ARCHIVED - Transcript, Hearing October 31, 2016

Volume: 1
Location: Gatineau, Quebec
Date: October 31, 2016
© Copyright Reserved

Attendees and Location

Held at:

Outaouais Room
Conference Centre
140 Promenade du Portage
Gatineau, Québec

Attendees:

• Chairman: Jean-Pierre Blais
  
• Members: Yves Dupras, Peter Menzies, Linda Vennard, Chris MacDonald
  
• Legal Advisors: Carolyn Pinsky, Daniel Finestone
  
• Secretary: Cindy Ventura
  
• Hearing Manager: Suneil Kanjeekal
  

Transcript

Gatineau, Québec

--- Upon commencing on Monday, October 31, 2016 at 9:02 a.m.

1 LE PRÉSIDENT: Order, please. À l’ordre, s’il vous plait.

2 Alors bonjour, Mesdames et Messieurs, et bienvenue à cette audience publique.

3 Avant de commencer je tiens à reconnaitre que nous sommes réunis aujourd’hui sur le territoire traditionnel des Premières Nations. Je remercie le peuple Algonquin et rends hommage à ses aînés.

4 This hearing is being held as part of the process to examine the policy issues surrounding the use of differential pricing practices by Canadian fixed and mobile internet service providers.

5 En bref, il y a différentiation des prix pour les services internet lorsque le fournisseur de services fixe des prix différents pour accéder l’internet selon différentes circonstances.

6 Par exemple, lorsque des applications précises sont exemptées d’un forfait de données mensuelles, une pratique appelée taux zéro, et aussi lorsqu’il y a des données commanditées.

7 Si l’on tient compte uniquement d’une ou deux pratiques isolées il peut sembler que certains fournisseurs de services proposent aux Canadiens des offres intéressantes.

8 Toutefois, il faut tenir compte de beaucoup plus de facteurs que le simple fait d’économiser quelques dollars.

9 Supporters of these types of practices generally contend that they enable consumers to benefit from free or discounted services.

10 Opponents argue that they are inconsistent with the common carriage requirements of internet service providers to provide their telecommunication services to all subscribers without undue or unreasonable advantage or preference.

11 Opponents argue that differential pricing practices enable internet service providers to act as gatekeepers, resulting in undue advantage to certain application providers over others and influencing a subscriber’s choice of content.

12 From this perspective, differential pricing practices by internet service providers are seen as undermining the principle of net neutrality which, broadly speaking, calls for all internet traffic to be treated equally.

13 In a nutshell, and since this hearing is starting on October 31^st on Halloween, the issue before today can jokingly be summarized thus: are differential pricing practices a trick or a treat?

14 En 2015, le CRTC a reçu deux demandes concernant les pratiques de tarification utilisées par Vidéotron pour son service musique illimitée.

15 Dans le cadre de ce plan de service sans fil mobile, les abonnés peuvent écouter de la musique en continu grâce à plusieurs services et ce sans utiliser de données.

16 Comme la portée de cet enjeu dépasse le cas individuel et comme le recours aux pratiques de différentiation des prix comme celles-ci risquent d’augmenter, nous avons décidé de mettre de côté les demandes concernant Vidéotron et examiner l’enjeu dans son ensemble.

17 Ainsi les Canadiens et les fournisseurs de services internet pourront bénéficier d’une politique règlementaire claire et transparente.

18 It’s also worth mentioning that these practices are emerging issues and are also a challenge for other regulators around the world.

19 They will be watching our proceeding as we consider important elements in defining how Canadians access the internet.

20 Lorsque nous avons lancé cet examen en mai dernier, nous avons accepté les observations et les interventions reçues par le biais de nos méthodes traditionnelles.

21 Nous avons toutefois réalisé que cet enjeu revêtait un intérêt particulier pour les Canadiens avides de l’internet, qui tenait leur propos –- leur propre conversation, pardon, à ce sujet sur Reddit. Nous les avons donc rejoints.

22 Je crois que nous sommes un des premiers organismes gouvernementaux à avoir utilisé Reddit en tant que plateforme officielle de consultation et j’en suis très fier.

23 Plus de 1 200 commentaires ont été partagés sur Reddit et nous avons ajouté ces points de vue au dossier public.

24 With the two rounds of written submissions today, parties have addressed the numerous issues raised in the public notice.

25 Most parties recognize that differential pricing practices involve à preference or advantage in favour of the internet service provider, the content provider or the subscriber, or discriminate among content providers and subscribers.

26 During this phase of the hearing we will be focusing on the following fundamental questions: first, can it be demonstrated that any preference or advantage confirmed by these practices is not undue or unreasonable.

27 Second -– and that any discrimination is not unjust; third, how and with what evidence. And finally, what regulatory measures, if any, should the Commission implement.

28 During our consultation some arguments -– some argued that the CRTC should ban data caps. These comments formed part of the record and we will likely hear further arguments during this hearing. We might even ask a few questions about it.

29 While we recognize that differential pricing cannot exist without data caps it would be unwise for us to pre-judge any potential outcome on this issue.

30 And I’ll add -- and despite what you may have read in a dubious article published just a few days ago in The Hill Times, this and no other is the CRTC’s most crucial hearing this fall.

31 You do not drive into the future looking through a nostalgic rear-view mirror.

32 Avant de commencer j’aimerais faire quelques présentations. Le Comité d’audience se compose des personnes suivantes: Monsieur Peter Menzies, Vice-président télécommunication; Monsieur Yves Dupras, Conseiller régional du Québec; Madame Linda Vennard, Conseillère régionale pour l’Alberta et les Territoires du Nord-Ouest; Monsieur Chris MacDonald, Conseiller régional Région de l’Atlantique et du Nunavut et moi-même, Jean-Pierre Blais, Président du CRTC. Et je présiderai d’ailleurs cette audience.

33 L’équipe du Conseil qui nous assiste comprends de Suneil Kanjeekal, coordinateur de l’audience; Carolyn Pinsky et Daniel Finestone, les Conseillers juridiques et Cindy Ventura, Secrétaire de l’audience.

34 J’invite maintenant Madame Ventura à expliquer la procédure que nous suivrons. Madame la Secrétaire?

35 MS. VENTURA: Thank you, Mr. Chairman, and good morning.

36 I would like to go over a few housekeeping matters to ensure the proper conduct of the hearing.

37 When you are in the hearing room I would ask that you please turn off your smart phones, as they are an unwelcomed distraction and they cause interference on the internal communication systems used by our translators. We would appreciate your cooperation in this regard throughout the hearing.

38 Interpretation services will be available throughout the duration of the hearing. We would like to remind participants that during their oral presentation they should provide for a reasonable delay for the interpretation while respecting their allocated presentation time.

39 Le service d’interprétation simultanée est disponible durant cette audience. Nous désirons rappeler aux participants d’allouer un délai raisonnable pour la traduction lors de leur présentation à vive voix tout en respectant le temps alloué pour leur présentation.

40 There is a verbatim transcript of this hearing being taken by the court reporter sitting at the table to my right. Please note that the transcript of each day will be made available on the Commission’s website on the next business day.

41 Just a reminder that pursuant to section 41 of the Rules of Practice and Procedure, you must not submit new evidence at the hearing. If you wish to introduce new evidence as an exception to this rule, you must ask the permission of the Panel of the hearing before you do so.

42 Veuillez noter que les documents seront disponibles sur Twitter, sur le compte du Conseil à @CRTCaudiences en utilisant le mot clé #CRTC.

43 Please note that the Commission was also be twitting the documents during the hearing at @CRTChearings using the hashtag #CRTC.

44 Please note that if parties undertake to file information with the Commission in response to questions by the panel, these undertakings can be confirmed on the record through the transcript of the hearing.

45 If necessary, parties may speak with Commission legal counsel at a break following their presentation to confirm the undertakings.

46 And, now, Mr. Chairman, we will begin with the presentation by the Canadian Media Concentration Research Project. Please introduce yourselves and you have 20 minutes for your presentation. Thank you.

PRESENTATION

47 MR. KLASS: Good morning Mr. Chairman, Vice Chairman, Commissioners, and CRTC staff. Thank you for giving us the opportunity to speak today about telecommunication providers’ differential pricing practices.

48 Before we begin, I would like to acknowledge that this hearing is taking place on the traditional territory of the Anishinaabe people.

49 My name is Benjamin Klass, and I’m a PhD student at the School of Journalism and Communication at Carleton University here in Ottawa. With me today are, to my left, Dr. Fenwick McKelvey, Assistant Professor at the Department of Communication Studies at Concordia University, and to my right, Marc Nanni, a chemist by training but an accomplished privacy expert by experience.

50 We’re here today representing the Canadian Media Concentration Research Project, which is the brainchild of Dr. Dwayne Winseck. Dr. Winseck is unable to join us today, unfortunately, since he’s lecturing in China. He sends his regrets, however, and I believe he even may be listening in from behind the “Great Firewall.”

51 It was almost three years ago today, or this month anyway, that I filed an application with the Commission, arguing that Bell ought to be prohibited from giving its Mobile TV service — a video streaming app — an unfair advantage by exempting it from the punitive data caps that it applies to other similar uses of its network.

52 Since that time, the Commission has found that Canada’s national mobile carriers were unjustly discriminating against their competitors, and that they collectively possess market power. These findings resulted in the introduction of wholesale rate regulation for the first time in the wireless market since the Telecommunications Act was introduced in 1993.

53 The CRTC made a similar finding in the market for wireline broadband when it extended wholesale regulation to fibre-to-the-home networks, again citing the carriers’ significant market power as a reason for taking action.

54 Even more recently, the Commission found that the incumbent network operators have been trying to charge their competitors unreasonably high rates, to the detriment of those competitors, consumers, and the public interest more generally.

55 On the broadcasting side, the Commission has been working steadily towards the goal of modernizing the cable TV industry, which comprises for the most part the same group of firms that operate on the “telecom” side of things.

56 Through its “Let’s Talk TV” proceeding, the CRTC has begun to unbundle cable TV by requiring distributors to offer “skinny basic” packages, by making sure that online video service providers aren’t tied to specific companies’ networks, and by introducing pick-and-pay channels so that customers can choose to pay for only the services they want.

57 The Commission also found that Bell and Vidéotron’s practice of exempting their mobile TV services from data caps ran afoul of the Telecommunications Act’s prohibition against undue preference and unjust discrimination in the provision of telecommunication services.

58 The CRTC’s decision to prevent Bell and Vidéotron from using the control they wield over mobile networks and content in an anticompetitive way was upheld on appeal at the Federal Court this summer, just as its decision to impose wholesale access rules on FTTH was upheld by the Cabinet earlier this year.

59 The basic principle in both cases is that the CRTC has broad authority to take measures that restrict carriers’ market power and anti-competitive practices.

60 So here we are today, and wouldn’t you know it, it looks like some of these carriers have turned a new leaf — they’ve decided to offer their customers services for free.

61 Unfortunately, while “differential pricing practices” might sound good on paper, the fact is that they’re too good to be true. Taking a closer look, what you find is that zero-rating, sponsored data, and similar practices tend to be, in fact, anti-competitive. They stifle innovation. They represent a power grab on the part of carriers who seek to act as gatekeepers to the Internet. They can impose substantial costs on third party service providers, and they raise concerns regarding people’s privacy.

62 No less importantly, the fact that the carriers are starting to use these practices on a wide scale and in conjunction with bandwidth-intensive applications strongly suggests that the data caps they’re placing on other uses of the Internet are arbitrary, punitive, and unreasonable.

63 In short, Mr. Chairman and Commissioners, we’re here with one overarching message: “There ain’t no such thing as a free lunch.”

64 As we’ve laid out in our intervention, one of the central, and perhaps “the” central legal and regulatory principles governing the telecommunications industry is “common carriage.”

65 Common carriage, in essence, is a principle of non-discrimination. It exists to ensure that those who own the infrastructure of commerce and communication do not unduly influence the ways in which people use that infrastructure. It comes to us from ancient Roman times, through the British common law, and was introduced to Canadian communications jurisprudence as early as 1891.

66 And while that may seem like ancient history, I’d like to point out that that very decision that took place in 1891, the Supreme Court of Canada’s case of Electric Despatch Co. v. Bell Canada, formed the foundation of the Supreme Court’s 2012 decision on the ISP Reference.

67 This decision held that ISPs are not broadcasters, even when they provide access to broadcasting content over the Internet. That decision essentially confirmed that common carriage is a valid and essential principle in the age of the Internet. The same principle was in operation in the court’s upholding of the CRTC’s Mobile TV case.

68 Today, the main tenets of common carriage are found in the Telecommunication Act, sections 27(2) and 36, which prevent Canadian carriers and other TSPs from unduly discriminating against, or offering an undue preference toward, any person, and prevent network operators from controlling the content or influencing the meaning or purpose of tele-communications.

69 We note here that common carriage does not constitute a blanket provision against all forms of discrimination. For instance, there may be good reason to treat dangerous goods with special care or to prohibit their transport. And certain social groups, such as seniors, students, charitable organizations, or disadvantaged segments of the population may often merit special treatment. However, common carriage serves as a guiding principle that places a heavy onus on carriers to demonstrate that discriminatory practices are justified and in the public interest.

70 In brief, common carriage is so important, and we place such a heavy emphasis on it in the context of today’s hearing, for three reasons.

71 First, it serves to protect and promote competition and innovation in the communications sphere, both amongst carriers themselves and also on the broader Internet. Second, it ensures that choice and control over how people communicate and who they communicate with rests with those people themselves, and not with network owners. And, third, it protects people’s privacy.

72 The increasing use of differential pricing by network operators challenges all three aspects of common carriage. Marc.

73 MR. NANNI : Good morning. My name is Marc Nanni. Continuing.

74 To be clear, DPPs are a form of discriminatory pricing where an ISP charges its customers different rates according to which online applications, services, or other types of information and communication the person wishes to access.

75 On the face of it, DPPs sound innocuous. Carriers like Telus and Vidéotron have likened them to coupons, or discounts, and argue that they’re beneficial for consumers.

76 While they may sound good at first, DPPs have nevertheless attracted a significant measure of controversy over the past several years, here in Canada and in many countries around the world as well, and for good reason.

77 A number of companies, including Bell, Telus, and Vidéotron, but notably excluding Rogers, argue that there is no evidence showing content-specific DPPs could lead to competitive harm and downstream markets. This argument should be dismissed by the Commission, as it was in the mobile TV decision.

78 The fact is that telecommunication regulators have long recognized the potential for discriminatory pricing to stifle competition in downstream markets.

79 This is true particularly in cases where the discrimination is tied to specific providers of content or classes of content, as our intervention has shown. From an economic point of view, this is precisely the reason why common carriage is important.

80 Reviewing the historical and international record, as we have done in our intervention, demonstrates numerous instances in which regulators have acted to prevent the harm that would inevitably result from such practices, beginning over 100 years ago and in recent years, from Canada to the U.S. to India and in numerous countries in between.

81 The common thread running through all the cases we’ve reviewed is that common carriage was upheld as a means to ensuring competition, innovation, and peoples’ ability to freely communicate.

82 Preserving online service providers’ ability to access audiences and customer bases without interference from gatekeepers in local access markets is simply a prerequisite for competition and innovation to occur. Otherwise, deep-pocketed players in established positions, such as Facebooks and the Googles of the world, will simply pay to play, while smaller providers will not get the chance to upset the status quo.

83 As we’ve noted, common carriage isn’t just about commercial concerns. It also protects people’s communication and expression rights, their ability to decide who they communicate with, and under what terms they decide to do so.

84 MR. KLASS: An open, non-discriminatory internet allows anyone with a connection to reach anyone else, on even terms, whether they’re another person, a business, a government -- who knows, maybe even they might be a dog.

85 Cable TV, on the other hand, gives users access to a limited variety of pre-selected channels with the choice to access extra channels for an added fee. That may be the reason why people are increasingly choosing the internet as their primary means of accessing audio-visual content.

86 The key difference between the two? Common carriage rules prevent ISPs from acting as gatekeepers, while gatekeeping -- that is control over content -- is at the centre of the broadcasting distribution market.

87 Some of the companies, such as Vidéotron and Telus, have argued that content-specific price discrimination doesn’t give them the ability to control the content or to influence their meaning or purpose. This would be contrary to section 36 of the Act.

88 Instead, they argue, they are just providing users with more of what they want.

89 These assertions, in our view, are plainly false. It’s a basic economic fact that price signals influence behaviour, and that’s exactly what differential pricing practices are all about.

90 By Vidéotron’s own admission, differential pricing practices are intended to serve as a promotional vehicle, driving traffic to service providers that a network operator has selected to be part of its differentially-priced program.

91 Promoting particular services, driving traffic to specific applications, if this isn’t influencing the purpose of people's telecommunications, I don’t know what is.

92 What’s more, despite Vidéotron’s claim that it is open to any online streaming music provider, the CBC remains conspicuous by its absence, as do the services of radio broadcasters who are also parties to this proceeding. What this all goes to show is that the service is in fact exclusionary; in other words, that Vidéotron is acting as a gatekeeper.

93 By introducing DPPs into the mix, ISPs and mobile carriers are trying to act as editors, publishers, or broadcasters, plain and simple. They’re treating the internet like cable TV, at a time when people are increasingly demanding the opposite.

94 I sat in the courthouse during the mobile TV appeal in Toronto this past January and listened to Bell’s lawyer argue that if Bell wasn’t a broadcaster, its mobile TV service would be in violation of section 36 of the Act because it clearly controls the content of the service.

95 This position is, however, unsurprisingly absent from Bell’s submission to this proceeding.

96 In fact, Bell’s primary argument throughout the mobile TV proceeding was that it is a cable TV provider, not an ISP, and so it doesn’t have to abide by the Telecom Act. A revealing position to be sure, and one that was thankfully rejected by the CRTC and the Federal Court.

97 The ISPs’ motivation for acting this way isn’t hard to understand. They want to get more value out of customers by offering more of something that costs next to nothing to provide.

98 But by attempting to assert their own rights as editors, broadcasters, or speakers, the ISPs wrest control away from users, which we submit is clearly a one-sided trade-off and not one that favours the public interest.

99 So how can the companies argue that treating the internet like cable TV increases users’ choice, when you can already access anything you want and communicate with whomever you want over the internet on even terms?

100 You’ll have to ask them over the course of this proceeding because I don’t know the answer.

101 Fenwick?

102 DR. McKELVEY: Thank you very much.

103 I understand some of the -- my argument here will introduce new evidence and I'm wondering if I a have permission to introduce this evidence in support of our submissions earlier?

104 THE CHAIRPERSON: Why were you unable to provide this evidence earlier?

105 DR. McKELVEY: It's a great question. Working with a -- by any serendipity honestly, I'm working in a new project working in measurement in Canada trying to understand and apply new internet measurement techniques to understand the operations of systems such as differentiate service provisions.

106 And it was only in early September that I knew of a methodology that is going to be presented at the internet measurement conference this November 2016 that I was -- found out about it and it was basically over the past last month that I was able to actually deploy and use this measurement and try to support arguments we've made previously as part of our submission.

107 THE CHAIRPERSON: Okay. Well, we'll take your arguments under advisement but go ahead and ---

108 DR. McKELVEY: Okay.

109 THE CHAIRPERSON: But that doesn’t mean ultimately we will necessarily admit it on the record.

110 DR. McKELVEY: That's totally fine and I'm happy to elaborate on questions. Certainly it's not my intent at any point to undermine these proceedings.

111 THE CHAIRPERSON: Sure. Please continue.

112 DR. McKELVEY: Differential pricing practices also raise privacy concerns related to the use of deep packet inspection and other technical internet traffic management practices.

113 Access providers clearly require greater awareness of their customers’ activities to engage in DPPs that they otherwise would. Such practices require methods of deep packet inspection, and the data generated from these techniques can flow to places where it doesn’t belong or be used in ways that are not permitted.

114 Since September, I’ve been investigating how Vidéotron provisions its zero-rated “Unlimited Music” service. My methodology involved signing up to a Vidéotron service, reviewing my contract, and finally conducting tests of the service.

115 I was primarily concerned with two aspects of the contract.

116 First, Unlimited Music in the contract "excludes the downloading of songs and other non-musical content". This detail above suggests that the service requires or could be using something beyond IP-based classification that Vidéotron described in its prior submission to the CRTC.

117 Second, that Unlimited Music "allows a maximum flow of 128 Kbps". This factor suggests to me that the service might be tied to technical internet traffic management practices such as throttling.

118 To investigate these concerns, I collaborated with the same team of academics and researchers from Northeastern University and the University of Massachusetts who recently audited T-Mobile’s Binge-On service.

119 The method we used manipulates packets to understand how providers apply different treatment to network traffic.

120 I worked closely with this team to test my own Vidéotron service. Specifically, we sent packet flows containing data that resembled music services that Vidéotron zero-rates to the wrong IP address, the idea being that Vidéotron, if it was only using IP identification as it claims, these packets would be counted against my cap.

121 The result was surprising. Despite being sent to the wrong IP address, the data was zero-rated anyways. These findings strongly suggest that Vidéotron is inspecting application layer data; in other words, that it is likely examining the application layer data using deep packet inspection as part of its Unlimited Music service.

122 These findings raise concerns that DPI devices previously used for traffic management practices are now being repurposed for unrelated marketing initiatives such as the Unlimited Music service, and that information gleaned from these devices is being connected to billing and accounting systems.

123 While I do not find any evidence of throttling in Vidéotron’s case, the fact that T-Mobile has been found to throttle in connection with the Binge-On shows that it has happened elsewhere.

124 These findings strongly suggest that Vidéotron’s unlimited service is not in line with the expectations laid out by the Commission in the ITM Framework and runs counter to common carriage principles more generally.

125 I am confident in my methodology, but I do wish to stress that these tests point to the need for further investigation.

126 And I’d be happy to answer further questions after our presentation.

127 MR. KLASS: In the time that remains there’s just one more concern that I’d like to address, but in my view it’s a big one, the data caps.

128 Virtually all of the network operators argue that the data caps are outside the scope of this proceeding and should be ignored. However, civil society groups, academics and individual intervenors have disagreed. And perhaps most importantly, the comments of hundreds of internet users on the CRTC’s Reddit forum were unanimous. I cannot find a single commenter who supported the continued use of data caps. On the contrary, many people asked the CRTC to limit or outright prohibit the caps.

129 In 2011, Michael Geist estimated that a carrier’s cost for transferring 1 gigabyte on a wireline network was 8 cents on the high end, a number that has surely come down since then; yet carriers like Bell and Rogers continue to charge in the neighbourhood of $1.50 to $3 per gigabyte in overage fees.

130 From this perspective, it looks an awful lot like data caps and overages render the effective price of a megabyte at many times its cost to deliver. This does not reflect a competitive outcome, plain and simple. However, what would normally be considered a market failure has been accepted, begrudgingly, as the debates on usage-based billing that have been going on for years demonstrate, as a means of managing traffic, an economic internet traffic management practice.

131 But that justification went out the window when Bell started offering access to 10 hours of video from sources it selects for an eighth of the price it charges for the same amount of mobile Internet access to other sources.

132 If the carriers aren’t worried about zero-rating -- zero-rated traffic causing congestion, how can they contend that other traffic travelling through the same pipes is going to? And if that traffic isn’t causing congestion, how can they justify continue to using -- how can they justify to continue using price signals that were set above competitive levels in order to dampen use?

133 Internet users have an economic interest in paying a fair market price for the services they purchase. And competition is supposed to deliver fair market prices by driving them towards marginal cost. But zero-rating and other differential pricing practices are a big red flag indicating that that just isn’t happening.

134 In our view, you cannot have a discussion about differential pricing without acknowledging these problems.

135 Today I don’t know and I don’t have an answer on whether the Commission should ban data caps across the board or whether that’s the right thing to do, but I do hope you’ll recognize that data caps are a serious issue that needs to be addressed and that it gets the attention it needs throughout this proceeding.

136 In conclusion, we hope the Commission will prevent carriers from continuing to engage in content or application-specific differential pricing practices, for which the downsides far outweigh whatever benefits the carriers think they have.

137 We also hope the Commission will listen to Canadians who do not deserve to have unnecessary limits placed on their ability to freely communicate.

