions currently benefit from a larger degree of competition.
  14. Based on the Matrix Study, the remaining regions in Canada could likely benefit from additional competition and could have prospective entry and expansion under this policy. The policy could accelerate Freedom’s planned entry and expansion as Freedom would be able to gain a base of subscribers before entering a market. It is also possible that Xplornet could choose to enter as an MVNO and build a network in the Maritimes where they purchased 600 MHz spectrum, accelerate their competitive impact and deployment in Manitoba, and possibly enter other areas across Canada. Eastlink could also sell beyond their existing network footprint in Ontario and the Maritimes and Tbaytel could also expand in Northern Ontario. Finally, Iristel could offer services in the Northwest Territories, the Yukon and Nunavut.
  15. The CRTC will likely want to preserve the incentive for regional carriers to continue to invest and expand, and avoid incentivizing funds intended to invest in building out rural areas being diverted to investments in the cities. Therefore, a key element of this policy’s design is enforcing build out requirements. To do so, the Bureau recommends the following.
    1. First, and at minimum, the Bureau recommends the CRTC set a clear expectation that if this policy does not substantially improve competitive outcomes, by the time the CRTC conducts another wireless review in 5 years, then the CRTC will mandate broad MVNO access in markets that are still lacking competition. Setting this clear expectation may itself incentivize MVNO entrants to adhere to the CRTC’s policy objectives.
    2. Second, the potential MVNO may also be required to make credible commitments to ensure they will meet deployment requirements. Such commitments might involve:
      1. financial penalties if the MVNO fails to meet the agreed upon timeframe for facilities-based entry; or
      2. up-front financial investments that would be substantial, but still on a smaller scale than if the candidate were to enter as an MNO.
  16. For option (i), the CRTC may consider a separate administrative monetary penalty (AMP) in each area where a service provider operated as an MVNO but did not meet the build requirement. The AMP should be high enough to ensure that their incentive to build out in rural regions is maintained.
  17. With effective policy design, this remedy has the potential to stimulate and accelerate competition from wireless disruptors in the short term, without the risk of undermining continued network builds and upgrades by wireless disruptors and incumbents.
  18. As regional carriers grow and expand their networks, the factors that discouraged them from coordinating with the Big 3 may be eroded in the long-term. Regional carriers may no longer have cost and market share asymmetries, network differentiation, and a lack of multi-market contact with the Big 3. Factors that make the wireless industry susceptible to coordination like barriers to entry and price transparency will likely remain; however, the likelihood of coordination is decreased with the expanded number of competitors. This is an important risk to be monitored as competition progresses; however, the Bureau has no evidence of coordination among the Big 3 and regional carriers to date. Were coordinated behaviour amongst the Big 3 and regional carriers may be observed, a broad MVNO access policy may be a possible solution

d. Approaches to rate setting

  1. The second major competitive dimension to an MVNO model is the mechanism used to decide the appropriate rate for wholesale wireless access.
  2. In general, the Bureau is not in favour of price regulation. However, as demonstrated in the Bureau’s broadband study,Footnote 229 the Bureau recognizes that in markets with material competitive challenges, there is potential to spur competition through government intervention.
  3. These interventions are not without cost, and in fact, the costs of these interventions are often borne indirectly by consumers. Should the CRTC decide to impose a mandated MVNO regime it must weigh these costs against their competitive impacts.
  4. There are three high-level methods of determining an appropriate rate for wholesale wireless access: cost plus, retail minus and negotiation backstops.

Cost plus

  1. Rate setting based on a cost plus methodology creates a wholesale access rate by estimating the infrastructure cost for an MNO to deliver wireless service to consumers and applies additional margin to provide the MNO with a fair rate of return.
  2. A cost plus approach is attractive in that it does not create the incentive to raise retail prices in response to mandating access and allows incumbent MNOs to receive a fair rate of the return for the investment in network infrastructure.
  3. The clearest drawback to the cost plus methodology is the complexity of determining the appropriate cost base on which to calculate the corresponding rate and the length of time associated with that process. For example, it took several years for the CRTC to determine the final wholesale wireline access rate on a cost plus basis and that rate has already required material adjustment.
  4. A follow-on result is that a cost plus approach is slow to adjust to changes in underlying network infrastructure. This is problematic in an environment where Canadian MNOs are moving towards implementing 5G networks.
  5. The complexity of cost plus and reliance on detailed private MNO data makes rates determined under this approach vulnerable to gaming by MNOs. As much as possible, a cost plus approach should use auditable data already in use for the routine businesses operations of MNOs.
  6. One potential mitigating factor is that cost plus allows for differential treatment of portions of a MNOs network infrastructure. A mandating regime that employs cost plus could include incentives to favour investment in next-generation network (NGN) infrastructure.