138 Thank you for your time.

139 THE CHAIRPERSON: Thank you very much. And Mr. McKelvey, just to be clear, the little discussion we had on the evidence didn’t count against your 20-minute presentation time.

140 So thank you. A bit like an exemption to data caps, I guess.

141 Now, there’s three of you, so I’m going to leave you to manage who answers the questions. And I’m going to suggest as well that if one of you is answering and the other ones don’t pipe up, we’re going to assume the other people on the panel will agree with the answer provided by one of you; is that okay? Yes? Okay.

142 So in your written presentation and again this morning in your oral presentation you’ve made a distinction, which I found interesting, between viewing telecom carriers on the one hand with respect to their internet service provision as either under a publishing model or a common carrier model. And I was wondering if you could perhaps elaborate on the important distinction between those two perspectives vis-a-vis the ISPs.

143 MR. KLASS: The basic distinction, as I understand it, is that an ISP acts as a mere conduit, as a neutral carrier of other people’s communications. And it’s that those people make decisions as to which sources of information, which other people they connect with.

144 Whereas, an editor or a publisher chooses pieces of information, news articles, television programs to display to an audience.

145 In the first case, there is no measure of editorial control on the part of the person who owns the network; whereas in the second, it -- control over the network lies in the network itself. It’s a center out perspective from the editorial side and it’s a sort of a dispersed or distributed system of control on the ISP side.

146 So I mean, I think that that’s the distinction I see between those two things.

147 With respect to the Internet, if prices are equal for all different sources of information, the people who are using that don’t have to reckon with the consequences of choosing one website over another, for instance, or one online video streaming service provider over another on the basis of the price that their network operator charges them.

148 Whereas, if the network operator starts to select different services to receive preferential pricing or to not have a price at all, the customer has to start thinking to themself, am I going to view this particular content if it’s more costly to me, or might I choose something that the network owner, who in an ISP model wouldn’t have any influence, has preselected.

149 THE CHAIRPERSON: And why is that important? Is it solely -- that distinction solely important for the three points you’ve identified under -- at page 5 of your presentation today, those three items, or are there more reasons why that distinction becomes critical?

150 MR. KLASS: Well, I mean, I think that the three distinctions that we’ve drawn or the three main points that we’ve highlighted, being influence over competition and the ability of firms that rely on intermediaries to succeed, people’s ability to choose the information sources and who they communicate with on their own without influence from network carriers, and people’s privacy, are the three main concerns that would be associated with common carriage.

151 THE CHAIRPERSON: The main concerns or the sole concerns?

152 MR. KLASS: Well, I mean, I think that those are the concerns that we’ve elaborated using the examples in our argument here and in the written documents, but I mean, I wouldn’t limit it -- you know, I wouldn’t say that they’re the sole concerns.

153 THE CHAIRPERSON: Right. Well, perhaps in the further steps in the proceeding you’ll be able to elaborate if you think that there are other concerns ---

154 MR. KLASS: M’hm.

155 THE CHAIRPERSON: --- beyond these three. That’s fair enough. But these are the main concerns in your view.

156 Now, when one looks at the distinction between this publishing model versus the common carrier model, help me understand why it’s not just a relevant issue when looking at vertically integrated companies that own or control content as opposed to all ISPs.

157 MR. KLASS: Sorry, the question was whether this is a concern only for vertically integrated carriers or ---

158 THE CHAIRPERSON: Right. That would own content as well.

159 MR. KLASS: Okay. So I think in the case of the Bell Mobile TV case, Bell is a vertically integrated carrier and in that proceeding I argued that those carriers should merit specific concern. The incentive to favour a company’s own content or affiliated operations I think is fairly straightforward. It’s fairly obvious. It isn’t as much so in the case of a company like Vidéotron, which is vertically integrated, but in the connection with its unlimited music service it isn’t giving a benefit to its affiliated content operations. However, it still is picking and choosing various different sources.

160 So while Bell’s commercial motivation to give its own TV stations an advantage is obvious, because if more people view their TV channels they get more advertising revenue, they get subscription fees from the mobile TV service and so on. There still is an advantage being conferred upon a select group of service providers in the case of Vidéotron’s unlimited music.

161 I’ve raised a number of times over the course of this proceeding the exclusion of CBC and radio broadcasters from Vidéotron’s service. And, I mean, the CBC has an app, unlimited music streaming online. They’ve got 40 channels they advertise for it, which I understand you guys have outlawed on radio too. But it resembles -- like if I’m making a choice of what music streaming service I want to listen to online, those are substitute goods, CBC or Spotify. But if Vidéotron is saying I don’t have to pay a data fee to listen to Spotify and I’m trying to make that choice, that’s going to influence me, regardless of the fact that Vidéotron doesn’t own Spotify. It still created an advantage for that company.

162 And it’s -- I think it’s been explicit in the record of this proceeding that it wants to have reputable commercial music providers as part of its service offering, and it’s excluded a number of radio-broadcasters. There’s a coalition of them that submitted to the earlier proceeding, as well as the CBC.

163 And that -- I mean, that’s an editorial choice. It seems arbitrary to me. Why should I have to make that type of decision when all these things are equally available on the network?

164 THE CHAIRPERSON: You mentioned in answering that question, music-based radio services or music-based other, like CBC, music service.

165 Why do you stop there? And these are streaming services on top of that. What would happen -- what’s your view with respect to one could examine, for example, have a podcast of music, not necessarily a streaming service; something that somebody goes and gets.

166 There are radio services that are all talk, all sports. That distinction that some parties are saying that as long as you treat all similar types of service in a similar way, apparently gets us around unreasonable and undue. What are your views on that?

167 MR. KLASS: I think there’s two ways that you can look at it and from a technical perspective, it’s all just bits.

168 So Videotron isn’t doing anything special with regard to music from a technical perspective. Someone in the company is saying, okay, it’s music that we are going to favour here, and people can use entertainment services as much as they want. But if they want to listen to the news or if they want to listen to a podcast or to use YouTube to find out how to fix the drywall on their house, they’re going to have to pay an extra fee.

169 From the second -- so that’s sort of an economic consideration. It’s also a bit -- it’s a speech consideration because Videotron is saying that music is a certain kind of speech that we think is more valuable than other kinds of speech to people; that they’re demanding that this is a service they want, more than, say, a podcast or something else.

170 That gets you into this issue of content control and influencing the purpose of communications. So as I understand it, these are separate but connected issues.

171 On the one hand, Videotron is creating a preference for whatever class of services that it’s chosen, despite the fact that when it’s fulfilling its function as a carrier, it’s doing the same thing for both of those things.

172 And the second is that it’s making this editorial decision. It’s saying, “we think we know what people want. We’re going to use a pricing mechanism to drive people towards that or to fulfil our perception of what their demand is.

173 But it involves a value judgment about what kind of speech people should be -- should be or what they think people want to be doing on the internet. And I think Fenwick had something to add.

174 DR. McKELVEY: Thank you very much.

175 I’d say, to the prior question, there continues to be a question not only about vertical integration but also of convergence in general. And this is a widespread issue in telecommunications and broadcasting, as the two are merging constantly.

176 And one of the -- you know, to echo the Commissioner’s comments about moving forward in the future, how do you develop policy that’s appropriate for the internet? And I think that in my reflections, and this is something that as an academic you’re always questioning or thinking through, but there is a growing evidence that there’s a need to distinguish between, say, broadcasting-like activities and telecommunications-like activities.

177 And this is certainly, you know, as I’ve participated in the Discoverability Summit, there’s a need to try to say what are next generation broadcasting questions and what are next generation telecommunications questions.

178 And to me, zero rating fits into the latter category of next generation telecommunications questions, because it really, you know, begins to raise questions about what has been the history of common carriers and what are the responsibilities about providing a technical conduit; and the responsibilities of being an impartial or respecting privacy and respecting people’s control over how they communicate.

179 And to the second point about, you know, some of these concerns is that in order to provision this and in order to provision something like music, to be able to differentiate between music or Videotron, at least in the study of T-Mobile’s Binge-On, it requires a greater awareness i-telecommunications companies of the activities of their customers.

180 And I think it’s an important question to raise early in the proceedings if we’re going to allow this type of practice. What are the privacy obligations and responsibilities?

181 You know, the simple truth is that if you’re, you know, in some ways that -- rating can be administered, it requires the inspecting of, you know, their customers’ packets to lower levels or application levels specifics to be able to say this is the music services and not. And it becomes, I think, an incredibly complicated and potentially fraught responsibility placed on telecommunications providers to suddenly be responsible for deciding what counts as music and what doesn’t count as music.

182 And I think that even in the review as Videotron’s music unlimited service, it’s interesting in reading the fine print, they allow certain services but they prevent listening to sports podcasts.

183 And so it becomes an incredibly complicated, I think, system to administer at some times and how it’s technically administered raises, I think, legitimate privacy questions.

184 THE CHAIRPERSON: Well, let me build on that because you’re suggesting, and I wouldn’t disagree, that we have to look at this from a future perspective. What should be down the road.

185 So globally, what outcomes do you think we should be focusing on through this process? I notice, for instance, at paragraph -- in the Executive Summary, paragraph 16 of your Executive Summary, and I’m quoting here that we shouldn’t -- the goal is not

186 “to hobble the advent of the internet and mobile wireless networks as rival distribution networks for audiovisual media and entertainment.”

187 Is that the ultimate outcome or is there more?

188 MR. KLASS: In that passage, I believe, we’re referring to the FCC’s various determinations to protect the internet as an open communication channel as against carriers who are what we call diagonally integrated, that is they operate across different segments of the same markets.

189 So whereas vertical integration refers to a situation in which, say, a company owns content and the means of distribution, diagonal concentration would be a situation in which a company owns related but distinct segments of the same markets. So what I’m talking about here is a wireless network and a wireline network or Cable TV and internet, where Bell or Rogers are seeing their cable subscribers growth decline and the actual absolute numbers of cable subscribers starting to go down.

190 This is a revenue stream for them and they’re concerned that their margins are decreasing, the overall revenue is decreasing and that they’ve got this other part of their network that is basically gobbling up a revenue stream that they’ve already had, right?

191 So the incentive structure, when those two companies are the same, is not -- it’s different from when those companies would be separate. If I’m a cable company alone and I’m losing subscribers, I’m worried, right? But if I’m a cable company that also owns an internet company, I’m seeing an opportunity to make up for those losses.

192 And so I think what we are trying to worry about here, what we’re trying to express is that a company who has those types of concerns shouldn’t be hobbling what would normally be an internet business that is just purely designed to attract people to its business; it wants people to use as much internet as they can without worrying about the fact that those same people are taking revenues away by leaving the cable system.

193 THE CHAIRPERSON: Right. Fair enough.

194 I guess my question isn’t so much about what are the potential risks as what are the positive outcomes do you think will flow from the position you have together taken in terms of perhaps adoption of the internet more broadly in society?

195 The deployment of networks, either wireline or wireless, investment in network deployment, the promotion of competition, privacy. I’m looking for the positive outcomes. I know you’re mentioning risk but what are you trying to achieve?

196 I realize it’s a 100,000-foot question, but that’s how I start and then we move on to more detailed questions afterwards.

197 MR. KLASS: I mean I think, yes, from 10,000 feet, what I would like to see is a network where the default is people use it for their own purposes, where it’s open to not just people who want to use it as a platform for a speech, or a platform for commerce, or to watch TV but that they can -- a platform where a person with a good idea can connect with audiences.

198 This is the end-to-end principle that you maximize the number of people on the network and it increases the value in a disproportionately large way to how many people jump on. It’s called a network effect. And I don’t think you get that when you have an editor in the centre making decisions about who can use the network in whichever particular way they want.

199 So keeping it open maximizes the benefit to society as opposed to a system that is designed for a particular use by a particular group of people with a particular set of ideas.

200 Fen?

201 DR. McKELVEY: I will admit that I often lack or long for a 10,000 view perspective on some of these things, and so I’m very actively trying to work on understanding just the scope of this.

202 To me -- to say two things -- is that one quite specifically is that these largely are the questions in my mind about common carriage is an important principle to preserve because it, I think, is the historical one of understanding how communication systems function, and that in the future this is something that, I think, we’ve seen borne out, it’s an important principle to preserve to prevent undue market power and to be able to kind of set limits on questions quite broadly about control and responsibilities.

203 I think that in my own observations you’re watching it becoming much more easy and much more effective to kind of manage and manipulate internet traffic for very legitimate purposes, and I think that this is an important check on some that direction.

204 And the second point I’d make in terms of -- is about these concerns about privacy and about network investment. Certainly there is an ability to be more aware of customers’ traffic and common carriage gives some legitimacy to why that should be protected.

205 And so I think that -- well, one of the things to say is in the future it would be great is, you know, there are ways that ISP’s don’t feel responsible or have some sort of legitimacy about giving or granting their customers privacy and I think that’s the fair objective.

206 And the second point is simply that in looking at how some of the zero rating might be a provision, this is a network investment. This is something that’s being invested in the network. And so I think one of the questions before the Commission is about what type of network investment do we want to incentivize and do we want to incentivize investment in infrastructure and infrastructure that’s capable of providing more significant zero rating services, or do we want to incentivize investment in capacity and investment in reach.

207 I think that, you know, to me, the question that hasn’t necessarily been answered is, you know, how much we’re seeing investment in zero rating services at the detriment potentially of investments in other aspects of the infrastructure.

208 I think that, to me, you know, quite honestly, if you’re thinking about where the internet needs to move in terms of the direction, where do you want it to go, I think, you know, reach, scope, affordability, these are things that I’ve had the privilege of being able to participate at the Commission about, and, to me, I don’t see as much a benefit, you know, and particularly because these questions that you are rating weren’t something that was present in my own anecdotal experiences at the basic services offerings, you know, as a network investment.

209 And so in some ways I think it’s important, you know, to send signals about investing in network infrastructure that protects customers’ privacy and that tries to find applications that preserve the common carriage principles.

210 THE CHAIRPERSON: Now, you mentioned that there may be legitimate purposes. We’ve heard from some parties in this proceeding, particularly with respect to content made by Canadians, often referred to as CanCon. In your view, is there a countervailing public policy argument that that type of content, or the promotion of it, or the availability at lesser cost because they would not count against data CAM, does that make sense?

211 Is it a public policy objective that can validly be balanced against the other public policy objectives you were mentioning such as net neutrality?

212 MR. KLASS: Some groups, certainly the CMPA, seem to be making that argument, but, I think, any such judgment is going to necessarily involve only a subset of the cultural products and the availability of various different types of Canadian content and, you know, were the policy -- were the objectives of public policy, then or otherwise, already available.

213 And you mentioned that it makes things -- it could be something to pursue because it makes things cheaper and it could place Canadian content at an advantage against other forms of content, which reminds me earlier you asked are there any other policy considerations or any other important considerations with respect to common carriage.

214 And, I mean, I can’t believe I missed the most obvious one, and its affordability and price. And common carriage is -- it also involves the idea that rates for these services are just and reasonable because society recognizes that they’re valuable for everyone to access.

215 The Commission has forborne from exercising it’s -- the traditional method that it’s sought to achieve this goal. As I mentioned in the end of the presentation, the market is supposed to provide a situation in which rates are affordable or they’re driven towards costs which is considered the fair price.

216 And a lot of the companies are coming in here and saying this is a net benefit for people, we zero rate things, it makes things cheaper, it brings people onto the network and so on, but I disagree strongly with that argument.

217 First of all, this is not the same as what happened in India with Facebooks re: basics or what’s happening in other parts of the world where large internet companies are striking deals with network operators to say we’ll provide free access to people who don’t have to sign up for a mobile plan to access the network, they’ll just be able to have a device and for free access the content that we’ve selected.

218 Now, that has its own problems but it certainly does have the argument that people who otherwise would have no connectivity can at least have some.

219 That’s not what we’re seeing here in Canada. We’re seeing Videotron offering service -- unlimited service only to subscribers of its premium plans. You already have to be able to pay to get your foot in the door before you can start accessing these free services that Videotron is suggesting are so beneficial.

220 To suggest that giving people data for free once they’ve already paid the rate for access to the network is kind of like saying a prisoner should be happy that we’ve stuck them in a room with a window or a cell with a window, okay.

221 If carriers are allowed to pick and choose the various different sources of information that are zero rated, it also allows them to keep the caps. It gives them an incentive to keep the cap high.

222 I think a lot of people, particularly in the media, seem to have forgotten that Rogers was originally part of the mobile TV proceeding because it was also zero rating, it’s any place TV service. It voluntarily stopped doing that before the proceeding was over. But in their arguments in that hearing they made no bones about the fact that this is an experimental service. The price that they’re offering $5 for 10 hours or thereabouts is an introductory price that they’re intending to offer customers to induce them to purchase a service they’re not familiar with and once people start using it more they raise the price. Like this is on the record of that proceeding.

223 So I think that the idea that you can take Canadian content and put it in this free tier of service, you know, that’s a worthy goal but the goal we should really be striving towards is making the whole network affordable for people so that they can choose not just a subset of Canadian content that some carriers have decided is worthy of being put on a pedestal but whatever Canadian content they want, which might be on You Tube, it might be a podcast, it might be a blog post, but it’s certainly not something that they have to select from a pre-set package of things that are cheap while they have to, you know, limit or meter their use of all other things.

224 THE CHAIRPERSON: Right. A very complete answer but perhaps not to the question I asked. The question I was asking is, is there a public policy case to be made with respect to CanCon that it should be given a special treatment that you sort of go a little bit more on the publishing side than on the common carrier side when it happens to be Canadian made content?

225 DR. McKELVEY: Thank you very much.

226 And it’s a challenge because this is -- it feels like there’s two worlds that I keep isolated. And I’m sure this is probably a common sentiment in the Commission about the broadcasting world and the telecommunications world and watching this kind of bleed back and forth.

227 I think the honest challenge is that is this the best forum to have that conversation about. I think the two points I’d make, quite honestly, is there’s great debate right now about what’s proper cultural policy going forward. I myself am trying to participate in that debate with great challenge about questions about Canadian content.

228 And certainly in participating in the Discoverability Summit and actually being, you know, committed to that term as a way of describing some of the challenges we have in kind of next generation broadcasting, there are concerns about how we discover Canadian content.

229 And these ones have to do with the accountability of large platforms like YouTube, these have to do with, you know, the influence of algorithms. It’s a really complicated conversation.

230 And I think it’s really hard to have it potentially in this forum, you know, which is one about telecommunication providers making claims about kind of the underlying conditions of the network itself.

231 And the -– so I would say that there are potentially, you know, public policy investigations to have about that.

232 I think that it’s hard to make them within a kind of tradition of common carriage. And I think, you know, also quite honestly are we convening the, you know, all the parties that are part of that conversation.

233 And, you know, it’s interesting in Bell’s submission they described their project as “improving discoverability of digital content”.

234 And I think that that raises, you know, legitimate concerns about how it’s seeing what this is and whether you’re watching a common carrier telecommunications company moving into having an ability of recognizing a network ability of discoverability and influencing that.

235 So all I’ll say is that, you know, the key takeaway is that, one, there’s a lot of contention about that.

236 And I certainly respect the idea that there are ways of promoting Canadian content online. I’m not entirely convinced that the best way is zero rating, as a way of doing that, but I don’t want to foreclose that conversation.

237 And I think the challenge is that this is very similar conversation that happened at the basic service offerings that remember about what counts as kind of a minimum of basic service.

238 And it’s a hard position for the Commission to be in to start picking and choosing what types of traffic are eligible for zero rating or not.

239 And I think that that’s one that raises a wide public policy debate and in some ways, you know, it’s difficult to have that effectively done where there are legitimate concerns about telecommunications and common carriage.

240 THE CHAIRPERSON: Right. I’m a little bit surprised by your answer, because I would have thought that a clear more principled approach would have been say, in your position, would be not that we’re against Canadian made content, but that the principle is mere conduit. Mere conduit.

241 And whether we’re supportive or not through other means of Canadian made content, that principle is the one that one should not be playing with.

242 DR. McKELVEY: I mean it’s a much nicer, eloquent answer than the one I gave, but I think that there is –- the truth of that is that well there’s an emphasis on our presentation about common carriage.

243 And then when asked to talk about this issue a bit more broadly I am saying that there is a variety of things that could take place.

244 But that yes, I think your point -- and yes that common carriage is the overriding issue in this particular discussion about Canadian content and public policy.

245 And yet I just want to be sympathetic that I do think that it is an issue and I don’t want to dismiss that issue. I just don’t think that common carriage –- that I think it’s, you know your point, is that the conduit is, you know, is the central concern here but -– yes.

246 MR. KLASS: I’m going to try and answer the right question this time.

(LAUGHTER)

247 MR. KLASS: I think I can say unequivocally that it’s the position of the CMCRP that the best means of achieving cultural policy is cultural policy and not telecommunications policy.

248 The thing that got us -– I mean the -– the thing that started the ISP reference was a reference question from the CRTC, our ISP’s broadcasters, and it was in the context of the possibility of taxing ISPs in order to provide into the diminishing fund to produce Canadian content. The answer from the Supreme Court and the Federal Court of appeals was no; you know?

249 And today in one of the companies there was arguing against the proposition that ISPs should be treated as broadcasters, I believe it was Bell.

250 So when the question is should we have to pay into a tax, should we face obligations related to Canadian content through telecommunications policy the companies they don’t want to, you know, nobody wants that.

251 And I think that it’s best to deal with cultural policy without throwing this principle of neutrality out with the bath water, I suppose.

252 THE CHAIRPERSON: Okay, thank you.

253 As I was reading your documents in your presentation, you refer to a lot of the history around common carriers and as you pointed out it goes way back in its application to telecomm carriers and then its application to internet telecomm carriers.

254 I was struck that the origins of it was largely to the benefit of the common carrier, as it protected them from liability.

255 And along the way that shield from liability, especially in Common Law, got converted into -– well got it added on, not converted, but something got added on with respect to a positive duty of not discriminating unduly.

256 Is it your view therefore -– well first of all do you agree with that analysis, you don’t have to if you don’t, but assuming that analysis is correct, could a common carrier stop acting as a mere conduit if it was ready to assume the liability associated with being a common carrier?

257 For instance right now there is a shield against defamation, a shield against certain copyright infringement and so on.

258 What would your view be if the common carrier, the erstwhile common carrier, decided no I’m stepping on the other side of the fence? I’m going to be a publisher and as a publisher I also have all the burdens of that?

259 MR. KLASS: I think that the carrier -– I mean that’s a question of convergence. We have so much vertical integration here in this country and it looks like the trend may be following in others, but the answer to that question is a yes with permission.

260 In the -– throughout the mobile T.V. proceeding Bell was -– Bell held that it has the rights of a broadcaster when it wants and it, you know, it can escape the common carriage obligations, but you know in other cases it doesn’t have to follow those obligations that are associated with liability and so on. You know, it wanted to have its cake and eat it too.

261 Section 36 of the Act, the Telecommunications Act, that is, says except where --

262 “Except with prior permission from

263 the Commission a carrier shall not

264 control the content or influence the

265 purpose or meaning of

266 communications.”

267 Or something to that effect.

268 That section is not subject to forbearance. The carrier has to ask permission from you and in discussions that I’ve found about this section, which are fairly sparse throughout the sort of regulatory history, what’s available on the website in any case, particularly in the review of regulatory framework 1994-19 which, you know, was the big review that took place after the Telecomm Act came in.

269 They discussed the issue of telephone companies and common carriers entering into content provision, information service provision and broadcasting activities.

270 And every time the Commission says, you know, we consider that this is something that we want to encourage, at the time the telephone companies were not dominant internet providers.

271 They were still discussing things like video dial tone, but the expectation always was that they would do this on a common carriage basis and where they decided to become involved in broadcasting they would have to ask permission from the Commission. They would have to obtain the licence required of broadcasters.

272 So I think the answer is that yes a common carrier could -– can enter into the field of broadcasting, but two things have to happen.

273 One, it has to follow the rules that apply to broadcasters and number two is that it cannot continue to avail itself of the shield that common carriage provides against things like liability.

274 THE CHAIRPERSON: At paragraph 95 of your written submission you are, I think, particularly critical of the positions in recent studies of the C.D. Howe Institute, of Fraser Institute and the Montreal Economic Institute.

275 I wouldn’t mind you elaborating on that. And tell me if I’m reading you correctly when I would, I think -– I don’t want to put words in your mouth, but you would be saying that those studies that want to put a lot of emphasis on competition principles, of ex post analysis of whether something is anti-competitive or not, are perhaps missing the essential nature of tele-communication common carriers in that they are common carriers, and therefore, there’s a special overlay above and beyond merely anti-competitive; that they have a special positive duty. It’s one thing to apply ex post competition law analysis to, I don’t know, a shoe manufacturer but when you have a common carrier, there are other issues at stake. Is that -- would that be correct? I’m reading your criticism of those three institutes’ studies?

276 MR. KLASS: Yes, yes. I mean, I think that when we’re talking about communication, we’re talking about people’s speech. We’re talking about people’s ideas. We’re not talking about what colour of shoes that they have decided to put on today. We’re talking about people’s ability to interact with others in society; increasingly to interact with democratic institutions; to realize their value as human beings. And I think that that is not something that a nice tidy assessment of whether a carrier has the ability to raise the price above competitive levels in a relevant market can capture.

277 THE CHAIRPERSON: But it has become fashionable in some quarters to criticize the CRTC in saying it’s an old institution not required anymore, that you should just rely on broad competition policy proceedings and laws.

278 MR. KLASS: I think that the competition authorities have already got their hands full in a number of industries and I agree with some of the things that they do and some of them that they don’t. But this -- the level of technical complexity, at the base, for the communication’s industry, the pace with which it moves demands particular attention.

279 The history is informed. The regulatory and policy decision making is informed by a long history specific to the common carrier industries. And, for instance, we discussed the case in 1910, a Board of Railway Commissioners case between the Western Associated Press in Winnipeg and the CPR Telegraph Company, which was a railway company that also owned a telegraph.

280 The railway company made an exclusive deal to distribute news from the Associated Press in the United States throughout Canada. And the Western Associated Press was a competing news service, bringing news to a variety of newspapers in Winnipeg and surrounding areas. Through the use of discriminatory rates, CPR sought to put WAP out of business.

281 And the Board of Railway Commissioners outlawed this saying that if the CPR is allowed to give its own affiliated news services news for free or for a reduced rate while competing newspapers are required to pay a per word fee or a metered rate, then all comers would be put out of business. That the CPR’s newsletter would be the only one available in Winnipeg. And while on its face side it looks like a competition issue, you know -- the Western Associated Press, it wasn’t, from an economic perspective, in competition with the CPR because it was a not-for-profit. It did not return money to its shareholders. It existed solely to bring the news to the people of the West.

282 And so, I think that that issue is something that would be missed by a pure competition analysis. You know, that what we’re talking about here isn’t just widgets being produced in a factory. It isn’t people just expressing their economic preferences. It’s expressing things that are more fundamental.

283 And, I mean, this is the reason that the second section of the Charter says people have the right to freedom of expression, including the use of the media. This is the reason why the freedom to receive and to impart information is in the Universal Declaration of Human Rights. Those aren’t considerations that a competition authority considers, but a communications regulator is required to. Fenwick.

284 And I mean, did you -- I mean, I could comment further on the issue of ex post versus ex ante but I get the sense that’s ---

285 THE CHAIRPERSON: Well, I’m going to get in a moment, so you’ll have that opportunity. At paragraph 72 of your written submission, or right after it actually, you’ve included a Table 1, and -- that looks at, you know, various jurisdictions, approach to zero-rating and differential pricing around the world. And, of course, on the one hand there are countries that, in your view, have banned it. Then there was a middle group that have certain restriction but not all, you know, under conditions, I guess. And there’s another column for -- where there are no rules.

286 And I was wondering, you know, in your submission you described the high level — and you referred to it earlier — the high level of vertical and diagonal ownership in Canada. In your view, is that nature of that ownership model that Canada’s developed over a period of time sufficient rationale to make a more aggressive, maybe the wrong word, aggressive, extreme or more on the left side, I guess, of that table, a prohibitory perspective as opposed to a permissive perspective in light of the ownership nature in Canada? Would that be something we should draw from your analysis?

287 MR. KLASS: I think that there are warning signs. When a company operates -- I mean, I sort of discussed this briefly earlier. But the data that we examined from a research group based in Europe that follows these things very closely and at a very granular level of detail, called Rewheel, investigates the consequences of various different ownership structures on pricing and availability of services in mobile wireless networks.

288 And one of their main headline findings — and I would get tripping over all the various different details if I tried to rattle the numbers off here but — was that carriers who are stand-alone mobile network operators are offering much higher levels of data to their customers than carries who are connected to wireline networks, so the diagonal integration. And carries at zero-rate services are offering much smaller data caps to the open internet than carries who don’t zero rate.

289 So I think that -- in response to your question, here we have, in Canada -- because of the high levels of vertical integration, because all of the mobile wireless network operators are also operating wireline networks, now that Wind has been purchased by Shaw and that Mobilicity’s gone out of business, and that the fact that all of these companies are also operating broadcasting distribution undertakings, in my -- or most of them, if not all -- it’s simply a different incentive structure than it would it be if they were separate companies.

290 So, the fact that you see stand-alone mobile network operators offering significantly more generous access to the internet, it’s not that the company’s cost structure or the cost structure of that operation is different in that they’ve somehow found a way of doing the same thing for cheaper, it’s that their business incentive is different.

291 And so here, when we have cable companies, as I said earlier, who are worried about losing subscribers and have frankly what I consider to be limited competition in wireless markets and wireline market although I -- you know, I mean, I understand the Commission’s approach is to bring about more competition through the wholesale regime, but I don’t think we’re there yet. You know, in the introduction to our presentation today, I highlighted I think that strides are being made, certainly. But you have to take those into account. You can’t just look at a wireless company when it also owns a wireline network and has introduced LTE networks with extremely high capacity but maintains the same data caps that it had before it deployed LTE.

292 If I was in that position, I’d be worried that someone might use unlimited internet on a mobile phone to say stop subscribing to their home internet. And that’s not a worry that a company who only operated a mobile network would have because they don’t care if the people who operate the wireline networks are doing business or losing customers. So, I think Fenwick has something to add.

293 THE CHAIRPERSON: Perhaps I’ll ask the question slightly differently just to add to focus to it.

294 We’ve had companies in this country seek and obtain ownership structures that, I think in your view, are highly vertically and diagonally integrated. And, as a result of that, the price of having sought and obtained those permissions to create that ownership structure in Canada might be a more prohibitory perspective when it comes to differential pricing.

295 That’s the consequence of their own request to have the ownership structure they requested.

296 MR. KLASS: Yeah, I think if you can more easily identify the risks of bad behaviour related to market incentives that wouldn’t exist in a more competitive or less integrated market, that makes a better case for imposing a rule ahead of time if you can see what the problems are before they hit.

297 THE CHAIRPERSON: So -- sorry, did you want to ---

298 DR. McKELVEY: No, I -- yeah, I just wanted to just present the vidence that, you know, at the beginning has been said three years ago, the impetus of a lot of this was that there was -- the zero-rated services were also ones that were broadcasting services. So, you know, the evidence we have in Canada is that the first, you know, the first crack, the first opportunity to use these zero-rating services, that wasn’t necessarily to offer more affordable programs. And certainly I’m not saying that it couldn’t be done, but given this ability, the first thing that’s done is to actually zero-rate their broadcasting traffic. And that was something that, you know, was ruled against, but it certainly raises questions about why that was the first innovation.

299 THE CHAIRPERSON: Right. Well, in light of that, I’d ask you then at the other extreme, should a future framework be more permissive for smaller ISPs or new entrants because that risk is not as prevalent?

300 MR. KLASS: There’s research out of LSC by a person called Alissa Cooper who wrote a PHD dissertation comparing the United States and Britain’s regulatory approaches to discrimination. And I mean, the findings were -- I mean, I don’t think they were surprising. They might be to someone unfamiliar, but there’s more discrimination in proportion to a larger amount of firms and competition in the British market than there is in the less permissive United States where the markets are more concentrated.

301 I think that if -- in a perfect world, if you had a perfect market, which doesn’t exist anywhere, but the more competition you have and the more choices people have, the more -- certainly the more leeway there would be to allow these types of practices. But we don’t have that.

302 And, you know, I listen to David Colville, a former, long-time CRTC executive at a conference at University of Ottawa last year say that when we were discussing competition in the ‘70s and ‘80s we had high hopes, but here I am standing before you today and I can tell you it hasn’t lived up to what we expected.

303 THE CHAIRPERSON: Right.

304 MR. KLASS: So ---

305 THE CHAIRPERSON: And, of course, he’s associated with Eastlink.

306 But again, my question was, should we be allowing a more permissive and less prohibitory approach? I mean, we have to make decisions. Theories, asking questions, that’s all great, that’s wonderful realm, but you, as experts, when you’re looking at this, do you think it would be wise for the Commission as the outcome of this proceeding to allow for a more restrictive prohibitory approach or framework on differential pricing for large, dominant, important players that have asked to consolidate their ownership structure over time and take a very -- a more permissive approach to smaller ISPs and entrants?

307 MR. KLASS: I think that the central principle the Commission needs to protect is common carriage. And that involves not unduly treating different types of traffic differently. And so I think that the approach that the Commission should take is prohibitory. That -- in particular, content and application-specific differential pricing practices should be prohibited ahead of time. And in the case of vertically integrated companies, there should be a particular -- it should merit particular concern.

308 THE CHAIRPERSON: M’hm.

309 MR. KLASS: I don’t think that treating different companies differently from the perspective of a framework that the Commission establishes is a good idea if it serves to undermine the general principal that discrimination should be -- should not be the norm and that people should be able to choose their own types of information and where they go.

310 THE CHAIRPERSON: Okay. I understand. And it’s a principled approach, a bit like the one you took earlier on Canadian-made or Canadian content that’s -- there’s a concern, therefore it should be not watered down generally.

311 DR. McKELVEY: And I just want to say that it is a challenge and I think that there is that -- the rationale for the principle approach is that it’s a legible policy response in that it also protects these -- the traditions of common carriage.

312 And certainly, to your earlier point about what happens if you just switch from being a telecommunications entity to a broadcasting entity, I mean, that raises questions about whether you’re talking about Internet service anymore. Like certainly they’re welcome to do it, but could they then advertise that they offer Internet service, you know?

313 And I think that one of the things that is here is that there’s a sense that the Internet is a common carrier and that’s something that we’re willing to -- that we’re here in principle to protect. And I think that certainly there are lots of directions, but I think moving into, at least from our point of view, there’s a need to reiterate and reinforce that idea, you know, at least as our part of this intervention to be a strong voice about the necessity, importance of common carriage.

314 THE CHAIRPERSON: I have a question about Section 36. And I’ll just do a little preamble because my interpretation of your arguments here, it seems to be -- and you base yourself on paragraph 47, for instance, of your written submission studies done by, I’m not sure if it’s a man or a woman, B.A. Cherry, that outlines the essence of the common carriage obligations. And the history starts in the common law. And I take your position that that common law has largely been codified through the Telecommunications Act, and if I’m not mistaken, you’re saying that it’s been codified in Section 27 and 36.

315 And you seem in this proceeding to be invoking Section 36 as part of our analysis. And I’m having a little difficulty understanding your argument in this respect, and maybe you can help me out on this. Because Section 36 says, as you know, except as otherwise provided by the Commission:

316 “...a Canadian carrier shall not control the content or influence the meaning or purpose of [the] telecommunications carried by it for the public.”

317 In this particular instance, assuming a ISP reduces the rate or somehow has an arrangement or a zero-rated completely particular application, like a streaming music service or any other, they stream the service in their -- in its entirety. So I’m having difficulty understanding your argument that somehow that they are influencing the content or influence of that message in that particular streaming service. I’m having difficulty understanding. Maybe you could explain to me why Section 36 is somehow relevant here. I’m not saying it’s not -- it’s irrelevant to neutrality more broadly, but in this particular instance, why is it relevant?

318 MR. KLASS: Well, as I said in my oral remarks, a price signal influences people’s economic behaviour. And in the past, for instance, during -- in the ITMP framework, the 2009 framework, the Commission indicated that it would treat instances of blocking through an analysis based on Section 36. As I understand it, a service provider who blocked access to telecommunications would be running afoul of Section 36.

319 Now, a carrier who offers -- who creates a limit, an economic limit on the Internet access that you can use but freely allows people to access choice -- sites that they’ve chosen can be effectively blocking people’s access to the broader Internet. If you run into your data cap, for instance, you can’t use the Internet without purchasing more Internet access. Or in cases where the throttling is imposed, I know that some ISPs throttle once the data cap is reached, the zero-rated traffic would be -- you know, the person would have access to it, but they wouldn’t be able to keep using their regular Internet without either the throttling or running into an economic block.

320 THE CHAIRPERSON: Right. I’m not dismissing that there’s an economic issue at stake and that Section 27 might not be relevant. I’m asking about Section 36 because I’m not seeing an interference with the content.

321 MR. KLASS: Well, I mean ---

322 THE CHAIRPERSON: The content is being streamed, for instance in that example, unchanged.

323 It may be -- there may be other economic incentives to use or not use that stream service or that application, whatever it might be, but nobody is interfering -- at least I can see -- with the actual content of what is being streamed.

324 Perhaps not treated economically the same way but the actual content, I'm not seeing it interfered with and so I'm trying -- so I'm struggling with your suggestion that section 36 is pertinent.

325 MR. KLASS: Well, I mean I'm not sure ---

326 THE CHAIRPERSON: Or am I reading section 36 too narrowly?

327 MR. KLASS: I certainly would say -- certainly wouldn't say that but I think we might have sort of a different interpretation of it.

328 In 1999, the carriers had to -- or a number of the carriers in any case applied for permission from the Commission under section 36 to become involved in the internet services, the content of the internet services that they provide. And this is in the context of running company email -- or websites and email service.

329 A company that runs an email server isn't changing the messages in the emails but they had to apply for service because they became involved in influencing what type of telecommunications people use. I mean the website, clearly a website is a piece of content but the email is another thing.

330 And so in 2005 I believe it was -- well, I won't use that example but I mean I just think in the case of blocking content, that's not -- it's not as if a carrier has gone and changed the text or opened a letter and written something different on there but it's that they've blocked access to it to a certain portion of the content. And so I mean I think it involves both of those things in my view.

331 THE CHAIRPERSON: Right. So you're saying to the extent that it is -- it amounts to blocking, that's how section 36 gets invoked?

332 MR. KLASS: It's the same -- it's of the same type. And I mean I think also in the research that went into putting this together, my understanding is that section 36 is an evolution of an article that was -- came from Bell's charter imposed in 1968 designed to prevent Bell from entering into the cable TV business.

333 As the incumbent telephone company, it had control over support structures. It actually owned the wires that cable companies installed to carry their signals and so on. And the Commission or the Parliament I think at the time was worried that given its incumbent position, it would dominate the cable industry.

334 THE CHAIRPERSON: Right.

335 MR. KLASS: The cable companies weren't doing anything other than selecting channels and distributing them to people, which is an editorial function in my view. They pick a bundle -- they bundle together channels and they offer those channels together to their customers.

336 I don’t see any difference between -- I mean substitute the word "app" and "channel" and what's the difference between what Vidéotron is doing? They're selecting a bundle of apps that customers can offer and on a preferential basis, they're not blocking access directly to the rest of the internet. But I mean in cases, like I said, where there's economic considerations, it may amount to the same thing. But they're making editorial decisions about what is and isn't included in this offering.

337 So I mean, in my view, that amounts to a form of control. Vidéotron gets to say commercial music streaming services are part of our program and non-commercial talk radio isn't.

338 THE CHAIRPERSON: Okay, thanks. You might want to give further thought to that in the further stages in this proceeding.

339 Perhaps I'm going to turn to the elephant in the room. That's data caps.

340 At paragraph 60 of your submission, you might want to pull that up, should I draw a conclusion that you believe that data caps are not just and reasonable?

341 MR. KLASS: Yes.

342 THE CHAIRPERSON: Okay. That is what you're saying?

343 MR. KLASS: Well, I mean can I put a caveat on that or ---

344 THE CHAIRPERSON: Well, no, it's your answer. You can put whatever caveats you want on that.

345 MR. KLASS: Okay. I mean I don’t have any -- I don’t have special insight into the carrier's costs, specific carrier's costs in the case of the wireline networks. There is a per gigabyte cost or per megabyte cost in the wireless tariffs which is currently under debate but I mean I see the legislated roaming cap, which is based on retail revenue, is -- was initially in the range of 35 or $37 per gigabyte. I believe it's come down a bit since then. But I see overage charges on retail internet service plans in the range of 50 to $60, twice or more again what -- or nearly twice, I'm sorry, and on its face, that seems like a pretty amazing margin.

346 There may be cases in which the data caps are justified and particularly on wireline networks. I did a survey of the websites casually in a lead-up to this hearing and in some cases, service providers are offering caps in the three, four, 500 gigabyte range and one terabyte range.

347 So I thinks it's hard to make a blanket statement across the border and that's why I haven't come here and said get rid of the data caps because there's a lot of ins and outs.

348 Some of them may be justified and I believe that there is -- there are certainly reasonable justifications for managing traffic but when I look at the lowest-tiered packages and I see a cap of say for instance 25 gigabytes on a 5 megabyte per second connection, when they're offering someone a 30 megabyte per second connection with an unlimited cap, I mean I just -- I can't see how that is possibly justified in the name of managing network congestion when that person who's got the 25 megabyte or the 25 megabyte per second connection isn't paying proportionally more for their, you know, unlimited cap versus the 25 gigabyte one, the person with the lower cap can't possibly congest the network.

349 So I mean I think that there's cases in which it's very clear that the cap isn't related to its stated purpose and then there's cases where it may be reasonable and people will never reach the limit.

350 THE CHAIRPERSON: Right. In fact, at paragraph 62, you suggest that the caps have not risen substantially despite substantial investments and deployment of networks.

351 Did you do any longitudinal analysis of those data caps that you might be able to provide us proving that point, or is that all there is in your assessment?

352 MR. KLASS: I think I'll turn it to Marc and then come back.

353 MR. NANNI: Well, during the initial filing, Vidéotron dropped its cap. I don’t recall the number, from 6 to 5 or from 5 to 4 gigs, something like this, on their mobile. So it's obvious price is going up, cap is dropping. Simple as that.

354 THE CHAIRPERSON: Right. It's just that you're making a blanket statement at the first sentence of paragraph 62 about basically all the carriers and I was wondering if you'd actually done an analysis of that or not?

355 MR. KLASS: I'd be happy to provide that as an undertaking. I have a considerable amount of archived marketing material from the companies. And so if you'd like me to produce something ---

356 THE CHAIRPERSON: No, not necessarily. I was just wondering if there was anything backing the statement you've already made?

357 MR. KLASS: I'll have to get back to you on that one.

358 THE CHAIRPERSON: So there is no study currently; is that correct?

359 MR. KLASS: I'm sorry. You're referring to paragraph 61 here?

360 THE CHAIRPERSON: Sixty-two (62) where you're saying:

361 "Carriers across the country have made significant investments in networks."

362 And then you draw the conclusion that notwithstanding those investments, the data caps have largely not risen substantially.

363 I'm taking your argument to be and I was wondering if you had actually done research on that linking the actual across various carriers of their data caps? If not, that's fine. We'll just move on to another subject.

364 MR. KLASS: I haven't produced this study that is exhaustive.

365 THE CHAIRPERSON: Okay. Thank you.

366 So let's just assume for a moment that data caps are prohibited or curtailed in some way, perhaps not as a result of this proceeding but in a follow-up proceeding, what in your view would be the impact on the implications -- the impact and the implications for network providers, particularly their incentive to invest in those networks, the potential congestion that this might cause?

367 You may have seen the recent publication of the CMR, the demand of Canadians for data seems to be using our word skyrocketed. Would you have a concern that if there was a curtailment or a prohibition on data caps, that that would cause network congestion?

368 DR. McKELVEY: I think that the one key takeaway is that there's genuine concerns from our part about the effectiveness of data caps as a tried management practice, and that there is a lot of competing rationales for why data caps are in place and we don’t have privy to that reason. So, it is something that we’re here to kind of raise concern about, you know.

369 In terms of responding to network congestion, while I’m not a network engineer, there's a variety of different strategies that have been implemented. Certainly we do have -- I mean we do have technical internet traffic management practices in place, and those are legitimate concern.

370 And also, if you look at organizations or carriers like Comcast in the United States, and they’ve move towards more user-specific congestion management. So in the case of high usage users in Comcast in -- at least my last review of their submissions to the FTC related to the 2008 throttling of peer traffic -- they moved away from throttling-specific applications to -- and this is just -- they moved away from application-specific to user-specific management.

371 And so all to say is I think there is a variety of different ways of managing congestion, and I think there's a number of legitimate options. And I think that the overarching, you know, concerns with data caps remain to be proven as the most effective. And I think there's a variety of different abilities on the part of carriers to deal with these congestion issues. So I don’t think that data caps or, you know, a ban or changes to data caps would impede them significantly from also being able to manage congestion.

372 THE CHAIRPERSON: As you look at networks, would you agree with me that congestion in light of how people are using -- residential customers are using their internet connection, that the congestion is mostly occurring in the evening on -- period of time during the day part and yet -- would you agree with that, that that’s ---

373 DR. McKELVEY: I’d love to know for sure. I mean, for example, my work is within the M-Lab system, you know, that’s my expertise with internet measurement, it has difficulty measuring congestion, it can be used. I haven’t done a conclusive study about congestion at time of day. Certainly, we need to, you know, I don’t know off the top of my head the Sam Knows -- the recent report on that of whether it measures congestion of time of day. I mean I’d like to know, I’m not entirely sure and I'm not disagreeing, I’m just saying I don’t have any clear evidence.

374 THE CHAIRPERSON: But would you agree that it’s more a time of day issue than -- I’m having difficulty because it’s counterintuitive that the studies I've seen are that the peak usage are in the evening, and that you have to almost build to that, and that’s the congestion that you want to manage through economic or other means, yet the data caps are monthly?

375 DR. McKELVEY: It’s a very fair point. I would say that the economic -- the technical internet traffic management practices when they were being used before they were phased out by Bell and Rogers in 2012, both were, you know, often were constant and then at other times during at least 2009 were time of day. So there seems to be in terms of effectiveness to your point, if you try to manage congestion it seems that technical internet traffic management practices are more granular and specific in their managing of these peak points of congestion.

376 And I -- by no means do I -- I'm not contending with your claim that congestion is a time of day issue, I’m just saying that I don’t have clear -- I don’t have myself evidence to support that, but I'm not – so I’ll take your word for it. And if that were the case, then you could say that data caps do then -- there might be more effective ways of managing that congestion that are more time-specific.

377 THE CHAIRPERSON: Wouldn’t you be concerned though if data caps were somehow constrained -- from a public policy perspective going forward -- that this would -- that this might lead to higher rates for all subscribers?

378 MR. KLASS: Again, the extremely complex sort of nature of these situations was what constrained us from coming and saying, “We want the data caps to be banned”, because the first thought that comes to mind is with the wireless code, you want to -- they got rid of the three-year contracts for the wireless code, went to two-year contracts and then the carriers raised their prices and started blaming the CRTC, right.

379 You can't -- I mean I can't predict every response that the companies are going to do. And here they are saying, “We are going to have to raise the prices if you get rid of data caps, because we won't be able to differentiate our products.” And I -- all I can say to that is that there are other ways that the carriers can differentiate their various different service offerings than data caps. They're already doing it, particularly in the wireline networks with respect to capacity.

380 So as I was saying previously, it’s hard to imagine that a person running a five megabit per second connection is going to contribute to network congestion in the same manner as a person who’s running say 100, both of which are on offer from many of the ISPs that are going to be showing up here. And number two, is that, you know, you see that there are plenty of carriers who are offering unlimited internet access and their networks aren’t catching on fire or bogging up.

381 MTS in Manitoba on its wireline networks is offering unlimited -- unlimited data as a standard feature on its internet plans. SaskTel does the same in Saskatchewan. And even on the wireless networks, both of those companies are doing that.