Retail minus

  1. Forgoing a complex calculation of an underlying cost base, a retail minus approach sets wholesale access rates by applying a fixed percentage discount on current retail wireless prices.
  2. The primary benefits of a retail minus regime is the relative ease with which a wholesale wireless access rate can be set and the publicly available nature of the data used to make the pricing determination. The former allows the wholesale wireless access price to respond to changes in market conditions, while the latter removes the opportunity to game the price through misrepresentation of underlying data.
  3. The goal of a retail minus approach is to set the discount at a level that allows MVNOs to provide competitive pressure to the market, while still fairly compensating MNOs fairly for their implementation and maintenance of network infrastructure. This is the primary source of regulatory complexity in a retail minus regime.
  4. There are also important decisions on what constitutes the appropriate retail rate, as MNOs routinely offer a variety of plans and price points. Retail minus requires the CRTC to develop a methodology to take this variety of price points and develop a rate charged to MVNOs. Another benefit of retail minus is that the calculation depends on publicly available data, reducing opportunities for gaming.
  5. However, retail minus does include the potential for another type of incentive problem that may be even more problematic for an MVNO regime. In a market with coordinated behaviour lacking a disruptive player, retail minus pricing creates an incentive for MNOs to set higher retail prices compared to under a cost based approach.Footnote 230 This also reduces the ability of MVNO entrants to provide effective competitive pressure as their wholesale access cost rises with the retail price.
  6. The current dynamics of the Canadian market make this risk worthy of consideration. This is particularly acute in regions without the competitive pressure of a wireless disruptor.
  7. Retail minus has been employed relatively rarely in international jurisdictions, often complementing a broader cost plus access regime.

Negotiation backstop

  1. The final approach to rate setting is one that relies on regulatory backstop in the event of the break down of commercial negotiation rather than pre-determining pricing based on retail or underlying cost figures.
  2. The goal of the negotiation backstop is to remove barriers to agreement and incentivize parties to reach terms that balance introducing competitive pressure and fairly rewarding an MNO for their infrastructure investments.Footnote 231
  3. The negotiation backstop would address complaints by potential MVNO entrants that MNOs are not willing to engage in good faith negotiations for wholesale access. By mandating a wholesale wireless regime and implementing a negotiation backstop, the CRTC would communicate to parties its preference for the establishment of MVNO arrangements, and shift incentives for both parties towards seeking good faith agreement.
  4. This approach allows for a greater level of flexibility in the determination of wholesale access rates as arrangements may be flexible to changing market dynamics and underlying cost structures. By emphasizing the input of market participants, it is more likely that the agreed upon price reflects an approximation of a market outcome, rather than a purely regulatory outcome.
  5. This approach aligns with international experience, wherein the Bureau’s study of other jurisdictions revealed that instances of regulatory intervention involving wholesale rate regulation are very rare. Most regulators opt for a ‘lighter touch’ approach by mandating fair and reasonable commercial negotiations, with a regulatory backstop in place in case negotiations fail.
  6. Arbitration processes can be costly and time consuming. As such, the design of the arbitration process and the selection of a qualified arbitrator is critical to the success of a negotiation backstop regime.
  7. While intervention methods can vary, an approach that has seen some degree of success both domestically and abroad is final offer arbitration. Under this approach, both parties submit a final bid at the conclusion of negotiations and the arbitrator selects the offer deemed most appropriate. This provides an incentive for both parties to submit offers that an arbitrator will consider reasonable, given the alternative offer is likely to be more favourable to the opposing party.
  8. Given the history of complaints about commercial negotiation experiences in the telecommunications market, the design of the arbitration process should be mindful of opportunities for participants to unnecessarily delay or derail proceedings. Arbitration will involve some level of cost and time commitment, but the process should aim to be as efficiently as possible to reduce the burden on parties.
  9. The selection of the arbitrator is equally important to the success of any rate setting based on commercial negotiation. Selecting an arbitrator with a deep understanding of the economic and technical characteristics of the market is critical to avoiding unnecessary delay and reaching an outcome that introduces competitive pressure to the market.
  10. The CRTC is well positioned to play the role of arbitrator under a final-offer backstop approach. The CRTC’s regulatory skill set and knowledge of the sector give it a comparative advantage in the ability to reach fair decisions to support a more competitive wireless market.