382 So to me, when I hear claims that we’re not giving, you know, “Our networks will blow up if we start offering unlimited data, we won't be able to differentiate our products if we start offering unlimited data”, that sounds like fear mongering to me. And it sounds like something that the carriers don’t want to do. But at the same time, I can't say what they will do if the Commission does that, if they won't find some way of raising the prices. So that's our position.

383 THE CHAIRPERSON: Right. You don’t think Canadians appreciate having more control over the rates they pay, based on the choices they make in the marketplace between this offering or that offering for internet service with or without data caps? And in fact, they're in control, they can pay what they want or not want, live with caps or not caps.

384 MR. KLASS: That may be the case, but I think the big problem of what you say is that there aren’t these offers available in the marketplace, particularly in the wireless market. They have started to come in in wireline, but really what our problem is and as I discussed in the intervention -- and I won’t go on too long with this -- is that if they're saying that we put these caps in to manage the congestion, if the prices per megabyte is so much higher than what it costs to provide the company, I don’t see a justification.

385 So they may find a way to raise the price regardless if you take away the cap, but they say -- the company is saying on the one hand that, “Okay, we’re going to offer a low -- a low affordable internet plan for granny who doesn’t want to download Netflix and so on.” But they put a cap on it that’s so low that she can't possibly update her computer or will face an extra charge. And you know, I think it’s the people who need these -- the affordable services that mostly don’t want to see those extra fees at the end of the month, right.

386 So, I mean, I think we’re sort of getting into the same type of discussions that took place at basic service, but I suppose that one option would be to require that their basic service level is offered on an unlimited basis below a certain bandwidth. Or perhaps a situation in which the Commission puts a limit as it did in the wireless code on the amount of over fees that someone could incur in a month, or require consent before those fees are incurred.

387 You know, I don’t think that necessarily getting rid of the data caps across the board -- like that’s a blanket solution, and rarely are you going to find a one-size fits all type of solution in these situations. But the point I'm trying to make generally is that the carriers are saying that the sky is going to fall if you do anything about the data caps. I don’t believe that. And I definitely think that there is a problem with the situation we have right now that can be solved.

388 THE CHAIRPERSON: Have you given some thought on how the Commission would do that in light of the fact that the CRTC is generally forborne from retail internet pricing?

389 MR. KLASS: I mean I’m not a lawyer, so I’m just offering a lay -- I mean I think that’s obvious by now -- but I'm just offering a layperson’s perspective on this. But as I understand, the Commission still retains its power under Section 24 of the Act to impose conditions on services offered by companies. And if they did it in the wireless code, then I don’t see any -- I mean, again, I’m not a lawyer, but if it happened in the wireless code, I don’t see why it can’t happen here. If ---

390 THE CHAIRPERSON: Yes?

391 MR. KLASS: The Commission could also -- I mean, I don’t -- again, I’m not a lawyer, but the Commission could find that the market isn’t sufficiently competitive. If it sees evidence that the prices are so far out of wack with what it costs the companies to provide, it’s kind of like a smoking gun. So ---

392 THE CHAIRPERSON: Okay.

393 MR. KLASS: But I don’t know what the implications of that are from a legal perspective.

394 THE CHAIRPERSON: Do you think that curtailing in one way or another the DPPs would have a positive or a negative impact on the current data caps offered in the marketplace?

395 I’ve heard it -- well, we certainly discussed it when we introduced the wireless code, that there seemed to be competition in the wireless -- of wireless providers not on the cost of basic telecommunications service. They were competing on devices.

396 One could argue, I guess, that in internet, they’re not actually competing on the internet price. They’re competing on offerings that they’re adding that are not telecommunications service, a particular app being -- or not counting to a data cap.

397 So hypothetically, if we were to curtail differential pricing practices in one way or another, do you think it would likely cause the telcos to compete more on the basic price and, therefore, raise the general data caps that we are seeing in the market?

398 MR. KLASS: The evidence that we’ve seen from the reports we cite in our written intervention suggests that carriers who don’t zero rate services offer larger data caps. And that some, in the case of several regulatory decisions in Europe, raise the data cap following a prohibition on the zero rating.

399 So I mean that’s what we’ve presented here.

400 THE CHAIRPERSON: Okay.

401 MR. KLASS: I also want to point out that the -- I think that it adds an unnecessary layer of complexity and possibly complicates a consumer’s ability to choose between services by adding these extra things.

402 So when you say carriers are choosing to compete on the basis of value-added services, zero-rated services, I don’t know exactly what that means by compete, by being more competitive. Does it mean they’re trying to attract subscribers? Yes, but if -- I mean, I think, you imagine a situation in which one service provider is offering zero-rated access to your favourite streaming app, that you subscribe to and that you watch at home and one isn’t. That’s going to be a factor in whether the other carrier is a viable option for you.

403 And so I think that you can imagine situations in which having these zero-rated offers are -- would make it harder. Like in a competitive market, my understanding of the basic economics is that consumers have to be able to substitute. They have to be able to switch from one to the next. And if a company has got some exclusive deal that none others are offering or are incapable of offering, it makes it harder.

404 So that’s less competitive, in my view, and it’s something that the Commission, I think, has dealt with elsewhere in the -- for instance, in the digital media exemption or the preferential treatment on the basis of a particular mobile operator.

405 THE CHAIRPERSON: Right.

406 MR. KLASS: And also on the hybrid VOD. But I mean, that’s my position that I don’t really think that that’s actually competitive from an economic perspective and in fact it could actually make it harder for people to choose.

407 THE CHAIRPERSON: In your documents, you make a lot of reference to international comparisons. And of course, the risk of starting to do international comparisons is that you go down a rabbit hole rather quickly to figure out, well, is the legislation similar or the technology similar? Is it a developed or underdeveloped country? Is the geography the same? There’s a lot of puts and takes and then we have like mini trials as to whether, you know, Norway is comparable or not.

408 Keeping that in mind, which is a single market that you’ve looked at which you think is the most relevant to the Canadian situation with respect to DPPs, obviously, and why?

409 MR. KLASS: I think that the most closely comparable country to the United States with respect to this issue is -- or with Canada. Canada. There you go. I ---

410 THE CHAIRPERSON: All right. Okay. So you think the United States is ---

411 MR. KLASS: Yeah, I think it’s the United States, yeah.

412 THE CHAIRPERSON: Okay.

413 At paragraph 91, and this is -- it gets back to you that you wanted some airtime on the notion of ex post and ex ante. At paragraph 92 you describe the approach of the Indian regulator. And certainly many parties in this proceeding have suggested that, you know, that if there’s an issue it is a potential issue and therefore we should look at it on an ex post basis as opposed to an ex ante basis.

414 Do I take it from the elaboration of paragraph 91 where the Indian regulator said, look, an ex post approach is so labour intensive that the litigants to come to file complaints are not on an even playing field and that there’s no administrative simplicity in that. And as a result of that, the lesson from paragraph 91, I guess, would be that Canada should take an ex ante approach to DPPs. Is that what I should read from paragraph 91?

415 MR. KLASS: Yes. Yeah, and I think there’s -- that there’s plenty of examples, as Fenwick mentioned earlier. Well, there’s the World of Warcraft throttling issue that Rogers, you know, basically repeatedly and for a long time, you know, was found to be incorrectly using its ITMPs. There’s my own personal experience of Mobile TV. And, Fenwick, I believe you had something to add?

416 DR. McKELVEY: No, I just wanted to say in thinking about this and I’ve just been reviewing the ITMP complaint put forward by the Canadian Gamers Organization and Teresa Murphy about the World of Warcraft throttling. And I think it’s a really -- you know, as an academic, I think it’s a really exceptional story because it, you know, is about few average Canadians working for over two years to bring about a concern they had with how the network was working.

417 And it really raises legitimate questions about the, you know, the after the fact enforcement about just how difficult it is on the part of publics to be able to participate and actually gather the evidence. And the fact that, you know, from the, you know, early 2010 when they found there was a concern about this until the, you know, the eventual decision by Rogers to stop doing this in 2012, it’s a long period of time that does change habits and is, you know, it’s a long time on the internet.

418 THE CHAIRPERSON: M’hm.

419 DR. McKELVEY: So all to say is that, you know, the precedent in Canada has raised concerns about that, just not necessarily that it wasn’t effective, that it just was a very difficult thing and really relied so much on public labours to keep it going.

420 THE CHAIRPERSON: As I mentioned in the opening remark, many parties to this proceeding have readily admitted that DPPs are discriminatory or a preference. And the issue then becomes whether it’s undue, unreasonable.

421 And our approach the Commission has taken in many cases when it comes to these sorts of issue a reverse of onus provision. And since that there’s a widely -- it would appear, maybe others will disagree -- some consensus that it is -- there is some preference or discrimination and the real issue is a factual one is whether it was undue, unreasonable. And in light of the general way the Commission reverses the onus on these sorts of things, and based on your comment that an ex ante rather than ex post is a better approach, I’d like your comment on the potential that the Commission would generally prohibit DPPs unless those that want to use them can factually make the case that they should be allowed. Does that meet your ex ante concerns?

422 MR. KLASS: I think in a similar way to our position on usage of base billing, I don’t -- I want to step away from putting down a complete blanket, you know, all generalizations are false, including this one; right? There are cases that you can imagine where providing special treatment, as we mentioned in the oral comments, providing special treatment is perfectly justified.

423 For instance, in the case of a VRS I believe as Immediate Access Coalition is arguing, I think there’s an easily demonstrable case to be made that people with special needs, hard of hearing, you know, should have access to these types of things, particularly, I mean, given the fact of data caps, that access to a communications technology that’s readily available but takes up a lot of data, for instance, to use sign language, shouldn’t be withheld because of these facts, right.

424 Like, I don’t like the idea of the data caps. From what I can tell, it’s not reflective of a fair price. But accepting the fact that they’re there, that doesn’t mean that people should be punished because there’s a rule that says you can’t give any special treatment to anyone ever.

425 I think in ---

426 THE CHAIRPERSON: Well, that’s not what I was proposing. I was proposing as a hypothetical that generally they’re prohibited unless you can make a case

427 ---

428 MR. KLASS: Right. Yeah, I know, and ---

429 THE CHAIRPERSON: --- and then would you reverse the onus.

430 So in the VRS situation, a case might well be made.

431 MR. KLASS: Yeah, I’m trying to agree. Yes, I think that that’s -- you know, there would have to be exceptions to the rule.

432 THE CHAIRPERSON: And I take it there might be other examples where it might be more likely one could conclude it’s not unreasonable or undue -- for instance, we’ve heard on the record about apps for account management, data monitoring, bill payments, operating system updates, how do you treat those?

433 Do you allow them generally unless an ex post prevents them, or do you start off with the principle that they’re all the same, it goes against the principle of a simple conduit, and it’s up to the carrier who might want to provide one of these to make the case?

434 MR. KLASS: I think in that particular scenario the analytical framework that PIAC has proposed in its second comments provides a rigorous test that could be applied in an ex ante situation.

435 I mean, I’d worry if the Commission was to turn it around and do it in an ex post fashion that it would be burdensome in the same way as the standard analysis would be.

436 But in the context of an ex ante analysis, I think that they have really provided a solid framework that could apply to people seeking exceptions from the rule.

437 THE CHAIRPERSON: Right.

438 I wouldn’t mind getting your views on privacy concerns when it comes to operating system updates.

439 I was struck hearing comments at one point in Nunavut where, as you know, because of the nature of the satellite delivery networks, it can be quite expensive to update an operating system, yet sometimes it’s essential because there’s been found to be a privacy issue and there are patches being provided, yet there are people up there that will wait months to come down south to update rather than do it immediately.

440 Do you have a view on that and whether -- how the Commission should consider those sorts of issues?

441 Because I do think it’s a -- I’m applying a privacy lens.

442 DR. McKELVEY: You know, I want to say those concerns are real and something that I take quite seriously so I appreciate them being raised here.

443 Certainly one of the largest concerns with privacy is the -- that it can often be seen as a luxury and something that people have -- you know, because of economic reasons have a better access to privacy. And certainly that factors into operating system updates and a general concern about security about maintaining devices.

444 Again, to echo earlier comments, you know, these are really significant concerns taking place in the north and ones that, you know, my colleagues at the First Mile Consortium have addressed and discussed. And so one of the things that I would raise is a concern that, you know, is that the most effective way of ensuring that there’s security updates and -- well, protecting their privacy rights, which are quite real and legitimate. So by no means am I diminishing them I’m just saying the most effective metric.

445 And just to say the inverse of that is the case of Jacob Ajit, who is a 17-year old student at Thomas Jefferson High School for Science and Technology in Fairfax, Virginia, and he was experimenting with his T-Mobile phone and found that if he put any applications or any files in his speedtest folder he was able to bypass the data usage and zero rating provisions.

446 And certainly this has also been the case taking place in Africa in some of the Wikipedia zero rating basics.

447 Because it raises a whole other host of security concerns that the inverse of that where now people are really invested in reverse engineering how networks function so as to take advantage of some of the ways that the zero rating services are provisioned.

448 So I think that in some ways is the inverse could be potentially true is this also raises the whole other of security and privacy concerns.

449 THE CHAIRPERSON: Right. And may lead to perverse behaviours?

450 DR. McKELVEY: Yeah, but, I mean, that’s not to say -- and again, it’s not to say -- I think that -- this is also, I think, one of the frustrations with the zero basics in general is that, you know, seeing -- at least having a chance to participate in the basic service offerings and seeing the amount of work and dedication by all parties participating in that and how tremendously difficult the decision is, it seems strange that now that zero rating becomes, you know, propped up is a solution to many of these access to broadcasting concerns because I think it’s a much more complicated -- like the ICT for development in a variety of these things about, you know, inclusive broadband are much more complicated than simply whether it’s a data cap or not. I think the barriers are much more robust.

451 THE CHAIRPERSON: Some parties have suggested that these DPP’s are innovative -- are an example of innovation. Pretty hard question, but what, in your view, is innovation; are these innovative?

452 And do I take it from your comments -- I’ve noticed that you’ve said they likely only benefit well-known brand names and not really upstart or edge innovators. So is there innovation in providing DPP’s, assuming -- well, actually, you might want to define what you think innovation is first.

453 MR. KLASS: I mean, I think a good example of innovation is the Bell Lab’s invention of the transistor back when Bell was conducting research and development. It’s -- we wouldn’t be here having this discussion today if it wasn’t for that basic scientific innovations that have wide effects throughout the economy, society, that affect peoples’ lives.

454 From an economic perspective, innovation can be considered the discovery of methods for providing services or goods more efficiently.

455 And I don’t see how zero rating or differential pricing practices themselves can qualify. The use of discriminatory pricing, as I understand, is nothing new. The idea of applying it to the internet might be somewhat novel but not innovative.

456 Now, I think the question -- I mean, I would respond by saying what does it add to the potential for innovation. Even assuming that smaller carriers could avail themselves of these services, it comes at the expense of equal access to other services that are available on the internet but that haven’t.

457 Having to -- a small company having to spend its time and resources to adapt its model to comply with the standards required by a zero rated service, for instance, such as Googles AMP music service, it’s not costless, right. So instead of spending time doing what they’re trying to do the company is spending time trying to comply with service standards that are required, unnecessarily in my view, by the zero rated provider.

458 And this is not just restricted to the case of one provider, because if numerous different ones are offering zero rated services you lose the benefit of an interoperable network. This is a consideration that was front and center for the Indian regulator in its decision to ban zero rating that one of the benefits of IP, of internet protocol, is that it’s a single standard that allows software or hardware to operate across all different networks.

459 And I think another implication of that generally -- and this is the last thing I’ll say before turning it to Fen -- is that you can -- people can already reach customer bases, they can already test out new products on the internet, and it’s a level playing field.

460 If you introduce differential pricing it only tilts that one way or the other and so people start considering whether to try a new product or use a new product on the basis of whether their internet carrier is marketing it or whether their internet service provider is offering it more cheaply, that takes away from the decision making process that says do I want to try to do this or do I want to test out this new service on the basis of whether I think it’s a good idea.

461 Fen?

462 DR. McKELVEY: I just think the -- I mean, this might be a question that could be further posted by van Schewick on -- I think she’s speaking on Friday, because her book is really front and centre discussing this question of innovation and how Internet governance and these issues have an impact on innovation. And largely one of the things they would talk about, and Ben also mentioned, is this idea of end user innovation, which is idea that people can, without too much marginal cost, change their habits or select new channels or applications.

463 And certainly, you can think of it the host that -- what we’re going through in large part is driven by end user innovation. And certainly, that’s not to downplay the kind of growing global concentration in the market, and I certainly want to be acknowledging that, but I also think that you can say that zero-rating has a tension with end user innovation because people are less incentivized to try potentially new apps because they’re worried it might count towards their data cap.

464 I mean, look at the coverage of Pokémon Go where one of the first frames about the coverage of Pokémon Go was this is going to ruin your data cap. And from a technical stand point of view, like I didn’t see it. I didn’t understand it. But that was a common frame of immediate coverage.

465 And so what happens when we develop an Internet where this new -- apparently -- I think literally there was one study that was saying I think -- I mean, it was like 72,000 years, I’m totally incorrect, but the benefits from people walking around more because of that game.

466 And so certainly, all the think is that, you know, Pokémon Go, you know, is an example of might -- what you may call end user innovation where people decide to play this game. And that radically changed, you know, how we use public parks and policies -- I mean public parks, not policies, but public parks this summer. And, you know, I think that that’s something that needs to be factored in is that, you know, people were worried about their data cap and were worried about their data cap with that game and some of the consequences of that.

467 And so the second thing to say about innovation too is that on this front, you know, I still think that -- and, you know, I’m frustrated with myself as, you know, that I haven’t generated more scholarship about this, so it’s -- I’m partially to blame. But we don’t well know that some of the effects in Internet measurement we talk about quality of service and quality of experience. Quality of service is how it’s provisioned. Quality of experience is what’s -- what happens, how do people feel about that?

468 And, you know, we do have some marketing studies that people will, you know, watch a video after two seconds and then stop doing it. But I really think that there’s some evidence and any of the studies I’ve seen say that, you know, data usage and data cap leads to more anxiety about how people use and contribute to the Internet, I mean, participate online. And all they say is that, you know, could there be innovation in making that transparent? Would innovation actually be that people feel less anxious about using the Internet? And would an innovation be, you know, in somewhat, you know, that data usage isn’t something that so imbues how people understand their everyday use of the Internet.

469 THE CHAIRPERSON: My last area of exploration before I turn to my colleagues deals with transparency. It has been suggested that one could be more permissive with respect to zero-rating or other DPPS if subscribers had access to -- transparently to mechanisms to track the data they’re actually consuming using these applications and therefore seeing transparently the associated costs they’re avoiding. So that’s sort of a transparency that is subscriber facing.

470 And others have suggested that we need maybe transparency that is competitor facing, that is -- that somewhere clearly, perhaps on a website of an ISP, one would have to disclose exactly the nature of the arrangements and the details associated with a provider providing zero-rating or another DPP, so that, I guess, an edge provider could see what’s going on and maybe try to tailor their innovation to that.

471 So one’s -- that’s a competitor facing transparency and the other one is a subscriber facing transparency. What would be your views with respect to either of those or both?

472 MR. KLASS: I don’t -- I mean, I -- I don’t see how with respect to the provider facing transparency, it seems to me to be an added and unnecessary cost to have to have legal or regulatory staff as part of your business searching out and determining whether you’re getting as good a deal to reach customers from an ISP as you are -- as your competitors. I mean, it’s something that certainly I think would favour established businesses over ones who have not proven a business case. And that’s to say nothing about non-commercial applications or users of the Internet who would be crowded out.

473 So in the scale of benefits versus downsides, I mean, I don’t see what problem adding transparency is going to solve from that side. And I mean, in the context of the Commission, I mean, I think that that’s, you know, should you choose to impose some kind of a framework guiding these practices, you would certainly want to have that type of information.

474 And I mean, I know with respect to Vidéotron’s unlimited music service there’s been lots of assurance, you know. And I think there was a very brief submission, a one-page submission that says all we do is have them sign a piece of paper and write their IP address down. I mean, I have a hard time believing that that’s all there is to it. But when assessing whether these things are reasonable or not, that’s information you would want. I don’t know how it would help the subscriber.

475 And in terms of transparency, I mean, I just think of the transparency requirements that are associated with technical ITMPs that were put in place as part of the -- that framework. And I mean, I’m thinking about things like Xplornet’s “transparency” about its network traffic management practices. And I mean, I think it’s -- as I understand it, they’ve got at least four different types of practices that can apply at different times. They can apply at -- according to whether you’re using the network from a particular place in the country, whether you’re using specific applications. I mean, I think you quickly go down this rabbit hole where too much information -- too much information about complex topics just isn’t useful.

476 I mean, in terms of use -- data usage monitoring tools and so on, I mean, that’s definitely beneficial. I think that those are, for the most part, already available on the market from carriers themselves and from third parties, as well as built into the operating systems of devices themselves. I mean, I know the phone that I’ve got in front of me will display data usage by application.

477 So I mean, I just -- I don’t -- I don’t really see that adding transparency requirements to differential pricing practices is solving a problem.

478 THE CHAIRPERSON: Thank you very much.

479 Mr. Menzies? Vice-Chair Menzies? Thank you.

480 COMMISSIONER MENZIES: I just have one sort of follow-up in your discussion about ex post and ex ante. Sort of moved into a couple of different areas where you talked about what might and might not be acceptable as a zero-rating or a DPP. And you mentioned VRS that you thought that might be okay.

481 So what I’m trying to get at is if the principles of net neutrality and common carriage, and even from what I understood your interpretation of Section 36, mean that there should be no preferential treatment, I mean, all content should be treated the same on -- when -- as it’s flowing through the Internet, which I understand and that’s a principled approach.

482 Wouldn’t the regulator itself in choosing what might and might not be a superior moral product, such as VRS to qualify for DPP status, wouldn’t, in your view, the -- or is it your view that the regulator may interfere with content flow and differentiate between different types of content on the Internet and but that carriers may not?

483 MR. KLASS: I think the nice thing about the Commission and these types of processes is that anyone who wants to can come in and have a say in the policies. So just -- I mean, just actually to back paddle just a little bit, I don’t think that I’ve ever actually said anywhere in any of my interventions said all content has to be treated equally without exception. So it’s definitely not our position that common carriage or network neutrality insist on equal treatment without exception.

484 These are social constructs, in my view. They’re rules that are debated in a democratic process and that are evolved through sort of an organic legal and social interpretation along with technology, along with social norms, along with the industrial organization, you know. And along the way decisions get made about how they’re going to apply.

485 I mean, I think you’d have a hard time finding any sort of a hard and fast rule that says all X has to be treated equally. Some people might hold that view, but I’m not one of them. And I think there are certain points at which decisions have to be made about for, you know, different treatment of things.

486 I don’t have nearly as much concerns about differential treatment when it doesn’t involve content-specific or application-specific treatment of traffic, for instance, peak time pricing versus off-peak time pricing. If the carrier’s not making a decision directly selecting a particular application provider or content provider, I think that that’s far less problematic than one in which they’re picking and choosing particular people or companies.

487 So I mean, I think in general my response is that you -- sometimes those decisions can be in people’s benefit and things like VRS might qualify for them.

488 COMMISSIONER MENZIES: Okay. Thank you. The answer’s clear.

489 THE CHAIRPERSON: Just one final question maybe you -- because there’s a legal component to that you might want to take it away and answer it as an undertaking.

490 I’m hearing you say that generally you don’t think that DPPs are in the public interest. Yet Parliament at Section 27(6) seems to actually explicitly contemplate it, discriminatory pricing. Because at 27(6) Parliament stated that:

491 “Notwithstanding [27(1) and 27(2) that] a Canadian carrier may provide telecommunication service at no charge or at a reduced charge

492 (a) to the carrier’s directors, officers, employees or former employees; or

493 (b) with the approval of the Commission, to any charitable organization or disadvantaged person or other person.”

494 There’s two ways of looking at that -- well, there’s a policy question. The policy question is, hasn’t Parliament in a sense said that, yeah, you can do that in certain circumstances, and sometimes with or without the approval of the Commission. And the second question is -- and this is a matter of interpretation and that’s why I think you should take that away and think about it -- that Parliament has actually turned its mind to the situation and has fully occupied the field by saying these are the circumstances in which you can make a distinction between no charge or reduced charges. In other words, when a -- you’re providing it to a circle of people that work for a company or a charitable organization or disadvantaged persons or -- and that, I guess, might go to the VRS situation. I’m not sure.

495 So perhaps my question to you is based on that very long question. Is there a lesson from a policy or legal perspective to be drawn from what Parliament has stated in Section 27(6) for this proceeding as it relates to DPPs? Would you be ---

496 MR. KLASS: Is that -- sorry, is that an undertaking?

497 THE CHAIRPERSON: Well, I’m asking you, are you able to undertake to do that ---

498 MR. KLASS: What’s the date?

499 THE CHAIRPERSON: Fourteenth (14^th) of November.

500 MR. KLASS: Okay.

501 THE CHAIRPERSON: Do you agree?

502 MR. KLASS: Yes.

503 UNDERTAKING

504 THE CHAIRPERSON: Okay. Great. Thank you very much.

505 Those are our questions for you. Thank you very much. It’s been a while. But as you know, as you follow these hearings, questions always take longer at the beginning of the hearing. Then as you hear our questions, the people at the end have the advantage of having crisper answers. But it’s not a criticism it’s just a realization that that’s the nature of hearings. So thank you very much. So we’ll adjourn for a few minutes until 10 -- sorry, 11:35. Thank you.

--- Upon recessing at 11:24 p.m.

--- Upon resuming at 11:37 p.m.

506 THE CHAIRPERSON: Order, please.

507 Madame la sécretaire.

508 MS. VENTURA: Thank you, Mr. Chairman.

509 We will now hear the presentation by Roslyn Layton appearing by Skype from Denmark.

510 Ms. Layton, do you hear and see us?

511 MS. LAYTON: I do indeed. Can you hear and see me?

512 MS. VENTURA: Yes, I can. Thank you.

513 You may begin and you have 15 minutes for your presentation.

PRESENTATION

514 MS. LAYTON: Thank you.

515 Good morning. I’m honoured to have this opportunity to make testimony on differential pricing. Net neutrality and zero rating are the focus of my study. I perform empirical and primary research and make comparisons across countries about Internet policy.

516 My colleagues include engineer Silvia Elaluf Calderwood of Oxford Brooks University in the United Kingdom, and economist Bronwyn Howell of Victoria University of Wellington, New Zealand. We work with original source materials such as traffic data, regulatory statutes, and official government and corporate documents. We also conduct surveys and interviews of users, entrepreneurs, and app developers.

517 Regarding the testimony earlier today by the CMCRP, the record should be corrected for the table they provide on page 33 of their report, which has either outdated or incorrect information. They assert that countries have ruled definitively on zero rating and differential pricing, but this is not the case.

518 Many of these countries have never conducted hearings or investigations at all. Moreover, there is ongoing litigation in a number of the countries with -- including the United States, India, and Netherlands, meaning that the statute -- the status of differential pricing of broadband is in flux. In Slovenia, the court found bans on zero rating to be a violation of their country’s net neutrality law and they were subsequently struck down.

519 My written comments include the relevant statutes for the countries in the chart so that you have the most up to date information.

520 My testimony today will cover five questions which telecom regulators should ask as they want to investigate differential pricing. As you well know, the United States has a case-by-case approach to zero rating. The EU law does not in fact mention the term zero rating, but the non-binding guidelines developed by BEREC, Body of Electronic Regulators for -- European Regulators for Electronic Communication, also suggest a similar approach. In neither case are whole or partial bans supported.

521 The key concern against zero rating or differential pricing is that if some content is offered for free, then it will foreclose other content. There is one instance in which this could be true, in a perfectly competitive market for content. Such a situation requires price taking participants, homogeneous perfectly substitutable products, no search or transaction costs for users, no externalities, and no barriers to entry or exit, and the perfect divisibility of output.

522 But neither are content nor infrastructure access markets perfectly competitive. Both have high fixed, sunk costs which violate price-taking assumptions. At best, the content market is either monopolistically-competitive or oligopolistically-competitive. Therefore, if we don’t have a perfectly competitive market, we cannot assume that zero rating, free data or differential pricing behaves as critics assert. However, we can study what happens in the market when zero rating is introduced, and we can ask critical questions to see whether it’s necessary to adjudicate certain instances.

523 So the first question is, to what degree are applications and content substitutable? A regulator would like to know what close or perfectly substitutable applications over the broadband provider’s connection are likely to be foreclosed by the zero-rated application. We have a real world example from Chile, which I also detail in a paper submitted to this inquiry.

524 In January 2013, 2 net neutrality advocacy organizations complained to the Chilean regulator Subtel that Virgin Mobile, a new MVNO with less than 1 percent market share, offered zero rated WhatsApp. Subtel replied that the offer did not violate Chile’s net neutrality law because no blocking or throttling took place. Discussion ensued between the regulator and the advocacy organizations via email, but the case was closed.

525 More than a year later in March 2014 a new regulator came to office who had earlier been employed as the head of the organization which made the complaint. Without any public hearing or investigation, the new regulator ordered the specific promotions called “free social networks” be stopped. Importantly, this is not a ban on zero rating or differential pricing as people believe.

526 In fact, zero rating continues in Chile today. It is just not marketed or discussed as such. Moreover, after some embarrassment in the media and a request from the Wikimedia Foundation, the Chilean regulator had to clarify that Wikipedia was expressly allowed.

527 This particular regulator has just been removed by the Chilean President who indicated dissatisfaction on his leadership to facilitate increased market share of entrants and MVNOs. That a small and entrant operators are constrained in their ability to compete because of overly strict net neutrality rules is a consequence that regulators need to consider.

528 This incident speaks to the political nature of net neutrality policymaking and is indicative of many situations in which advocacy organizations, not consumers, call for bans on zero rating.

529 However, there is a way to determine empirically whether harm occurs to consumers or competition. One can study the traffic to applications before and after zero rating occurs. My paper details how to do this.

530 Since launch in Chile, WhatsApp was a popular service from its beginning. When zero rating began, WhatsApp actually lost amount of traffic on certain devices, where on others it stayed constant. But its competitors, which included Twitter, Skype, Badoo, Google Hangouts, Emoji, LINE, Telegram, imo, Talking Tom and Viber, amongst others, remained popular and did not experience a change in traffic as a result of the zero rating of WhatsApp during the given period.

531 This can be explained because each of these apps has a different value proposition and appeals to a different market segment. Therefore, the zero rating of one does not decrease the traffic of the other. This is because these apps are not substitutable.

532 In a similar analysis, the Slovenian Competition Regulator proposed an empirical analysis by looking at the amount of zero-rated data versus all paid data and found that the potential impact in the marketplace to differential pricing was but a few cents on a 30-euro monthly subscription. This amount was not significant from a regulatory perspective and did not justify intervention, according to their view.

533 The second question is, does the zero-rated applications cost the broadband provider less than equivalent usage of non-zero-rated applications?

534 Now, this instance may be relevant for rural and remote areas where there is a difference in delivery cost.

535 In the case of New Zealand, about 90 percent of the desired content comes from offshore. This is largely entertainment. It is relatively expensive for ISPs to deliver foreign content versus providing content hosted locally.

536 The ISPs in question hoped that by zero rating the 5 percent of locally hosted New Zealand content would reduce the demand for foreign content. It turned out that the free offer did not make a difference for the end users. It did not change their traffic patterns.

537 This example shows that content is imperfectly competitive and not substitutable, even if it’s offered for free.

538 A related example is Greenland, a country slightly larger than Alaska with 55,000 people. Here, TELEGreenland has data caps on fixed line service to compensate for the use of video entertainment such as Netflix which is expensive to deliver.

539 However, vital applications such as telemedicine and education are designed to be bandwidth-light so they can encourage usage. In such cases, it makes sense to zero rate the socially beneficial applications to encourage their use but to charge for privately beneficial applications as these costs can help recover the cost of the network infrastructure.

540 The third question, is price differentiation designed to spur adoption amongst low-valuers or people with low income?

541 It begs the question as to why differential pricing, if it is as harmful as critics purport, that we allow it in every other aspect of life, particularly to make goods and services more accessible to people of low income.

542 We see this across many industries in which the youth and the old receive discounts, or when offers are made to increase usage at off-peak times, for example early bird specials.

543 Differential pricing is common among digital goods and services. For example, Microsoft offers a range of software products free for students but the enterprise version comes at a premium.

544 The freemium model is common in Skype and Spotify where basic service is free but the premium service requires a fee.

545 So similarly, Facebook uses Free Basics across some 50 countries to stimulate adoption of first time users. This version, Free Basics, is particularly helpful to deliver mhealth or mobile health, as I detail in a separate paper.

546 Some 10 million people have joined this program to access free health services, particularly for AIDS and maternity. These services include free messaging to determine whether AIDS medicines are counterfeit, free reminders for prenatal exams, health information on diseases and disorders, and remote patient care.

547 The research I have conducted with developing country entrepreneurs and developers attest that without Free Basics, they would not have obtained these new users. These platforms that are both free and tailored to the first time user is essential to stimulate adoption.

548 A number of countries, including the U.S. and Canada, have a problem in which a certain percentage of people will not adopt the internet at any price. I believe that the Free Basics model, a free platform with socially beneficial content, could be helpful to give these individuals a reason to try the internet.

549 Therefore, regulators should encourage zero rating or free data, particularly when the goal is to stimulate adoption.

550 The next question is, who makes the zero rating complaint?

551 This is the instance when the regulator needs to pay attention to who makes the complaint, as it may reveal important competition information.

552 If the complaint about differential pricing comes from an existing content or app provider, it may be indicative of an incumbent wanting to foreclose the entry of rivals. However, the complaint could be credible if a potential provider says that such an offer forecloses its entry.

553 We have also seen cases where users want to limit the entry of other users. This was particularly interesting in India where reportedly one million responses were received by the Indian regulator protesting that the poor should not have free internet access. It bears mention that the poor themselves did not participate in the hearing as they had no means to do so.

554 In one way, it is encouraging that people want to be involved in the regulatory process, but the system can also be manipulated.

555 For example, the global net neutrality lobby is well funded by U.S. corporations and foundations to the tune of $25 million annually.

556 A number of internet advocacy organizations use state of the art database and marketing systems to conduct political action on a range of causes around the world.

557 For example, in the recent BEREC consultation on net neutrality guidelines, one-third of the nearly 500,000 responses came from people in the USA.

558 In the net neutrality rulemaking in the U.S., 125,000 responses came from a Canadian advocacy organization, in addition to many thousands of mails from other countries.

559 Thus, when President Obama claimed that four million Americans -- four million people from the USA contributed to the process, this is in fact not true.

560 While regulators may enjoy the buzz of proceedings and the high number of copy-paste comments, it can represent a challenge to accountability. Regulators, by law, are accountable to their national governments and the citizens they serve, not to foreign activists.

561 Moreover, the abuse of digital tools can distort reality.

562 Media scholar Maria Loblich conducted a detailed ethnographic study of 13 internet advocacy organizations in the net neutrality debate in the U.S. and she concluded that their cyber activism tended to favor causes driven by elite, white, and educated users.

563 She also critiqued a media bias in coverage of these groups, the view that they collectively and accurately represent “the consumer” or “the public”. It is suggested that such “clicktivist” efforts marginalize the poor and the unconnected in favour of a tech-savvy male, a suggested dynamic of Reddit for example.

564 So therefore, it's important to pay attention to who makes the complaint about zero rating.

565 And finally, my last question, what are the user’s search and switching costs for content and application?

566 The internet is an experienced good. It cannot be valued until it’s tried. Moreover, platforms such as Google, Netflix, and Amazon take advantage of the attention economy, making efficient use of an individual’s limited time and attention.

567 A user may have set up a profile, preferences and payment details within the platform. These dynamics help make the platform sticky and users, therefore, can be reluctant to switch or try other platforms because of the costs.

568 As such, platforms exert a kind of market power over their users. It is in these situations that it is in fact the new entrants which most desire the differential pricing because it helps reduce the user’s search and switching costs to try a competitor.

569 We have all experienced this ourselves. If you have ever gone into an ice cream parlour and they may have a new flavour of the month or something, you will be unlikely to try it if you have to pay for a full-priced cone or cup, but you may be willing if you have a free sample to try and there you may then go on and purchase the flavour.

570 In this proceeding, some have called for partial bans on differential pricing, for example when a small content provider would pay a network provider for the cost of traffic. But such a ban is most onerous for small and medium content providers because they have the most difficult time to compete in the content market which is highly oligopolistic. It is small providers that most need the pricing flexibility.

571 Prior to working in academe, I had a career in search advertising. To advertise successfully in Google, for example, companies, including small businesses, would generally have to have spent a lot of money, maybe $50,000 a month or more to get economies of scale.

572 However, by working with a broadband provider, they could get an advertising opportunity at a much lower price point and perhaps better targeting. And they can contract with the advertiser to lower the cost to the end user, which provides a valuable financial benefit to the end user which they don’t enjoy today.

573 In any case, content -- in fact, as we experience it today, we, as users, are subsidizing advertising through our mobile phones, whether we like it or not. Ads can comprise up to 80 percent of the content -- of the data on a mobile subscription. By preventing differential pricing, regulators are unwillingly strengthening the oligopoly of the advertising platforms. Broadband providers are perhaps the only players which could create a credible sense of competition at scale for these large Silicon Valley companies, for the most part.

574 In any case, content providers, app providers and broadband providers are already experimenting in ways to subsidize the end user. The app mCent offers a means to deliver data to end users. It has more than 30 million users in 93 countries across 300 mobile carriers. A similar idea is contained in toll-free apps, in which users collect data credits from content providers which can be added to the subscription balance.

575 The complexity of this issue impugns the premise of the scrutiny on differential pricing, which falls asymmetrically on ISPs, itself a form of discrimination. There is no parallel obligation on content and application providers to refrain from activities that would inhibit innovation in the network or the ecosystem or from “picking ISP winners”.

576 Unsurprisingly, inconsistencies have emerged around the world. For example, flat-rate pricing of broadband subscriptions is, in effect, zero rating of all data. The “gamers and grandmothers” scenario in which game users pay lower unit costs are because grandmothers are paying more. Why then is it not acceptable to -- why can we subsidize the gamers but not the grandmothers?

577 Following the same logic, why is differential pricing allowed for different access speeds? Isn’t this a form of discrimination? Why do we say it’s okay or “neutral” for consumers to pay for the privilege of higher speed, but not when content and application providers seek to pay for the same thing? The simple ‘one size fits all’ principle is insufficient to govern a complex, dynamic internet ecosystem with multiple multi-sided markets. There is much to learn here and caution before rulemaking is warranted.

578 THE CHAIRPERSON: Thank you very much, Ms. Layton. I’ll put you in the hands of Vice-Chair of Telcomm, Mr. Menzies.

579 COMMISSIONER MENZIES: All right. Thank you. In your submission, you gave a number of examples of differential pricing practices using telephony, used movies and other analogies to show that it’s illogical for the same practices not to be freely applied to the internet.

580 There is a body of opinion out there though that argues that regulators and others err when they view the internet as a -- merely as an extension of telephony, and that there’s a lot of unique characteristics and nature to the internet which make that a serious error to draw from one to the other.

581 So maybe you can clarify for us why do you think the internet is no different than telephony when it comes to differential pricing areas.

582 MS. LAYTON: Well, thank you for the question. I think you have hit on which are perhaps -- you know, maybe one of the key debates in tele-communications policy we have. Is the internet just an extension of the telephone or is it something different and therefore special and needs special regulation?

583 I think what I would like to point out is an inconsistency in terms of the freedom of pricing. And what I want to show is that when we look at price discrimination within the field of economics, it has over 100 years of literature behind using price discrimination to achieve social goals. You know, we can make an example of, let’s say, the Cancer Society of Canada wanted to pay for the content of a series of cancer videos to help people going through treatment.

584 In fact, that is -- it’s health -- these kind of health applications, which were -- what got sponsor data started in the United States with ATT working with United Health Care. They had a population of low-income woman who were trying to prepare for pregnancy and they wanted a way to encourage the woman to watch the videos without charging their data cap and it made a lot of sense. It made sense for their health insurance to lower the costs, to lower the morbidity, hospital visits and so on.

585 So I guess the question is why is the -- I wouldn’t say that I disagree that the internet is an extension of the telephone network. What I do want to support is that pricing flexibility is necessary for many reasons, for many goals of why we need pricing flexibility. First of all, to make things more affordable for the poor. And, secondly, we need it for innovation. We can look at one of the most important scholars of innovation, David Teece, who has shown through complimentary assets that you need to be able to have marketing partnerships across the board. Zero rating and differential pricing, in many respects, are nothing more than marketing partnerships.

586 So, in many cases, when you’re restricting the ability of any provider to price, it is, in fact, a restriction on enterprise. It’s a restriction on speech. And it isn’t efficient to allow the various parties in the market to come together. So I’m not sure if I answered your question, Mr. Menzies, but I hope that gives a little more insight to what I’m trying to express in this testimony.

587 COMMISSIONER MENZIES: Okay. Well, yeah. The answer I got was that it’s -- the internet is not the same as telephony but it doesn’t matter when it regards -- from your point of view, in regards to differential pricing practices in terms of that.

588 So, just to follow up with that, isn’t there an issue though with -- and to continue with telephony analogy, you know, you can price obviously -- things were priced differently for long distance for a number of calls, minutes, et cetera, et cetera, mobile, all over the years, but isn’t the better analogy for what we’re looking at specifically in this hearing that essentially this is like allowing a free telephone call to your sister, but charging you to call your brother or your other sister, in that the provider is choosing for you what you may listen to for free and taking and disempowering the consumer from the choice, as opposed to offering me so much free data that I may assign to whichever app I choose in a given month.

589 MS. LAYTON: Sure. Well, I’m definitely familiar with that criticism that somehow if the operator is selecting, that it’s the operator’s choice.

590 Now, I am not privy to any particular operator. I can only access the data that I get through the various platform analytic tools that I use, but what I would say is the operator -- the ISP or the broadband provider will only make an offer to the extent that its customers think that they would desire it. Any company who would be stupid to want to offer the sort of thing that they want and not what their customers want, will very quickly be turned out of the market. I mean, we have a graveyard of, you know, of ideas that, you know, maybe at one time operators thought made sense, that didn’t go anywhere. So I have no fear that people choose -- if they don’t want it, they choose not to use it.

591 You know, but what I would say to your point, though, what is emerging are multiple ways to do this. I think India is a very good example. They are actually exploring what’s called a ‘free data architecture’. They realize that this ban is -- really was a -- knee-jerk reaction was really too hard response. They weren’t really prepared to deal with it. But they want to find other ways to give people free data.

592 Now, you’re absolutely right. You could use a rebate or a coupon or even you could imagine a loyalty card in the same way that you fly, you get a mile on your credit card every time you fly on a certain plane, or every time you fly the plane you can get miles or free data, why not?

593 But where I can see the benefit here is that very frequently you can have an instance where an ISP could work with an up-and-coming start-up. This was the case of Spotify. Spotify today is perhaps -- it’s the only European app that’s really global. And if Spotify had not partnered with mobile operators, it would not have ever grown to be where it is. It did have a -- it wasn’t really from zero rating but it was because of the billing relationship. And what the -- what Spotify learned was that when free -- people who joined the platform for free never upgraded. But when they could become premium users when they were buying their mobile subscriptions, they were more likely to pay for the premium subscription of Spotify. So Spotify needed to have those paying users to pay and go its service and that was valuable for them to grow.

594 Now, today what we see a lot of zero rating services do is expand the market. So I know Pokémon GO was referenced earlier. This was zero rated by T-Mobile in the U.S. What that did was actually create -- open the market for other virtual reality games. So people who had never played a virtual reality game could try Pokémon GO for free and so the next time that another game would be in the market they’d be more likely to try it.

595 So I actually would submit to you that it’s not the ISP choice at all. It’s really designed to cater to what the users want and in many respects present apps that people would not find otherwise and because of those search costs that I described they wouldn’t try otherwise. They need to have that incentive to try. And frequently zero rating is not enough so you actually have to do more than that, but in some cases it could be enough.

596 So it seems to me to be too much to regulate it, it’s an overkill, especially if you’ve not established harm to consumers and competition in the first place.

597 COMMISSIONER MENZIES: Okay. But what about its impact on a competitive framework to that extent where the ISP is deciding that we consumers want music so it will decide which music service -- in a commercial agreement with a music service, it will decide which music service I will use or it will prefer for me, right, which is ---

598 MS. LAYTON: Sure.

599 COMMISSIONER MENZIES: --- like your ice cream cone -- your ice cream analogy, to a certain extent, except it’s permanent. It’s not a ---

600 MS. LAYTON: Right.

601 COMMISSIONER MENZIES: You know, for one week -- try this out and for one week there will be no charge against your data cap.

602 The only -- it seems to me, the only way to offer that fairly in a non-discriminatory fashion would be to offer it to all music services or let consumers choose between which music service they wanted that free data to apply to, otherwise the ISP is picking whether Spotify wins and somebody else loses in terms of that and it has an impact on competition in that sense.

603 If you -- so address that first please.

604 MS. LAYTON: Sure. Well, I think that’s a great question. We could look at this in a few ways because we have different examples of this.

605 First of all, if an ISP chooses to use one music service -- personally, I don’t have a problem with it. That’s a decision that they have to live by and it doesn’t -- because they offer it for free doesn’t mean that any person’s going to choose that.

606 I showed you a number of examples where many cases the operator had hoped that making it for free would change traffic and it did not, because this is the -- content is not perfectly competitive nor is music.

607 We can certainly look at the T-Mobile U.S. instance, which I think is very interesting. It has more than 40 music streaming services. And I haven’t done the research, but what I would wager is that many of them, because they are doing all music that way, those -- in fact, so just from a starting point, there are more music services in that T-Mobile bundle as a list than you can find anywhere else on the internet. If you go to Wikipedia or Google you can’t find a list of so many music streaming services in one place.

608 So, in many respects, that particular ISP has found a very interesting product which is promoting music as a whole, allowing people to experiment across a number of music services.

609 T-Mobile between that music and video gained eight million net new customers last year. In addition, it has consistently won the best consumer satisfaction awards and analysis by a number of independent tests. So there is really empirical evidence that people like that particular -- like that view.

610 Now, I will also say in Slovenia, which I think is very significant, you did have a case where there were music -- for example, Deezer, which is a French competitor, was being zero rated by the incumbent, and that was banned by the regulator, subsequently overturned by the courts.

611 So -- but it was in the scheme of things because there were four -- there were a variety of zero rated offers in the marketplace which had gone on for years with no complaints and one complaint was made which precipitated the regulator to ban a set of them and that arbitrary and capricious behaviour ended up getting the regulator in court.

612 So I think it would be -- rather than saying what could or should happen, we could actually look at the traffic; we could ask the music services themselves are you getting more or less users today; are you losing traffic because your competitors are zero rated. Rather than conjecture, we actually have the tools to go and determine this, and I think that’s a step that’s missing in these analysis.

613 COMMISSIONER MENZIES: Fair enough.

614 What about the argument that if this were to expand you can create virtually a parallel internet. Say you took a social media service somewhat like Facebook and it was zero rated and You Tube or other services carrying video were not, wouldn’t that allow the zero rated service to become essentially a carrier of all kinds of different items like streaming video and that sort of stuff, so basically a conduit and a parallel internet service? What would your ---

615 MS. LAYTON: Well ---

616 COMMISSIONER MENZIES: What’s your response to that argument?

617 MS. LAYTON: Well, first of all, I don’t have a religion about http and internet protocol. I would like to think that the internet is not the end of innovation and that other things can evolve.

618 There are a lot of challenges with the internet that we have today. There are a lot of ways that the particular architecture that wants to be enshrined by net neutrality has many shortcomings, particularly in the areas of security and privacy.

619 We have also a kind of design where there’s a CDN overhang. I mean, 80 percent of the traffic -- I think Sandvine is testifying later today. Eighty (80) percent of the traffic is being -- is video, and that was not what the internet was designed to do.

620 So we are actually kind of reverse and trying to -- the internet had not been thinking of those things and we’re trying now to adjust it for that.