Bureau recommendation

  1. Regardless of the access criteria the CRTC deems appropriate, the Bureau recommends the CRTC adopt a negotiation backstop for setting MVNO rates. A negotiation back stop strengthens a potential MVNO’s negotiating power with the Big 3 and has the benefit of being a lighter touch approach to regulation which can adjust more quickly to changes in the market.
  2. In contrast, the costing studies associated with setting a cost plus rate are long and complex. A cost plus approach is particularly problematic at this juncture with 5G technology on the horizon. Historical costing information will be of little relevance in setting rates. A retail minus strategy is unlikely to be a better approach; in a market that is susceptible to coordination this policy could have the unintended consequence of actually increasing wireless pricing.
  3. The Bureau recognizes that negotiation back stops exist in the wireless industry;Footnote 232 however, it is the details that are critical to the effectiveness of such a policy. A negotiation back stop has a demonstrated track record with other jurisdictions since it has typically been the preferred approach. The CRTC can therefore look to its international counterparts who have designed and tested such an approach, including Spain where their mandated MVNO policy using a negotiation back stop triggered sufficient competition to warrant deregulation of the market.Footnote 233
  4. For the dispute resolution mechanism an appropriate model to consider may be final offer arbitration (FOA).Footnote 234 FOA differs from conventional arbitration or a costing study in that it can create incentives to act reasonably, it is more efficient, and it can keep costs in check.Footnote 235 The CRTC benefits from having experience with this process.Footnote 236 The Bureau previously examined this process in relation to an enforcement matter and found that it can be an effective mechanism to remedy disputes.Footnote 237 The Bureau also made similar conclusions about final offer arbitration mechanisms in other contexts.Footnote 238
  5. As previously described, the arbitrator is a key policy design element. The CRTC is likely to be an appropriate arbitrator of disputes given the complex nature of these arbitration processes and its experience examining wireless carrier costs in setting wholesale roaming rates. Based on the Matrix findings herein that increased regional competition lowers the Big 3’s prices but does not affect network quality, the Bureau recommends that the CRTC look to prevailing retail pricing in more competitive regions in Canada to “sanity check” the proposals it receives.Footnote 239

e. Applicability to the national carriers

  1. Spectrum is an essential input for any wireless provider wishing to enter a new market and compete. As such, the Big 3 may assign an added value, often called a “foreclosure value”, to spectrum holdings due to their ability to preclude competition.Footnote 240 Foreclosure value can drive up the price of spectrum creating another barrier for potential new entrants. To combat this issue, ISED has employed policies to ensure new entrants are able to access the spectrum they require to continue to expand their networks. Canadians are currently seeing the benefits of these policies borne out in the development and expansion of regional carrier networks, and its associated impacts on pricing.
  2. That said, as demonstrated in Figure 18, The Big 3 continue to hold the lion’s share of spectrum in Canada, including critical low-band spectrum, likely due to historical advantages in spectrum allocation.

Figure 18 – Spectrum ownership in terms of serviceable population Footnote 241 Footnote 242

[## ##]

  1. In particular, it appears the scales have been tipped in favour of the Bell and Telus network. Rogers has explained that they are not able to support MVNOs without investing incrementally, as they do not have sufficient excess capacity.
    • “Rogers closely monitors its available capacity at all points of its network. Rogers only undertakes the necessary upgrades to increase capacity when a particular element in its network is approaching its limitations. This has two implications. First, Rogers is not actively seeking arrangements with MVNOs to fill spare capacity, and secondly, Rogers cannot serve MVNOs without investing incrementally more capital to expand network capacity.”Footnote 243
  2. Rogers’ claim is supported by the quantity of unused spectrum in their portfolio versus that of their national competitors. Bell and Telus hold a combined [##]% of the unused spectrum owned by the Big 3.Footnote 244 It is also supported by Rogers’ recent speeds slightly lagging behind those of Bell and Telus in the major cities.Footnote 245 That said, Rogers recently won the most 600 MHz spectrum of the Big 3,Footnote 246 thus this picture may change when that spectrum is to be deployed.
  3. Further, Bell and Telus frequently cite the cost advantage that network sharing provides them in investor calls. For example Bell has said “because of our significant wireline fibre investment, as well as our network sharing arrangement with Telus, we were able to maintain an industry-low wireless capital intensity of around 7%[.]”Footnote 247 Rogers on the other hand was reporting a wireless capital intensity of 14%, double that of Bell and Telus.Footnote 248
  4. This may be a relevant consideration to the CRTC in designing an MVNO regime and whether it applies to all there national carriers equally.

f. Additional remedies to explore

  1. Whether or not the CRTC plans to impose a mandated MVNO regime on the Canadian wireless market, there remain alternative and additional remedies to improve the level of competitive intensity in the Canadian wireless market. These remedies are focused on reducing the cost of switching, and barriers to entry and expansion.