621 So if, in fact, what you’re saying is coming true, I wouldn’t have a problem with it because that would be -- it’s actually what people wanted. And maybe this is the idea that the Indians are looking at is that somehow this set in stone idea that the internet was only one way and can’t evolve is out of pace with what people really want.

622 But what I would say to you, I think we can have multiple models without it being hurtful or discriminatory. I think that the multiplicity of models reflects the heterogeneous users, it reflects different budgets, and it reflects different desires. So it’s important that there is diversity in the marketplace and not a, you know, only one size fits all kind of way to pay to be on the internet.

623 COMMISSIONER MENZIES: Okay. And I just want to touch a little bit on the argument -- your argument in terms of affordability.

624 So I can see, for instance, if service providers only offered differential pricing practices to subscribers of high usage plans they -- I can see that they would be encouraging usage and subscription to the higher tiered plans. But, I mean, the last year or so we’ve heard quite a bit about affordability issues. And doesn’t the practice such as that just allow the rich to get richer, so to speak, and while people with fewer resources just remain with fewer resources?

625 MS. LAYTON: I wouldn’t see that in the research that I do. I mean, frankly, I’m not sure if it was what the Chairman was alluding to, but I definitely see an argument and I see a number of net neutrality proponents would support.

626 There’s a lot of arguments for free data platforms for e-government, for, let’s say, Canadian content that -- for example, health information, especially in light of your basic services -- and I think I’m on record for this -- your basic services hearing, that we should be identifying what are the socially beneficial services of the internet and finding a way to make them as affordable as possible.

627 If it was -- you know, whether -- if companies wanted to do it themselves they -- however you wanted to organize it. I actually think that’s a way to get the poor people more content and more data.

628 If you look at our other media industries part of the way of making radio, television and print more affordable to the masses was through advertising. We relied, to a large extent, on the private sector to pay for the cost of delivery so that they did not.

629 Now, I know in Canada you’ve had a certain set of regulations. It’s not exactly like the U.S. But you have had a blend of those things where public sector and private sector participate to help make it available for the poorest people.

630 In my research where I see small operators making partnerships with local app developers, they are particularly looking at programs where they can get customers and they are very much so helping the poorest people. There is no doubt about it and this is going on with free basics, because this is a set of users who are truly poor, that without that program they have no money at all to purchase internet access. So it’s the only -- it’s a sort of connectivity insurance for them. So I could see an argument for that kind of a program for parts of people -- for the truly poor in Canada as well.

631 COMMISSIONER MENZIES: Okay, I understand that. What about the argument that say -- let’s take it back, data caps exist. We are told -- the argument is that data caps exist to help manage the network, right. But don’t zero-rated programs make it clear that network management isn't required? That there's plenty of unused -- that there's plenty of unused capacity on the network when you can have -- when you're encouraging people basically to consume as much data as they want on these services for free. And doesn’t -- anyway, doesn’t that sort of create a scenario in which data caps become permanent? And doesn’t that work against ultimately against the development of the system to the benefit of the less well enfranchised?

632 MS. LAYTON: So what I would say, if you look at data caps over time, it is a -- they continue to -- they continue to be increased, there's more and more data, the speeds get faster, you get more data for your money. So if people don’t like data caps, I see it as a temporary problem. And there are definitely legitimate reasons to have it, you know. I might like to ride further on the train, but, you know, I buy a ticket for -- one set is not an unlimited ticket, so data caps are necessary.

633 And I think especially if -- to go back to your first question about, you know, is the internet telephony; look at the electricity industry. We pay for the electricity that we use. There is price -- there's price differentiation as well in electricity; in fact, more pricing flexibility than telecommunications. There's a number of zero rating of electricity for alternative forms of energy for example.

634 So the point that if somehow because there's data caps and zero rating and then that’s disenfranchising, I don’t see any kind of -- I don’t follow the argument and I don’t see any academic evidence for it, no test has been done to show that.

635 In fact, in general a lot of the critics against this have yet to share any evidence. Even I think the paper on T-mobile by Barbara van Schewick, she only cites two sources, one is a article from Wall Street Journal with an interview with Salon.com; and the other is from CTIA on a survey about, you know, how should mobile services be regulated. So those are not proper research to see if the premise that you're making does in fact happen. And I would rather than we conjecture about it, I’d rather we actually go and try to test it and find the data that shows that that’s happening.

636 COMMISSIONER MENZIES: Okay, thank you.

637 Given the nature of the Canadian wireless industry where in many parts of the country Canadians have access to maybe three or four wireless service providers in their regions, how would allowing DPPs improve competition -- you think competition in the wireless sector -- specifically, how would it improve competition in the wireless sector if the differential pricing practices and zero rating were to be permitted?

638 MS. LAYTON: So I think the fact that more zero rating has emerged or maybe more critique is indicative of a few things. First of all, that there is already a lot of competition in the market to begin with. And to simply try to compete on price, speed and volume, that is very difficult; it’s especially difficult for the smaller providers. So we can say right out of the gate that differential pricing is actually more valuable to the entrants and the challenging operators than to the incumbents. The incumbents of course have larger networks, so they have less need to price differentially. They can charge the monopoly price, versus the entrants who need to have the flexibility.

639 But what I would say I also see a value of this is to drive diversity and content. You know, Canada has a very important role that it places on Canadian content, and I don’t think you will be able to support promotion of Canadian content unless you engage in some kind of differential pricing. In fact, at the end of the day a lot of this is about how do we pay for content and who should pay?

640 And in many respects, what neutrality is doing is putting all the cost on the end user, which does make it difficult and makes it more expensive for the end users, so that the more that you have the content providers themselves participating in the cost of delivery and the cost of delivering the content, the less the end users have to pay.

641 So when we talk about differential pricing, it’s also looking at an idea of opening up the possibilities that the end user doesn’t need to pay the whole cost, and that there are other ways to get the content to the end user.

642 COMMISSIONER MENZIES: So would it be sensible to you when you were talking about product produced in Canada, for instance to contemplate the public policy in which Canadian product was zero rated, and non-Canadian product was not, so that essentially Canadian video streaming service would not count against your data cap, and a foreign one like Netflix would? Is that ---

643 MS. LAYTON: So I ---

644 COMMISSIONER MENZIES: Is that sort of what you're implying?

645 MS. LAYTON: That makes sense -- it makes sense to me. If you have the argument that Canadian content may be vital for our cultural, social, educational purpose, you could –-- how about the case of elections, maybe that you want to have -- to be able to zero rate the political speeches and that that would enable people to find out what is their representative saying. I mean, what about watching CPAC on a mobile phone and that that was a -- you know, that was a public service?

646 So there -- we could imagine a lot of opportunities where a government socially beneficial content was as low cost or free, would actually encourage the consumption. But again, it isn't guaranteed that it will. I mean you also have to remember the case of New Zealand, that for many users the demand for the foreign content was greater, even when they offered the domestic content for free. But by no means should you discourage -- for example, if there is a way that you deem certain content to be of vital national importance culturally, socially, what have you, to me that makes a lot of sense to offer that at a low cost or free.

647 COMMISSIONER MENZIES: Okay, some would argue against that proposition by saying that actually puts the government or government agencies in the role of regulating content on the internet. And I mean, when you mentioned, you know, trying to get accessibility to political speeches, can you not foresee a point at which somebody would say, “Only approved political speeches is being -- is available for zero rating, and unapproved political speech is not”?

648 And then you get -- people would just flip the argument around and say, “Well, everything should be free, there should be no data caps. And what are you doing, Mr. CRTC or whoever, is you're actually charging me -- I’m having to pay a data consumption fee on all this other content, and you are discriminating in an unjust fashion against me.”

649 MS. LAYTON: So, my main preference would be not to regulate it at all. I don’t see a market failure here, I don’t see consumer harm, I don’t see harm to competition; so I prefer that it would not be regulated.

650 If you have to do something that is a regulatory option, which I’m not -- don’t believe that you have to, I would certainly want to carve out the opportunities to maintain some kind of content that may be vital for people or connectivity insurance if you will. So I, you know, I think that that would be the way.

651 My first – my first best would be don’t regulate at all. But my second best is, if you have to regulate something, please recognize the importance of socially beneficial content first. And this goes to the point of there are -- there is something fundamentally different from an economics perspective of providing free access to let's say things that may be, you know, a free basics kind of health education type of a model versus, you know, guaranteeing that I can get my Netflix for free. Those are not the same thing. Or to say that, you know, oh music has to be free.

652 This is a difficult question politically, but I think that regulators, if you do have to prioritize, this is the case of Greenland, they decided to prioritize health and education above entertainment.

653 Now at the margins we can always quibble what’s, you know –- you know you may love hockey, maybe I love baseball, you know, and one or the other. Therefore don’t regulate unless there’s harm.

654 COMMISSIONER MENZIES: Wouldn’t it –- just to finish on that, wouldn’t it make more sense to simply allow people to -- if they wish to offer zero-rating, to allow genres, so to speak?

655 So speaking of baseball and hockey, so if they could say you could make an offering of so much music and the consumer can pick the music provide or sports and you can choose your baseball and I’ll choose my hockey and et cetera.

656 In terms -– that allows people to offer different rates and different volumes of –- or services, or multiple rates and differentiate in the marketplace, but at the same time avoids the ISP being put in the role of gatekeeper and allows the consumer to be fully empowered in terms of what choice they make.

657 MS. LAYTON: So I guess what I would say is do we call Google the gatekeeper when they are -- you know, when we do a search and they have an advertisement on the side by a certain company?

658 Are they not being the gatekeeper by showing us a particular advertisement at the top of the page or what have you?

659 So I would say here this idea that we need to regulate the telecomm provider and we don’t need to worry about the content side, which arguably I would say has more gatekeeping ability.

660 It’s easier in fact in Canada to change your mobile operator than it is to change your social network or to change your search engine, to change your video provider.

661 So the gatekeeping thesis I don’t accept because I would say, in this economy, in a converged world, it’s the ITs or the platforms, that Silicon Valley has the market power, not the telcos in the same -– in the same respect.

662 And the issue of, you know, isn’t it better that you know consumers just have a plain amount of data, the problem is even if you just give a plain amount of data you still -– you still promote the same entrenched -- you know, what’s popular today continues.

663 So the point of the zero-rating is the ISP has some information. If they want to zero rate X-Y-Z, they know something is valuable to their customers to provide and that they’re making the choice to do so.

664 I think this is a classic high-ec information problem that for somehow or another regulators think that they may have more information than the operator does. It’s not likely in this case.

665 There’s a paper done on zero-rating by William Rogerson from I think it’s Northwestern or Stanford, who looked at this very closely and in these kind of cases it’s not likely that the regulator would know better which bundle or which sport or music than the ISP would.

666 I mean the ISP, in fact, has access to a lot of information, but also what they choose they have to -– they take a risk, because the customer can say I don’t care I’m not interested.

667 As I said, a lot of the thing -– most of the stuff they do consumers reject and, you know, they have product developers every day trying to think well what will -– how do we win the customer from our competitors.

668 So if you want competition you cannot have it unless you allow prices to be flexible.

669 COMMISSIONNER MENZIES: Okay, thank you very much. Those are my questions. I don’t know if my colleagues have any or not. I’ll turn back to the Chairman.

670 THE CHAIRPERSON: Just a few quick questions. You know you suggested that we should test the data, know more what’s happening, and I’m struck by the fact that all the beneficiaries, the third party beneficiaries of zero-rating or other differential pricing, didn’t intervene in our proceeding.

671 And although we sent some RFIs to some parties, the answers that came back were rather thin. What in your view should I draw from that?

672 I mean you’re obviously a researcher, evidence based, one of the potential outcomes out of this is there’d be constraints on their ability, I guess, to be involved in DPP, yet they didn’t come to the party.

673 So what should I conclude from that?

674 MS. LAYTON: Well, so as I said I don’t speak for any of them. I’m not privy to their particular decisions and I don’t know the companies to whom you contacted or what the letters were or what they responded.

675 I would say that to be involved in a regulatory proceeding takes considerable amount of time and attention.

676 A lot of the companies who could benefit from zero-rating don’t have a person to follow CRTC activities, which I think is unfortunate. Many start-ups and so on don’t have any clue about a lot of regulatory issues.

677 I mean I do a test, you know, I ask a lot of folks have you ever heard of net neutrality, they say no and even start-up -– so a lot of entrepreneurs and so on are engaged in their day-to-day business and they don’t think about regulation, how it could help them, how it could not help them.

678 I don’t know about the case of the platforms or what their particular issue is, but what I would say small and medium sized businesses are really not my -– what I think, I haven’t asked them personally, but when I have worked with them they really don’t follow regulatory world because they don’t have time to. They have to focus on their business.

679 But I regret that. I think they could be valuable if they knew more about what was going on in your proceeding. I think it’s very important and could impact them.

680 THE CHAIRPERSON: But the lack of interest extends to large multinationals with huge market caps.

681 MS. LAYTON: So I would imagine they’re probably not Canadian and they probably fear to come in front of your Commission, understandably, because I think they’re probably benefiting quite a bit and maybe not paying their share of taxes, their share of the network costs and so on.

682 So they probably want to keep a low profile. I don’t know for a fact, but I would definitely say, like any number of countries who are –- take a lot of traffic from the United States, they probably feel like -– and understandably so, justifiably so, that these U.S. companies use the roads, they don’t pay for the cost of the roads, they don’t pay taxes and so on.

683 So I could imagine they would be reluctant to be put on the spot. They prefer all the regulation to fall upon the ISPs and not themselves.

684 THE CHAIRPERSON: Earlier you mentioned that the -– that new entrants might benefit from DPP as a way of differentiating themselves in the marketplace, if I were to put to you that those that are using DPP are not the new entrants but the large incumbents, would your conclusions be different?

685 MS. LAYTON: Well my conclusions are different only because I look at other countries where I could say the case of free basics, where you have hundreds of apps from developing countries that are leveraging Facebook’s platform where or the free basics platform to get their apps known and what they’re hoping for is that people use the free basics app as an onramp.

686 They actually -– the goal is that they get a first time internet user on their own free basics app and then transport that user to their flagship website. They don’t want them to stay at free basics forever because they can’t monetize them there.

687 THE CHAIRPERSON: Right.

688 MS. LAYTON: They want them to be on their mainstream app so I can’t ---

689 THE CHAIRPERSON: But I – let’s put ---

690 MS. LAYTON: Canada may ---

691 THE CHAIRPERSON: Let’s assume, it’s a hypothetical question, that in Canada --

692 MS. LAYTON: Yeah, sure.

693 THE CHAIRPERSON: -- those that are predominantly using DPP are not the new entrants but the incumbents. What does that say from a competitive perspective?

694 MS. LAYTON: The -– you mean the incumbent telcos or the incumbent content companies?

695 THE CHAIRPERSON: The incumbent telcos. Rather than ---

696 MS. LAYTON: That says to me there’s competition -– sorry.

697 THE CHAIRPERSON: Wouldn’t it -– it is counter, I think, to what you are saying earlier that DPPs could be used by new entrants to carve out a niche for them in a competitive market.

698 In fact I’ll put it hypothetically to you that it fits the incumbents that are using it. It’s to maintain and retain market share and in fact to prevent a more dynamic marketplace.

699 MS. LAYTON: Well what I would say, the company that maybe makes the most sense that I’ve studied would be looking at T-Mobile in the U.S., which is a large -– it’s the third or fourth operator in the U.S.

700 It’s hard to call them an incumbent, but they are -– they gained 8 million net new customer. That reduces the share of the -– two largest operators. I believe that’s a goal of the FCC.

701 Those customers have a high level of satisfaction. Now you could say in the case of Verizon and AT&T they are racing in their own ways to develop very different kinds of offers.

702 So I would actually say I don’t want to overstate the value of differential pricing and zero rating. I think its value is far overstated. I think in any market you may have one player that will really want to use it, like the T-Mobile case. You can see the large incumbents are going to find different strategies. Verizon is doing this go90, which is trying to be like -- you know, they bought Yahoo and AOL. And then you see AT&T doing a different strategy, which is purchasing Time Warner. So I see three incumbents doing three different things.

703 And with regard to your hypothetical, forgive me, I’m not -- I would say I would not overstate the -- I think that zero rating or this differential pricing is less important than we think. I don’t think it is as important as maybe the media would say.

704 THE CHAIRPERSON: Right. Some have suggested that this is an example of innovation and that regulators should be supporting innovation. Would you hazard a definition of what innovation is?

705 MS. LAYTON: Sure. And I think it was touched upon in your last hearing. You know, innovations are new and I think an attempt was made to distinguish between incremental and fundamental innovation.

706 Fundamental innovation being like those things we talked about in Bell Labs and the transistor. Fundamental innovation is essential to the internet storage, and transport, and video compression and all of these things, by the way, which were results of very large corporate R&D budgets and a lot of government funding.

707 What the fight about net neutrality is some sense about how do we protect this “application level innovation”, which for the large part is incremental, and I find that it’s making kind of a false choice between saying the application level innovation is more important than the fundamental network innovation, you know, I’d like to say that they could both evolve together and that they should evolve together.

708 In terms of zero rating, it is mainly valuable to -- in the short run it’s more valuable to the application level innovation, but in the longer run those companies who could use it to win market share could be able to support the R&D labs and those things to make the fundamental innovation but that wouldn’t be seen until you had a longer term perspective.

709 But, again, it is not -- I think its impact is too small. There’s too many other inputs that would be able to have a firm that could be large enough to create a Bell Labs once again.

710 THE CHAIRPERSON: Right. But here it’s really marketing innovation it’s not really technical innovation. Would you agree with that?

711 MS. LAYTON: Well, marketing innovation, yes. I mean, it could be technical in the sense that there may be marketing software, there could be engineering or some kind of a thing on that level, but we don’t -- we can’t really say that for any one company some amount of incremental innovation could be important for them in the whole scheme of things.

712 Again, I understand your challenge as a regulator, you need to be this -- you have a -- you want to have a crystal ball and you have to do the best you can with the information at hand.

713 I mean, I guess what I would say here is I don’t see the need that you have to necessarily make the choice and fall down and regulate something when the case for the consumer harm I haven’t yet seen. Maybe you have done the analysis, you’ve maybe seen it, and if you have printed and I’ve have seen it, I apologize I’m not familiar with it. But in this instance, wouldn’t it be fine for the regulator to allow both kinds of innovation to flourish?

714 THE CHAIRPERSON: All right.

715 My last question -- I’ll give you a chance to respond. But my impression of the language you use in your submission is the language of competition law, and it strikes me that perhaps you’ve set aside the fact that this is not an industry like any other, that these are common carriers, and you don’t seem to have addressed the common carrier aspect of this with, you know, asking for harm as opposed to realizing that as a principle the carriage providers, the common carriers, should be as neutral as they can be and if they’re not it is up to them to prove why it is beneficial.

716 MS. LAYTON: So let me -- maybe if I might address that with my academic background. I would call myself the person who is the, say, classic regulation in the sense of facilitating market entry, providing the -- you know, the basic services and enabling the spectrum of allocation and so forth.

717 The challenge I have with the common carriage assertion is that I find that this term had a useful life and that is attempted to be reapplied in the case of the internet, which I don’t believe is appropriate.

718 And in many of the industries that were similarly capital intensive, common carrier industries, we have removed the common carrier regulation and they allow to compete freely.

719 In the case of the internet, you know, we’ve had a long history of telephone networks which regulators very successfully deregulated. We have mobile networks. We have multiple ways to get online.

720 So I don’t believe the common carrier perspective is valid. I know many people who do. I know there are many scholars who build their careers on that kind of view. I don’t share their view. I do like to move to a general competition framework to reduce what I see as distortion and A-symmetry where the entrants are able to abuse the regulation against the broadband providers.

721 So when you talk about innovation, my biggest concern -- what I’ve just seen, for example, in the U.S. last week with the FCC, the privacy regulation falls harder on the ISP’s than it does the internet companies. The ISP’s are in a position to create competition to lower the broadband cost to the end users and that privacy regulation under the common carriage idea is being used to limit the market entry to the online advertising, which is probably the biggest problem of oligopoly.

722 I understand regulators are limited in how they can address that other issue, but to the extent that you can allow zero rating differential pricing you can give broadband providers more flexibility to compete with the elephant in the room, if you will, which are the Googles and Netflix’s of the world who many cases you might argue they are the ones who are the common carriers. They’re getting the majority of the traffic and they don’t have to submit to these kinds of obligations.

723 THE CHAIRPERSON: You may want it to be different, but the fact remains that the Canadian parliament, through Sections 36 and 27, have made Canadian telecom carriers, ISP’s, to be common carriers. Would you not agree?

724 MS. LAYTON: That is the choice of the Canadian people and the Canadian government. That is -- I respect that choice.

725 THE CHAIRPERSON: And therefore as a consequence when we operationalize that mandate parliament’s given to us we have to apply a common carrier analysis and not a competition law analysis, which I understand you favour, but that’s maybe not the world we have to live in.

726 MS. LAYTON: Well -- sure -- well, I would say I’m not speaking in position as a legal scholar here, and I certainly respect the Canadian perspective, the Canadian -- the challenges and the way that you look at it. I would still say the regulator has some bit of leeway in how they can interpret and apply the regulation.

727 And what I would simply offer again is that these are complex multisided markets and to bear in mind that in many instances I know that the tools that the regulator has they may not be able to apply them to all of the relevant players.

728 So I would have caution in -- and again, I applaud you for taking this particular consultation. I think it’s very helpful and interesting to see the different perspectives that appear and I think you’ll take all of those things into consideration when you make your decision.

729 THE CHAIRPERSON: Okay. Well, thank you. I understand.

730 So thank you very much, Ms. Layton, for participating in our hearing. We very much appreciate it.

731 And I think what we will do, Madam Secretary, is take a break for lunch until 1:45.

732 Thank you very much. So we’re adjourned.

--- Upon recessing at 12:42 p.m./L’audience est suspendue à 12h42

--- Upon resuming at 1:44 p.m./L’audience est reprise à 13h44

733 LE PRÉSIDENT: A l’ordre s’il vous plaît. Order please.

734 Madame la secrétaire.

735 THE SECRETARY: Thank you, Mr. Chairman, and good afternoon.

736 We will now proceed with the presentation by Sandvine. Please introduce yourself and you have 20 minutes for your presentation.

737 Thank you.

PRESENTATION

738 MR. BOWMAN: Thank you very much.

739 Good afternoon. Thank you for the opportunity to present. My name is Don Bowman. I am the Chief Technology Officer of Sandvine and Sandvine’s network policy control products are deployed in more than 300 networks in more than 100 countries. The primary driver of our business today is service creation. So that's the perspective that I'm speaking to you as a technologist.

740 Differential pricing is, we feel, important in a competitive market. If all players have to sell exactly the same product in exactly the same fashion, we think that only the largest ones will survive. So typically, our customers are the new entrants seeking to find an un-served niche by selling something differently.

741 Differentiation, we find, takes many forms. It can include device-specific plans. For example, my home alarm system which uses a SIM card, OnStar in my vehicle. Application-specific plans, for example, the streaming audio service that we've been talking about this morning, or social networking services.

742 It can also include time. You can be differentiated by unlimited evenings and weekends. We can, for example, have sponsored-data, the 1-800 business model when it was introduced, introduced billions of dollars of related ecosystems such as Yellow Pages and so on.

743 It can be quality-based, for example, based on availability guarantees or based on congestion guarantees. It can be consumer-focused, or it can be use-case focused, for example, internet of things, sensors, and smart-grids.

744 It can be time-based, a day pass, a roaming pass. These are often most appropriate, for example, when I'm in an airport about four hours.

745 Some differentiation is for operational or regulatory reasons. For example, the equivalent of calling 611 or 911 on a mobile device. It could be checking your balance while roaming.

746 Some differentiation is for altruistic or societal reasons. For example, the Free Basics by Facebook and Internet.org in which Sandvine has rolled out in conjunction with Digicel across the Caribbean.

747 With the support from the Government of Guyana, Sandvine’s customer GTT offers free 4G access to information and knowledge from 20 local and international sites regarding e-health and e-government.

748 Differentiation can be a different product rather than just different pricing. For example, T-Mobile’s Binge-On provides unlimited time of video for a fixed fee at a reduced quality. The user has the choice. They can choose to use the reduced quality unlimited time or they can have the original quality and variable volume.

749 Some applications such as sensors reporting on a smart grid don’t need the same network guarantees as others, for example, VoIP.

750 Pricing can provide an incentive to make the network more efficient in the same way that Energy Star ratings did so for appliances. That smart grid, if they have an incentive, they can spread the reporting over five minutes rather than having every household report at the same instant with a higher peak.

751 Virtual Network Operators on both fixed and mobile networks typically act to differentiate their offering to distance themselves from the underlying network technology for whom they would otherwise compete.

752 In the Canadian market, for example, we see TekSavvy with an unlimited off-peak data plan. In the U.S. market, we see operators like Karma with a freemium device. You get a device, you get a certain amount of data for free and if you can convince other people to sign up to the service, you're credited with additional value.

753 We also see specialists like TravelTab which has a no-contract tablet that has an included SIM with travel-specific applications such as GPS which is typically done with a rental vehicle.

754 Differential pricing is not new. We've had lower price long distance calling off-peak for years. I know that we're not, you know, looking forward to the future by the past but I'm saying that there have been reasons that we have come to these decisions in the past.

755 Differential pricing seeks to find a better expression of consumer value, either through price-certainty, which is one of the ones we found most powerful which is exchanging the variable volume pricing uncertainty for a fixed amount of time, which is called zero rating, zero rating of the volume, or through giving the consumer the ability to control costs, peak versus off-peak, or through device-specific plans. For example, a cell phone plan that may or may not support tethering.

756 As the Canadian market evolves towards shorter-term contracts driven by the inability to fully subsidise equipment within the realm of the contract, we think that we're going to grow in the amount of pay-as-you-go and in the amount of prepaid traffic.

757 In this model, price-certainty is more important than ever and provides a way to differentiate plans between data providers.

758 A simple volume-only plan is in effect differentiated here between these types of users and between the applications they use. Streaming audio will not see the adoption it has if users have to worry about a tap running every time they turn it on. You know, this is the main tipping point for that application to allow people to use it many hours a day.

759 Differentiated pricing can create enhanced transparency. For example, a $10 roaming pack is much easier to understand than the unknown bill you'll get when you get home from your vacation.

760 Most consumers are unaware of the volume of an individual application. Does a five-minute voice call use more or less data than a web page? It's something that most people don't know.

761 Would a plan that charged by the minute for voice not be easier to understand? Studies have found that zero rating some data on mobile can cause an increase usage in other access technologies, for example on WiFi which means more use of internet, spawns, more use of internet.

762 We've had one customer in Africa, Econet for example, that found that by launching a flat-rate Whats application, they encouraged a much higher overall utilization of their 3G cellular network. So overall, more people purchased more communications.

763 Sandvine has proposed a set of commercial best practices and also a set of technical best practices. We proposed these to our customers. They're available from our website.

764 We don't believe these provide unjust discrimination or unreasonable preference. Our overall suggestion is to provide a data plan that is across all applications in a class. We talked about the streaming audio class earlier. That might be music but it could also be voice. And there should be no network prioritization. So we don’t propose to intermix commercial charging with technical prioritization on transmission.

765 We also propose that for zero-rated type plans that there be no pay for access, that the money continue to flow in a single direction. This is different than sponsored data.

766 In sponsored data, you know, the voice equivalent industry of 1-800, it's a related ecosystem and we propose in that case it be open for all with an open API and that it be just a similar credit system and just reverse -- reverse in the direction of the toll.

767 Sandvine does not believe that there needs to be and either/or proposition. There is no single answer as to whether one type of charging is better for one person, one device, whether it be by time, by volume, variable or fixed. We all like it to be free. That's not the way it works.

768 Most other industries have multiple methods of charging for similar services, either all-you-can-eat versus à-la-carte. We think that the same should hold true for the data industry.

769 Differentiation is happening globally. It's growing. Regulatory rules around the world seem to leave the door open for this and to encourage a competitive market in Canada, we feel that virtual network operators must be able to distinguish their offerings.

770 Incumbents will be able to -- will need to be able to differentiate offerings to attract or retain customers as their equipment and subsidies subside over time.

771 Thank you.

772 THE CHAIRPERSON: Thank you very much.

773 I'll put you in the hands of Commissioner MacDonald to start us off. Thanks.

774 COMMISSIONER MacDONALD: Good afternoon. Thank you for participating in this proceeding.

775 Over the course of the week and already on the record, we're going to hear a lot about the good, the bad and the ugly associated with differential pricing and sponsored data and zero rating.

776 But from your standpoint as a solutions provider working globally with hundreds of other internet providers, I'm hoping that you are going to be able to educate me a bit on some of the technical aspects and considerations that go into such DPP practices.

777 But before we get into some of the technology, I'm just wondering if with your global set of lenses if you would be able to compare the Canadian networks and the experience that Canadians enjoy with their service providers perhaps in relation to other providers around the world?

778 MR. BOWMAN: So certainly the quality of internet in Canada is dramatically higher than the majority of places that I travelled. I mean it's easy for us to get caught up in the fury of the internet was down of a minute or two yesterday but the reality is that for most of the world, the internet is still an emergent technology.

779 The same is true for bandwidths. I mean it's easy to look to, you know, places like South Korea or Singapore which have, you know, high government sponsored fibre to the home, but the data rates that we enjoy are pretty topnotch overall. So there's still room for growth but I think it's pretty good.

780 With respect to the types of plans that we have available to us, they're reasonably similar. We pay more for data, as you are aware, but we're also a wealthy nation. So there's been a lot of innovation in the Canadian market over the years.

781 You know, we were one of the first markets to launch an unlimited social networking plan, which was done by Blackberry through operators such as FIDO, respond to Blackberry, which had the BIS and the BES type services. Some of the earliest players in mobile apps were in this market, like 724 Solutions and so on.

782 So I would say overall, you know, we're an early adopter of technology, but we sort of -- we run out of gas sort of in the last mile of getting it to, you know, fibre to the home and so on for a lower price.

783 COMMISSIONER MacDONALD: Would you say that the use of data caps is more or less prevalent in Canada than it would be in -- among providers in other countries?

784 MR. BOWMAN: Certainly the use of -- on the fixed line side, the use of charging on fixed by volume is a bit of a mixed bag. There are some countries that have done it, some that have not. Even the ones that have haven't gone universally. You know, in the United States there's a very large number of operators that offer usage-based billing on fixed, but certainly not all. The same is sort of true in Canada. Certainly, there's probably less than a third of the countries in the world that use usage-based billing on fixed.

785 In terms of the mobile side, I would say they were more or less similar. Maybe the packages are less generous than some of the countries, particularly the fixed replacement markets, where there really was never a merged, a fixed infrastructure, for example, eastern Europe. You know, people -- we talk about markets as either what we call fixed augmentation. I have a home internet and I have an away internet. Well, we talk of them as fixed replacement. I have one internet. Which one am I going to choose? In those latter markets you tend to have very large volumes of data at a very low price, but usually get a fairly poor experience out of it. That's been my experience anyway.

786 COMMISSIONER MacDONALD: With respect to the potential for poor experience, how do Canadian networks as they're currently built today fare from a congestion standpoint? Do you get the sense that significant investment would be needed to upgrade? Is our needs for upgrades imminent or what's the current state of affairs?

787 MR. BOWMAN: On the cable side, Canada's rolling out DOCSIS 3.1 today, which has probably got maybe three or four years of life before it becomes even, you know, even sort of average, so I think it's got a ways to go yet. We haven't had a lot of roll out of fibre, as you know, in the fixed line side. I'm a member of the SamKnows program at home and I watch my numbers fairly religiously. We use quality of experience with our customers. Business intelligence is a big part of our offering, and you know, they use that to measure the delivered quality in each of the links in their network.

788 I would say that the network's pretty top notch, to be quite honest. I mean, I -- there's very few times of the day when you can't get the majority of what you want over the internet in Canada. Now, this isn't going to be true in remote areas. You know, they have their own challenges, and it's not going to be true in all of the mobile networks in all of the locations, but if you take the broad average, I think it's pretty good.

789 COMMISSIONER MacDONALD: And when you're making that statement, are you comparing us to, for example, the United States? And I ---

790 MR. BOWMAN: Yeah.

791 COMMISSIONER MacDONALD: --- ask because, you know, a number of intervenors have mentioned that data caps tend to be lower in the United States.

792 MR. BOWMAN: So Sandvine doesn’t feel that data caps are an effective way to manage congestion in the networks, so our counsel to our customers is that there's a discipline called traffic management. We'd get the internet traffic management principles here in the past and that's a discipline of applying a tool, when needed, where needed. So it's not a broad average. So it's -- it doesn’t make sense to say, "the network is congested" any more than it makes sense to say, "the road is congested". It's at a time and a place.

793 When you talk about aligning the broad average of investment with the broad average utilization, that's a different story, so if you want to say, you know, over time, the average user does this so we're going to need to spend that average amount of money next year.

794 But we don’t feel that usage billing is an effective way to drive congestion targets. And that's because, you know, you mentioned earlier the peak and off-peak. It's -- you know, you could have unpredictable events; you could have, you know, two links to a site and one of them gets hit by a backhoe. You get congestion until you can get out there and fix the link. And you could have things that are predictable but from a business standpoint, very difficult to solve. You know, it could be like, SARSstock outside of Toronto or the Olympics. You know, you know what's going to happen and there's a certain amount of technology you can bring in to solve it, but you can't generally fix it economically.

795 And then there's the things that are plannable; you could plan over the course of three months, six months, nine months, and you see your trend of traffic going upwards. So that's -- we don’t feel that the usages is a way to drive either of those targets.

796 COMMISSIONER MacDONALD: Okay, so obviously technology is a component in that and where -- what level the data caps are set at. To put the technology aside, if we're in good shape on that front, do you get the sense that decisions about where data caps are set are largely based on competition versus technology, more competitors in the United States, so they differentiate themselves by having a bigger -- a larger -- a higher data cap than Canadian?

797 MR. BOWMAN: I don't know if there are more competitors per capita in the U.S. On the fixed line side it's probably fairly similar. On the mobile, I guess they have their four national players, but we really have three national players. I think the caps are largely set based on commercial, you know, competitive pricing differentiation. You know, you look at the different operators and they may have some plans that are -- you know, there's a bundle. There's -- you know, it's speed, it's features, it's volume, and it's differentiated from its neighbouring bundle.

798 One of the things we found challenging is, you know, the fixed industry never found a way to monetize a number of devices, a number of terminals, and that's what's currently driving the growth in the fixed line today. You know, when we first -- when Sandvine first started, growth was driven by one factor: new users coming on the network, you know, 30 plus percent per quarter. And then it was driven by two factors: those users were coming on very quickly, plus they were using a very disruptive application, peer-to-peer file sharing. The network looked like a (inaudible) symetric.

799 Then a streaming video came online. The number of net new users started to drop, but growth was now driven by a new application. And a new application is dramatically less efficient on the network. It took something like peer-to-peer file sharing, which used the network over 24 hours a day, and you squished all that entertainment into live, into 4 hours of the day. The network, the peak and the off-peak sort of stayed apart from each other.

800 Today, the growth rate of those applications, while still very good, the primary driver of growth is net number of devices. Since they don’t monetize it in the same way that the mobile industry does with SIM cards, some of them have looked to build plans around speed and volume. But as they’ve got into higher speeds, it's -- not all users respond. It's not obvious that there's two (inaudible) value in 100 megabit plan over a 50 megabit plan. A lot of people are content so they act to differentiate them.

801 COMMISSIONER MacDONALD: I'm struck when we think about usage, oftentimes when we're online or what we're doing online differs greatly, depending on time of day. And we see usage, you know, drastically increase in the evening. So I'm wondering when plans are being divided, why are data packages calculated on a monthly basis versus time of day. If -- you know, if we assume people come home from work and then they're streaming the news and the kids are on Netflix or what have you, wouldn't some type of scheme that takes that into consideration be more appropriate, much like utilities do today?

802 MR. BOWMAN: Well, some do. Some operators have peak and off-peak plans. You see that more commonly with satellite because their cost per bid is dramatically magnified. But you do see it. I mentioned the TekSavvy plan here today. I mean, there are some plans in Canada that have peak and off-peak pricing in it. I think it's a combination of something that approximates cost and also is understandable by the consumer.

803 You mentioned peak time, but the peak time varies by day; you know, vacation days and snow days have a different peak. On WiFi ISPs, their peak is around noon. It's not always the same for all operators in all locations. So it would be very difficult to sort of post a -- like, an internet busy signal and say -- you know, think about Uber with their surge pricing. Users hate it, but it's very efficient. So yes, the network would be more efficient if we had surge pricing, but I don't think people want it very badly.

804 COMMISSIONER MacDONALD: But since you're a solution provider, is that technically possible? Can the plans ---

805 MR. BOWMAN: It is.

806 COMMISSIONER MacDONALD: --- be designed that way? Can the appropriate monitoring equipment be installed to (inaudible)?

807 MR. BOWMAN: It could.

808 COMMISSIONER MacDONALD: Okay.

809 MR. BOWMAN: There's an IATF standard that we've been working on for many years that it sort of prices per bit and it uses economic means to build the network to the point where no bits are dropped and everybody is incented along the whole chain. It's a very nice idea, but it's -- again, from a consumer understanding point, it's very difficult.

810 You know, even the utilities that you mentioned, in a commercial environment, you can agree to have your air conditioners, you know, cycle on and off to, you know, during peaks. But from a home user, they publish a peak set of hours. It's an approximation of the peak, so I've set my pool pumps to run during these four hours, but it's unlikely that that's the exact peak in my neighbourhood.

811 COMMISSIONER MacDONALD: So there would be a certain amount of guesstimates that would go into ---

812 MR. BOWMAN: Well, from a technical standpoint, you can implement whatever you want. From a market acceptance and an explanation, I think it would be challenging to do more than approximate it with peak and off-peak.

813 COMMISSIONER MacDONALD: You mentioned TekSavvy as an example, but are there any facilities-based providers internationally that would be doing that today?

814 MR. BOWMAN: There are definitely satellite providers that do that are facilities-based.

815 COMMISSIONER MacDONALD: Okay. In paragraph 13 of your intervention you stated that -- I want to make sure I get this right:

816 “Conceptually differential pricing, I would include any pricing plan other than unlimited data as even volume-based differentiation in a plan can change the cost to subscribers.”

817 So, should I take from that that differential pricing is already widely in market, and this is -- when we talk about things like zero rating and sponsor data, that this is just the next evolution of differential pricing?

818 MR. BOWMAN: Yeah, so that’s our opinion, is that there's many forms of differential pricing, and I mentioned some of the longest standing ones we’ve had. You know, the Amazon Kindle was very successful, they produced a device, you bought a book, the book came sort of magically to your device, you didn’t have a relationship with the carrier. It happened to use the infrastructure of the cellular network. You know, my truck is magic with OnStar, I can check the amount of gas that's in it from here, but I don’t pay a subscription fee for that. These are types of differentiation.

819 Differentiation in terms of selling roaming packs, I mean that’s become the norm of the last -- the last few years, you know, you pay per day rather than the very high amounts, that’s been there. You know, with Blackberry launching the social networking plans in Canada, that was some of the earliest in the world. I mean those have been around here since maybe 2007-2006.

820 So, yeah, I think there's been a lot more differentiation that people give credit to. You know, it’s not always the same, it’s not always zero rating. When people talk about rating, they're talking about multiplying something by a price. So you can either rate by time or by volume or by both, or something else I suppose. So, yeah, and I think there's been a fair bit.

821 COMMISSIONER MacDONALD: Can you perhaps expand for me what the potential harms or benefits would be, both from a service provider standpoint and from a consumer standpoint if, you know, if we were to on one end of the spectrum ban all forms of zero rating and sponsored data differential pricing, versus open the floodgates and let market forces dictate?

822 MR. BOWMAN: Well, in the one end of the spectrum to have truly undifferentiated pricing, you would probably have to do a couple things. You would probably have to disallow charging by volume entirely, because otherwise you have an inherent differentiation because some applications use more volume than others. So, if you say you can't differentiate but you can put caps in, one gigabyte of cap is a heavy disincentive to use Netflix.

823 You would also probably have to increase the amount of transparency in terms of reporting. This is something that it’s very -- it’s a big struggle, so a lot of the world’s regulators, the CRTC included, have embarked on sort of sunshine clauses, you know measurement initiatives. And there's more measurement initiatives out there that you can shake a stick at, and they often give contradictory answers, you know, between M-Lab and Sam Knows and so on.

824 You know, the FCC in their last network neutrality proceeding put in the, you know, that you have to report the latency loss and jitter at the peering and interchange points, which I’m on an advisory committee for and we’re trying to discuss how to define that. One of the harms that comes out of causing people to publish the drop statistics, is that’s it's obvious that to a lot of people that higher drop is bad, but as a technologist that’s not actually true, you know.

825 The only way that the internet slows itself down to avoid running away is by dropping packets. There's a phenomena called “buffer bload” which is if I was an ISP and I wanted my drop stats look better, I know there's a bunch of latency to prevent it and that VOIP doesn’t work and gaming doesn’t work. So you can actually get unintended consequences, you know, publishing that stat and having people point to it, is going to cause the behaviour of making the network worse.

826 So, you know, if you go to the entire spectrum of banning any form of volume-based billing, then I think you’d disincent a lot of the investment in the network because it doesn’t align the average cost with the average usage.

827 In terms of disallowing some of the other ones like disallowing a device-specific plan, that’s just an opportunity lost. It might have been a new virtual network operator that was planning to do that that doesn’t launch in the market or something like that. It might be the Kindle that doesn’t come along. It's really going to be hard to say without going case by case.

828 COMMISSIONER MacDONALD: From a technical standpoint, whether data is -- counts against a data package or whether it's zero rated or it’s sponsored data, it’s still data traversing a network. So, I guess the first question -- so from a technical standpoint, there is no difference?

829 MR. BOWMAN: Well that’s not true actually. So a lot of people think a byte is a byte, but it actually can cost more to deliver. So ironically, you know, if you were to do 100 megabytes of streaming audio and 100 megabytes of web, the 100 megabytes of web would cost the operator more to deliver, and that's because of the bursty nature of it. So in a mobile network, it's more efficient to get on the network, deliver a big chunk of data and get off. It's more efficient from a battery standpoint, it’s more efficient from a spectrum standpoint. So there can be a difference in cost, it may not be dramatic. So the most expensive devices are ones that use a very constant rate but very low. So they -- they're constantly using up spectrum, using up batteries, using that up from other players.

830 COMMISSIONER MacDONALD: So then, what would be the incentive or perhaps it’s just purely a marketing incentive on the part of the service providers, what incentive would they have to technically zero rate one service over another? Why would they economically want to zero rate video for example, when they could zero rate access to a website and potentially that have less of an impact on their network?

831 MR. BOWMAN: So, for an equivalent volume delivered video will cost less than the website. I don’t know if that’s a significant driver in that process, but it will. That will vary depending on the network technology chosen, so that will be very different for satellite versus cellular versus fixed-line infrastructures. On the fixed-line infrastructure there will be not a material difference in cost there. So why would they want to do it? I think it is more about the differentiation of the service, you know, delivering a better consumer value or a more fixed consumer value to it.

832 In some cases, it may be that that’s the only convenient way to charge for something, it may not -- there may be no other method like, you know, in a censor-style network I may want to have millions of censors out there, I may want to charge per censor rather than by byte because I can sort of predict the overall amount that they're going to drive. I also know that it’s something that can be sort of squished during times of congestion.

833 There's value in having multiple -- you know, a heterogeneous network is actually a lot more efficient than a homogeneous network. If you have a lot of different classes or things all vying for it and some of them are more willing to accept latency and loss than others, you can get the network closer to that target of 100 percent utilization. So you may want to incent the use of a class of service that is able to withstand congestion, because it sort of brings your noise floor up in the network.

834 Particularly, you know, with the streaming data the way it is today, you're going to see a huge difference between peak and off-peak utilization. That off-peak utilization represents an opportunity, you know. We’ve paid for the capital to deliver that, the spectrum is out there, the battery is out there, the fiber backhaul is there, but nobody is using it, you know, it's terrible.

835 If you could find something that uses that -- now, there's only so much you can do with economics, you know. If you told me internet was half price, you’d just have to surf at 3 a.m.; I’m not going to take that bargain. But if you said that my smart car can download its firmware updates at 3 a.m. and the manufacturer saw that incentive back to the Energy Star rating, I think they would put it in. And we do see plans like that emerging today. There are devices that will shift the time that Netflix buffers its next video based on your predicted viewing habits since you're off-peak.

836 COMMISSIONER MacDONALD: So, some would argue that instead of getting into the complexities of zero rating and sponsored data -- because I'm sure it will be complex for a number of different reasons -- that why not just increase the caps, why not just increase the data allowance, why not just make things unlimited?

837 MR. BOWMAN: M'hm.

838 COMMISSIONER MacDONALD: From what I'm hearing from you, would I be correct in my understanding that that might not allow for as complete network utilization ---

839 MR. BOWMAN: Yeah.

840 COMMISSIONER MacDONALD: --- as a DPP scheme might?

841 MR. BOWMAN: Certainly I feel that that would lead to a less efficient network, and a more efficient network is in everybody’s best interest. You know, today there's -- depending on the market, between maybe 20X and 2,000X difference in the volume delivered on a fixed and a mobile network. You know, there's different reasons for that. And one of the reasons is the primary driver on the mobile network are smart phones, and I'm not using it the same way I use my home internet connection because of screen size and battery. But the second one is clearly price, you know, if you were to move to deliver that type of volume over the mobile network, you would need much more small cells, you would need more backhaul, you know, it would cost more money. You might need more spectrum to do it economically.

842 So, if you move to totally unlimited, some of those operators, they may not be able to provide that and their next best alternative will be to provide a lower quality of service in some areas. They may not be able to predict those very accurately; it might move around. That might in turn dis-incent the use of certain applications.

843 Like, I don’t know if you use voiceover IP on the mobile network; that’s a bit hit or miss around the world today. So if you use Skype, WhatsApp calling, BBM calling, they can work very, very well over LTE and they can work quite badly, and you don’t know in advance.

844 So I think that if you want that problem to get better rather than worse, you have to make sure that the average cost and the average volume sort of align with each other.

845 COMMISSIONER MacDONALD: Because Sandvine is involved in helping providers manage these types of solutions, if you could quantify for me in terms of complexity and cost ---

846 MR. BOWMAN: Yeah.

847 COMMISSIONER MOLNAR: --- what would -- what would it entail for an ISP to come forward with various zero-rated or subsidized content versus just opening up the pipe, because those are two separate ---

848 MR. BOWMAN: Sure ---

849 COMMISSIONER MacDONALD: --- conversations from a technology standpoint?

850 MR. BOWMAN: Sure. So there’s several different techniques for doing the differentiation. The first one is the one that’s enshrined in the -- in sort of the standards -- in the 3GPP Standards and so on. And that’s sort of how LTE and related services are delivered; there’s a technology standard called IMS; there’s a control plane; the person who has the application, they signal to the network and say, “I want to deliver this. It’s going to be on these coordinates. Okay.” That’s sort of how it’s done.

851 The second approach, we talked a little bit about this morning. It’s the sort of the IP list. You know, this is one that T-Mobile has documented as their API, for example, and they say, “Here’s my API. You bring your list of what you’re going to bring to me; we put it in a contract; and away you go.”

852 The third is using approaches more similar to ones like ours where you sort of infer signatures based on the content.

853 And then the fourth is you get them to explicitly stamp their traffic in some fashion; it’s called in-band signalling.

854 Each of those has a different cost. The IMS core infrastructure one is something that’s going to be in place, anyway, as a consequence of providing Volt and E-911. So there probably wouldn’t be a materially increased cost there. It’s got some flexibility and scaling concerns; it might limit the types of service that you could do this with. It’s designed around things like rich communication services like push to talk over cellular and sending pictures and things of that nature.

855 If you talk about the sort of the fixed set of IPs and the prearranged contract, it's very operationally complex for the content providers to do that. You know, there's a limited number of IPv4 address space left in the world. There's none left in the public pool, so you're kind of scrambling for it. And because of the nature of how CDNs work, they want to get the content as close to the eyeballs as possible. That needs a lot of IPs, but then if they have to segregate the IPs by carrier and content type for charging, it's expensive as well.