Seamless handoff

  1. Different from seamless roaming, seamless handoff (and in particular seamless handback) allows customers to enter roaming markets and return to home markets with no interruption in service (e.g. dropped calls, interrupted downloads / uploads). Regional MNOs identify this interruption as a detrimental customer experience that degrades the quality of their service offerings and weakens their ability to compete effectively.
  2. While there is a level of technical complexity and required investment associated with this capability, Shaw suggests incumbent MNOs already have this capability with one another:
    • “We are aware that Bell and Telus, for example, support seamless handover. A Bell customer can move out of a Bell service area and into a Telus service area without dropping a call. We believe the same can be said with respect to Rogers and Bell.”Footnote 249
  3. Extending this seamless handoff requirement to all carriers would work to level the competitive playing field and meet the intended policy goals of the mandated wireless roaming regime.

Changes to tower sharing and site access rules

  1. There appears to be room for improvement in the competitive potential of Canada’s tower sharing and site access rules. A number of parties to this Proceeding suggest the current rules create the potential for delay and obstruction of legitimate access to towers and sites by new entrants.
  2. For instance, Rogers suggests that the permit application review is unreasonably complicated and expensive, permit denials are unresponsive and lack sufficient detail and justification, claims of insufficient capacity are unfounded or misleading, the practice of issuing “partial” approvals results in piece-meal and inefficient installations, make-ready costs are prohibitive and Incumbent Local Exchange Carriers (ILECs) can also control access to utility poles.Footnote 250 Similarly, Shaw suggests:
    • “Shaw experiences significant delays in gaining access to ILEC (primarily Telus) support structures. The two most common delays relate to first, denials of access due to alleged cases of “no spare capacity” and second, undue delays engendered by ILEC insistence on performing make-ready work and lack of service standards for completion of such work by the ILEC...
    • …[d]enials of access to ILEC support structures on grounds of allegedly “no spare capacity” rarely detail why no spare capacity is available, including whether this is due to reservations of space by the ILEC for its own future use.”Footnote 251
  3. Narrowing the definition of “imminent and future use” and their associated timing windows could reduce an incumbent MNO’s ability to obstruct the installation of a new entrant’s infrastructure on existing towers and sites.
  4. Beyond these changes, a “negative approval” approach may improve outcomes. This would effectively reverse the onus to respond to requests for access to towers and sites. Under the current regime, if a request is not acknowledged within a set window, that request is not considered approved and the requester must seek assistance from ISED to remedy the situation. Under a negative approval approach, requests not addressed within the window are considered approved, incenting MNOs to respond more quickly to requests by new entrants.
  5. The Bureau is of the view that more aggressive tower sharing and site access rules could facilitate the expansion of regional carriers.

Reduced roaming rates and alternate means of negotiating roaming rates

  1. The rate associated with the CRTC’s mandated roaming regime is a key input to a regional MNO’s ability to compete against incumbents with national networks. Effectively priced roaming rates reduce the importance of establishing a national network of physical infrastructure and allow regional MNOs to attract customers with service needs outside their network.
  2. An inflated wireless roaming rate reduces the competitive potential for regional MNOs and does not meet the competitive intent of the CRTC’s mandated wireless roaming regime.
  3. There is evidence to suggest the current wireless roaming rate is inflated. In particular, evidence suggests that the current per gigabyte roaming rate is well above the per gigabyte retail rate offered to Canadian consumers; this indicates the rate does not reflect the underlying cost structure. In its initial submission, Eastlink submits:
    • “…pursuant to the wholesale roaming tariffs, the incumbents charge data roaming at $13.00-$14.00 per Gigabit of data, while the data rates they charge for their retail customers are at significantly lower rates. While the tariffed rates are much lower than the excessively high rates previously charged, the decisions have not included a review or other mechanism to establish reductions or reviews of wholesale roaming rates in cases where changes to costs associated with roaming should result in lower rates, or where other circumstances warrant.”Footnote 252
  4. Similar to the Bureau’s recommended approach above regarding rate setting in the context of a mandated MVNO policy, looking to retail market rates in more competitive markets to test the reasonableness of roaming rates would be valuable. For example, in Quebec, Bell “unlimited” plans with 10 GB of data (after which network speeds are throttled) are approximately $6.5 per GB at retailFootnote 253 but the roaming rates are between $13 and $14 per GB (depending on the carrier).Footnote 254

Device decoupling

  1. In its initial submission, the Bureau proposed various steps that could be taken to accelerate the disruptive effect of new facilities-based competitors, including steps to reduce switching costs. One example provided was separating the handset purchase from the service plan, which has been implemented in Israel.
  2. As part of its RFI on July 5, the CRTC asked the wireless service providers to provide their views on whether it would be appropriate to establish such an obligation as a way to facilitate switching. In their responses, a number of carriers suggest that such plans already exist in the Canadian market. For instance, in their response, Telus submits:
    • “...[M]any Canadian wireless service providers including Bell, Rogers, Telus, and Iristel also now offer separate device financing which separates the purchase of the handset from the wireless service plan. Device financing allows the customer to pay for the device in monthly installments. After the remaining balance of the device financing amount is fully paid, customers no longer make monthly payments for the device. At Telus, device financing is called Easy Payment. The Easy Payment device financing agreement is a separate agreement from the wireless services agreement and provides financing at 0% APR. There is a $0 upfront option for all phones in the Telus Mobility lineup, thus making it easier for customers to afford the latest smartphones. Telus Easy Payment is but one example of the device financing options that are available in market that have evolved without unnecessary wireless retail regulation.”
  3. While device financing contracts offered by Bell, Rogers, Telus and Iristel are separate from service contracts, they are not functionally separate. A service contract cannot be cancelled without also cancelling and paying the outstanding balance on the device financing contract,Footnote 255 thus eliminating any potential benefits to ease switching.
  4. Decoupling the device financing and service contracts can reduce barriers to switching by allowing a customer to cancel their service contract without needing to cancel their device financing contract. In this scenario, the customer is free to switch to a competitive service plan, while continuing to pay off the balance on their device with their previous service provider.

X. Conclusion

  1. It is clear that the Big 3 have market power at the wholesale and retail levels of the wireless market. Indicia and direct evidence support this finding. Further, evidence suggests that the wireless market is susceptible to coordination and that there has been coordinated behaviour among the Big 3.
  2. Wireless disruptors, with a market share between 5.5 and 20%, have increasingly been able to discipline the Big 3 in the markets where they operate which shows promise of more affordable wireless services for consumers, to the credit of policies aimed at promoting facilities-based competition introduced by the CRTC and ISED.
  3. However, this competition has not yet reached its full potential and a mandated MVNO policy applied broadly risks undermining the steps taken by wireless disruptors, without much certainty that the MVNO policy will significantly decrease pricing. If the CRTC intends to introduce a low access rate for MVNOs it should do so knowing that it is placing wireless disruptors at risk. With a high access rate, MVNOs are unlikely to have a significant competitive impact on the market.
  4. The Bureau believes there is a remedy that can stimulate competition in the short run while ensuring continued network deployment and facilities-based competition from wireless disruptors. This remedy accelerates and expands competition from wireless disruptors in the short run by allowing them to enter and deploy in new markets. Wireless disruptors, as shown by the Bureau’s expert, have a proven track record of significantly impacting pricing and data usage when they enter a market, and a proven ability to build wireless networks. This remedy does not put wireless disruptors at risk by narrowing their margins below current levels so they can no longer economically justify network expansions.
  5. The key to the success of this remedy is enforcing continued network expansion. A potential negative outcome would be wireless disruptors choosing to forego, delay or divert funds that would have been used to expand their networks and rely on mandated access. The Bureau has suggested three ways to overcome this challenge:
    1. upfront investment;
    2. fines for not deploying a network; and
    3. a clearly expressed intention to mandate broad MVNO access in 5 years if this policy is not successful.
    Additionally, policies which seek to lower barriers to entry and reduce switching costs may be considered in conjunction with this remedy to facilitate deployment and further market share gain by the wireless disruptors.
  6. The Bureau would be pleased to participate in the public hearing.
  7. For the purposes of this Proceeding, the designated representative of the Commissioner is:
    • Laura Sonley
      Senior Competition Law Officer, Competition Promotion Branch
      Competition Bureau
      21st Floor, 50 Victoria Street
      Gatineau, Quebec K1A 0C9