856 These -- none of these techniques would be anywhere near the investment you're talking about to materially expand a network, though, like, a small fraction. Like, if you look at a mobile carrier, the part of the budget we sit in is what's called the packet core, and that would be, you know, a few percent of what the radio network would cost. So that's sort of the overall.

857 Now, there can be some what are called OSS costs. OSS is the charging system and things of that nature, so if you were to assume a world in the future where they didn’t charge for anything and had no charging system, I guess that could be dramatically simpler. But if you're assuming some incremental amount of complexity in there in terms of professional services hours to companies like Amdocs and so on, it's a bit hard to say, but it should be a small fraction of the narrow capacity, I think.

858 COMMISSIONER MacDONALD: Would that -- would those same statements hold true regardless of the access type, or ---

859 MR. BOWMAN: No.

860 COMMISSIONER MacDONALD: Okay, so certain solutions would be easier to implement, obviously, on fibre versus DOCSIS versus satellite ---

861 MR. BOWMAN: Yeah.

862 COMMISSIONER MacDONALD: --- versus mobility?

863 MR. BOWMAN: So the mobile industry has a lot more tools in it, you know, things like voice over LT have driven a lot of design decisions in there. The fact that it -- much of -- the vast majority of the world is on true prepaid. We don’t have a lot of true prepaid in Canada. We have pay as you go, but prepaid means you put money in and it counts down towards zero. When you hit zero, you're cut off. And that's how the vast majority of the world consumes their internet today.

864 So because those tools are baked in all of the equipment, there's a set of standards that we've had to implement to do that. There's something called the online charging system, there's top ups and voucher management, so they have a ton of tools.

865 When you talk about something like DOCSIS -- so DOCSIS has something called PacketCable Multimedia. They have a reasonably good set of standards there, and when you look at something more like DSL or GPON, it would fall off a little bit from DOCSIS.

866 COMMISSIONER MacDONALD: So the next thing I was going to ask, because we were having the conversation earlier today, is whether DPPs are really innovative or whether they're just an innovative pricing strategy. So now that you have taken us in behind the network curtain, if you will, would you be of the belief that they actually do represent an innovative way to run and expand and transform your network?

867 MR. BOWMAN: Definitely. I mean, I wouldn't say necessarily all are, but there's quite a few things that our customers have done that are highly innovative that have -- you know, all parties involved have benefited. You know, I mentioned the case that we did in Zimbabwe. They had rolled out a 3G network and they had no material use of it. You know, people were staying on these unlimited talk and text plans on 2G phones, like, old flip phones. And by tariffing the over the top rich communication services in the text tariff, they dramatically increased the amount of utilization of the network, and then in turn, those people started using an internet experience.

868 I mentioned the Kindle earlier. I mean, you know, the Kindle seems fairly normal today, but the Kindle launched 10 years ago, and at that time, it was completely magic, that you could have a device that you could buy and the book would be there. It would always be there for you. I don't think Amazon would have been as successful as they had in the book ecosystem if they had not thought of that.

869 And that was a bit of an innovation a couple of different ways. You know, there's no way they would have been successful in trying to convince people to have a data plan on a secondary internet device at the time. And by just bringing it into the cost of the book, they were able to make that just more or less work for everybody.

870 You know, OnStar; I mean, OnStar's been around for a long time, and it's not as exciting as it used to be because lots of cars have technology in them, but the idea that you could have your doors unlocked remotely, I think that was pretty innovative.

871 You know, we mentioned there's a lot of different device-specific plans or other things there, but you know, you don’t know what's going to come next, and I'm very excited about the internet of things. I think for the internet of things to really take off they're going to -- somebody is going to need to figure out a way to service a large amount of devices on the network without worrying too much about how they're using the network.

872 I think if somebody on the other side is going to take the bargain to say that these devices are willing to be, you know, given at lower class of service so that I can build my network closer to 100 percent utilization, I'll price it less. You know, you're buying in volume. So I think those are innovative.

873 COMMISSIONER MacDONALD: You stated in your intervention that customers really weren't interested in buying a specific volume of data. They're more interested in what they can get ---

874 MR. BOWMAN: Yeah.

875 COMMISSIONER MacDONALD: --- and what they can do with that data and differential pricing is an easy way to explain to customers what they're going to get and it's easy to understand and it avoids bill shock. But you know, we've seen in the interventions that there are also a lot of pitfalls with programs like zero -- or potential pitfalls with programs like zero rating or sponsored data.

876 Are there any ways to make it easier for people to understand data and what they can actually get from it without perhaps going near the pitfalls of differential pricing?

877 MR. BOWMAN: So we've tried. I mean, we've put tools in place that give much more real-time advice of usage. You think about, you know, the -- you know, the wireless code of conduct and the roaming bill shock and you know, you think about time lag on that. So time lag is your main villain there. So what you don’t want to do is you get a notice to say, "Warning, you’ve now hit 900 percent of the maximum amount you wanted to spend." You want to be able to see it a priori.

878 So I think you can educate around that area in the same way, you know, like, I'll go back to my Energy Star example. A lot of people don’t know what a kilowatt hour is, but if they're looking at two dishwashers for the same price, and one uses less kilowatt hours, they can kind of get their heads around that.

879 I don't think that there's going to be any real value in society in trying to go back to primary school and educate people on, you know, what a byte is with respect to an internet application because they change. You know, Netflix, in this year, they dramatically increased the efficiency of their MPEG encoder so that they could use less CDN capacity, less disc capacity, less bandwidth, and deliver the same experience to the home. Their overall net bandwidth actually went down for the same number of hours of video delivered for the same quality. You know, that wouldn't happen, a) if there was a -- if there was no incentive around cost data, but b) you know, it would -- can you imagine going back and resetting people's expectations?

880 So I told you an hour of video used to be a gigabyte and this year, the new big thing is virtual reality, so it's two gigabytes, and next year, oh, well, the MPEG committee has figured out how to fix that, so it's 800 megabytes again. Like, that's very hard.

881 COMMISSIONER MacDONALD: I guess understanding that obviously it's important for customers to understand, you know, what they're buying and what they can do with it to avoid, you know, to avoid bill shock ---

882 MR. BOWMAN: Yeah.

883 COMMISSIONER MacDONALD: --- you know, being a primary problem, especially in a world where, you know, you maybe have the kids upstairs accepting the overage charges on their device. Can you quantify for me how big of a problem that actually is in the Canadian or in the global context?

884 MR. BOWMAN: No, I probably can’t. You know, the counsel we give our customers is nobody wants variable charging; right? So you think back to your cellular plan, you got the 100, 200, 400, 800-minute plan. You never went over because the overage charges were, you know, were a threat. You always bought yourself into the next plan. And they usually called you. You know, you’re at 600 minutes. You know, you’re going to hit your 800 minute. Do you want to move to the 800-minute plan? And that’s one of the permitted uses of data is to help users get them on the right plan.

885 So we feel that the best practice is to either have prepaid, which, you know, was for people with no credit and it counts down, or to have a model where as you get close to the limit you’re presented with an advice of usage. You’re now at 75 percent. Are you aware that this plan is available? This may better suit your needs. And then you don’t actually hit a cap plus variable amount. So that’s our best practice.

886 But, no, I don’t have information on the specific number of overages in the Canadian or other markets.

887 COMMISSIONER MacDONALD: In paragraph 21 of your submission from June, you outlined your recommendations for ISPs who were considering launching zero-rated plans. So I’d like to unpack that a bit if I can.

888 You identified lumping different applications and content into particular buckets or particular classes. And when we asked for some additional information you outlined a number of different potential classes of applications or content. In -- from your experience, how many different classes of services would there be out there that could potentially take part in a zero-rating program?

889 MR. BOWMAN: So twice a year Sandvine publishes the only true census of the internet which is our global internet phenomena. We publish how people around the world consume internet. And we have 11 categories that we use in there. And that’s probably the best answer here. I think it is a mix in each market of what things are good for technical reasons or commercial reasons to zero rate. Not all would probably respond well to zero rating. But we feel that there’s 11 basic categories, so real-time entertainment, online storage services, gaming, et cetera.

890 Some of those have different technical needs on the network. Like real-time communications has a different technical need in sort of the bidirectional latency than real-time entertainment does.

891 In terms of the practical ones that we see out there, by far the most common are the unlimited social plans and the unlimited messaging plans. So, you know, from the messaging stand plan point it could be WhatsApp, Tinder, Grinder, Viber, those types of applications. From the user standpoint they achieve the same goal. You know, they -- their social networks are usually forked. You can’t have -- like I would love to have just one messenger on my phone, but my wife uses a different one than my work colleagues so I have nine. So -- that’s not how I got to nine, but there’s a bunch.

892 So, there are -- those ones respond really well. Because from a consumer standpoint, the value we feel is in the number of hours they engage with the application for those types of things. So, you know, users that use Facebook, the average at the time we did the measurement was 18 hours a day of engagement. So from the user standpoint that’s the value.

893 A carrier’s cost is in terms of bytes. So we look at the ratio of those two hours of engagement to volume on the network to make the proposal for which things will respond best commercially to zero rating.

894 So the worst ones will be the ones the users hardly use or they’re not aware -- they use them by inference. You know, we talked about advertising earlier. It’d be really hard to convince a user that they should pay extra somehow for an advertising plan. This is not going to happen so.

895 But anyway, there are 11 categories that we have. There’s probably five, six categories that work really well to differentiate. The streaming audio one has been very, very popular. You know, for a lot of people that’s their commute time music, their XM radio replacement. And they wouldn’t do that at all unless they had the price certainty of its fixed fee. I’m not going to get to the end of the month and be out of balance in a prepaid world. I’m not going to get to the end of the month and have an overage. You know, I just know I’m going to commute on average an hour a day. One day was an hour and a half because there was traffic. That’s what I want. I want it to be fixed.

896 COMMISSIONER MacDONALD: And, I mean, you mentioned nine different messaging apps that you use ---

897 MR. BOWMAN: Maybe six.

898 COMMISSIONER MacDONALD: Maybe six. It’s still a big number in that context. Who should, in an ideal world, decide which app is in and which app is out of a particular class? How do you draw a firm line between two different classes of service?

899 MR. BOWMAN: So our recommendation and our best practice is to treat them all, all apps of a class the same and to make it open. So it may not be possible to do all of them. There may be diminishing returns. There may be one that is just not used in a market. But we recommend not just picking, you know, the market leader in an area and leaving others alone.

900 So if you look at how a lot of our customers launch their plan, they’ll launch it as an unlimited social plan. And underneath it they’re going to have a Twitter logo and a Facebook logo, I don’t know, MySpace logo, I don’t know, whatever, a bunch of logos. And then they’re going to have an -- probably not MySpace. And they’re going to have another one that says an unlimited messaging one. It’ll be Vibe or Line, Tinder, Grinder, Whatsapp, Hangouts, BBM, and they’ll have a little star below it that says if your favourite app isn’t here, write us and we’ll put it on a list.

901 COMMISSIONER MacDONALD: So then you would -- you wouldn’t see it as necessarily being an exhaustive list of every single app that’s out there. They’d choose a certain number of popular ones using some test and then additional ones would be added down the road upon request?

902 MR. BOWMAN: Yeah, so the -- so the way it currently works is there’s an open request. So the person that owns the app can ask or the consumer can ask to have their favourite app added. There isn’t currently a method by which applications -- there’s no taxonomy on the internet today. So from our standpoint, it would be easier if there were. It’d be very nice if there was a way that, you know, in the internet headers they could sort of stamp and say I’m this class -- I’m this rating class of application.

903 We have proposed that to some of our partners there to put it in the encryption header so the encryption header would allow it a way to be cryptographically secure so it couldn’t be tampered with, but still provide the privacy. And you could say I’m a messaging app. And then if you had a contract with them you could say, okay, well, messaging apps are allowed to run within this envelope. If you stay within that envelope you’re not cheating me and then I can offer it. And that would be a great technical integration point. But certainly the industry’s not there at the moment.

904 COMMISSIONER MacDONALD: What do you think the uptake would be? I mean, if I’m the customer of a service provider I’m one of maybe a million individuals subscribing to that service and they may only get five requests for Grinder, for example. So what realistic uptake do you think that there would be in an ISP proactively going out and trying to establish that contract based on either an individual request of a consumer or a request by the application provider?

905 MR. BOWMAN: Well, I think from the ISP standpoint, more doesn’t actually cost them anything. It actually gives them increased value. So they’re incented to have the biggest number on the list possible, because you’re not using two at the same time. It’s not like you double your usage by adding a second one there.

906 In terms of the, you know, the cost contractually, it’s going to depend on how tightly or loosely coupled you are. So we don’t currently propose that in the zero-rated type plans that there is any commercial coupling to the content provider or the application provider. There may be a contractual binding to say, you know, you agree to these terms, but there isn’t a commercial like dollars and cents binding there.

907 If they’re unilaterally done, which means that the ISP does it on their own accord, they go into equipment like ours and they hit the configure button and they make it go. There are some applications that we don’t propose doing that with today because they’re not accurate enough for some reason. But for the most part, it wouldn’t be that big a cost.

908 So I think they’re incented to do as many as they can. If there really is only five people using it, then I would suggest that it isn’t really a driver of the internet as a whole, per se. Most of the apps tend to be either very large or disappear quite quickly.

909 COMMISSIONER MacDONALD: From a technical standpoint, how do you tag that data? So you know that, you know, maybe I use Facebook and maybe I use Twitter. Facebook might be in, Twitter might be out ---

910 MR. BOWMAN: M’hm.

911 COMMISSIONER MacDONALD: --- of a zero-rating plan. How do you know that that’s -- that I’m going to Facebook and zero rate that versus something I might post on Twitter?

912 MR. BOWMAN: Well, in the specific example of Twitter and Facebook, both of those are 100 percent encrypted protocols today. The vast majority of the internet encryption occurs using a protocol called SSL, https to date. So our proposal -- we believe that by the end of this year into early next year we’re going to be about 80 percent by volume encrypted on the internet. So we’re big proponents of encryption.

913 In the specific case there, there’s a field -- when you set up the connection in the encryption setup, you assert I want Twitter and it inserts that I am Twitter. So in the same way your bank site guarantees you’re going there and you know it’s not somebody foraging it or phishing it, Twitter does that as well with their SSL exchange, and we use that as the signature for it.

914 So that’s the majority of the ones that are interesting are already encrypted today. There are some that aren’t. So they may use like the http host name or other fields.

915 COMMISSIONER MacDONALD: And from a management standpoint, both on the side of the app or the content owner and the service provider, how is that relationship managed? Because I’m thinking on the side of the service provider it takes significant time and effort to manage what could be a very, very long list of zero or sponsor rated ---

916 MR. BOWMAN: M’hm.

917 COMMISSIONER MacDONALD: --- sites versus, you know, I’m sure the Twitters and the Facebooks, the big guys in the world have large staff and employees that can help manage that relationship. Does that make it easier for them to get access than some of the smaller new application provider, someone just starting up?

918 MR. BOWMAN: Well, so sponsored data is different than a zero-rated plan. So in a sponsored data arrangement, you could not help but have a relationship with the content provider. There isn’t a lot of sponsored data applications in the world today. This one is perhaps about to grow, perhaps not. We feel that it’s going to -- there’s going to be a set of aggregators come in there that are going to hold those multi-party business relationships. So instead of every ISP negotiating with every content, it’s going to be every ISP negotiating with three aggregators and negotiating with every content. That cuts the number of contracts down dramatically. But there aren’t a lot of examples of that today.

919 In terms of the zero rating one, you know, most of those are done more or less unilaterally. So the ISP puts together a plan of these. In some cases they ask those providers to technically integrate to them. In some cases they don’t. But there is not the same sort of tie in.

920 The sponsored data one has a set of APIs, which are sort of technical interchange points and you publish that standard. And then anybody who wants to can come and program to that standard without asking for permission.

921 So that’s sort of, you know, the -- I’m trying to think of another example, but, you know, they’re trying to make the network something that anybody can buy something off of or anybody can use something off of, in the same way that anybody can buy something out of the app store today. You know, you don’t have to be -- you don’t have to have a relationship with Google in order to use them as a payment processor.

922 COMMISSIONER MacDONALD: So would all of the apps or content providers knowing that they would obviously be hoping tohave their service picker up by as many subscribers as possible on as many different internet service providers as possible, does that create complexity in the relationship from a technical standpoint? Do I -- you know, if I have a great new app or new service that I want to include in a zero-rated plan, do I need to build to one standard for Telus versus another standard for Bell versus another standard for Rogers?

923 MR. BOWMAN: Not currently. In the mobile industry the standards are pretty firm. So the 3GPP standards group has created all of the charging standards and they are more or less religiously followed across the board. So I wouldn’t say so, no.

924 COMMISSIONER MacDONALD: So it’s more of a global standard than the ---

925 MR. BOWMAN: That is a -- that is the global standard that is won out for almost all of the integration challenges that are out there. So that comes from the European Telecom Standards Institute, so that’s used for almost all charging service creation, rating, everything. It’s all done from there.

926 COMMISSIONER MacDONALD: Do you foresee any scenarios where a service provider could or should -- other than the app doing something illegal, other than that, are there situations that you could envision where a service provider would refuse to allow that app into a particular zero-rated class?

927 MR. BOWMAN: Well, they will definitely refuse to allow a class differentiated charging. So, you know, I go back to my example of hours of use and volume. The application that had the highest hours of use and the lowest volume was Facebook. The one that had the lowest hours of use and highest volume was YouTube. So it would be a rare operator that’s going to say, like I’d like to for the same amount of money have a lot more of that on my -- an unpredictably large amount more of that on my network. But they’re probably going to say that as a whole across video.

928 So T-Mobile is kind of forging new ground there with their video Binge-On program. I haven’t seen anyone else attempt that with video just yet.

929 With respect to audio, no, I think once you’ve got one audio app in there, they’re all very similar. So if you could make more of your users happy for the same network utilization, I think you would always take that bargain.

930 COMMISSIONER MacDONALD: What implications would there be for the service provider if on a routine or constant basis the data that was ascribed to zero-rated classes was actually outpacing the amount of data that subscribers were using from non-zero-rated or non sponsored classes.

931 MR. BOWMAN: Well, if you’re doing it right it is, because I think that’s the intent. It depends on whether the thing you’re trying to encourage more of has any different cost on your network or not. You know, so going back to my home alarm system at home. I don’t have a PSTN connection to my home. There’s only one provider which will only provide a monitor alarm system which was Rogers. And I think that they know exactly what that’s going to cost on the network, that those -- when your door opens and closes there’s only so many per day. And that’s a great way to yield based utilization of that network. Either want to sell as many of those things as they can.

932 You know, if you’re talking about, you know, doing zero-rated audio, you know, you’re going to build a business model together that says, you know, this is what I think the growth is going to be and the zero rating is going to cause an increase in that growth curve because you’re going to have more people using it and you’re just going to build that in. You’re going to say it’s going to be part of my customer acquisition costs, part of my technical rollout. Most of the operators we work with today have fairly simple rules for what they call capacity planning, when do I add additional capital. And they look at the quality at each link. And when it hits some target they add capital and they split the link. So in terms of -- it might be utilization, it might be -- we use a measurement called round trip time which measures congestion. So when you get a certain number of congestion minutes on a link you split it and now you have two. So I don’t think those rules typically change just because you’re doing zero rating somewhere upstream. The part of your business that’s paying for the capital and rolling it out keeps doing that based on observed demand, not projected demand.

933 COMMISSIONER MacDONALD: So if the intent is to ultimately have more zero-rated data than non-zero-rated data, if that is a goal, it ---

934 MR. BOWMAN: I don’t think I said that. I said it should grow faster.

935 COMMISSIONER MacDONALD: Okay. So not out -- so there shouldn’t be more zero-rated data consumed than non-zero-rated data, you just think ---

936 MR. BOWMAN: I don’t know if there should be or not, I guess is what I’m saying.

937 COMMISSIONER MacDONALD: Okay. Whether it’s more or less it would theoretically would be ---

938 MR. BOWMAN: Should grow more quickly at the beginning. That’s why you’re doing it.

939 COMMISSIONER MacDONALD: Okay.

940 MR. BOWMAN: There’s a pent-up demand.

941 COMMISSIONER MacDONALD: And presumably it will be, if not the majority, at least a sizable chunk of the data that a subscriber uses, so ---

942 MR. BOWMAN: Could be.

943 COMMISSIONER MacDONALD: --- does that mean that ISPs either would or perhaps entertain the idea of actually lowering their data caps as opposed to maintaining or increasing them?

944 MR. BOWMAN: I guess they could. I’m not sure. I mean -- you know, I think the size that they put on for their volumes they allow at a certain price is a combination of where they are, their generation of technology cycle, roll-out, you know, market pressures, et cetera. You know, like I said, I don’t feel it’s a way of driving capacity -- excuse me -- capacity gating on a network so.

945 COMMISSIONER MacDONALD: Do you have concerns around free data like this actually altering consumer’s behaviours? I think you’re trying to mitigate that to a certain extent by having classes that are encouraged to have the most content and the most apps, but do you still have -- do you have any concerns that I’m more apt to get my zero-rated Netflix over my non-zero-rated something else?

946 MR. BOWMAN: So, I’m one of the world’s biggest fans of the invisible hands, so I, you know, Adam Smith drives a lot of things. Our experience has been the more internet you use, the more internet you use. And it doesn’t really matter what that is. We can all think around the room to the first internet used. So for me, it was University of Waterloo, it was VAX, it was NetNews, you know, very textual experience. And the more I used of that, the more I used of it.

947 So, you know, we have customers -- we have a customer in the Philippines that they’ve done a promotion that you get 30 megabytes of data every day for free, but you have to -- every day you have to click I want my 30 megabytes. And just that fact of people using that 30 megabytes drove a ton more internet usage, because people started -- now that they’re using it they’re used to it. They want more of it. That just drove a ton of adoption. So that’s been our experience is, you know, whatever you give -- in the free basics by Facebook, you know, giving people access to Wikipedia and a couple of newspapers, it’s droven [sic] a huge amount of people using a lot more internet outside that box than you would ever imagine. The uptake was huge.

948 You know, my example in Zimbabwe with Whatsapp, like the uptake was enormous. You know, you had people that were never using any data experience and now they’re using a little bit of data experience and they’ve got a friend send them a link saying “there’s a thing called the web, you should check it out” next thing you know they’re internet experts. So that’s been our experience.

949 COMMISSIONER MacDONALD: So more usage lowers the cost. There’s economies of scale there. Providers can ---

950 MR. BOWMAN: There can be, yep.

951 COMMISSIONER MacDONALD: --- monetize their networks better.

952 MR. BOWMAN: Yeah.

953 COMMISSIONER MacDONALD: Does that also lead to one class of subscribers, perhaps high users, subsidizing a lower class of subscribers?

954 MR. BOWMAN: I think it depends how you setup the plan. I mean, I’m a firm proponent of if you use more you pay more, within reason. I mean, you wouldn’t ask me to pay for the network for everybody, but I use more than the average person, I think I do pay more than the average person.

955 So, you know, you’ve got the economies of scale there. You’ve got people are gated by the number of eyeball hours they have to spend because they’ve got a job, you know, the quality of the equipment they have available too, it’s battery life and so on, and how much they can afford.

956 But everybody wants to communicate. I mean, that’s the 21^st century. And it doesn’t really matter whether it’s for voting online or checking the hockey score, it’s -- you know, people want that. So, you know, we want more of that in peoples’ hands. What we found is the zero rating is a good tool to get more of that in peoples’ hands.

957 COMMISSIONER MacDONALD: So more usage lowers cost. Would that same objective also be met if not through zero rating if through providers increasing their data allowances or making them unlimited packages?

958 MR. BOWMAN: I think if you can come up with a way to make everybody truly unlimited and you can make the economics of that work and you can keep it sustained then that could be a different thing. I don’t think it will be easy to keep that sustained though.

959 COMMISSIONER MacDONALD: You had suggested an opt-in approach for zero rating. Can you explain that to me a little bit further?

960 MR. BOWMAN: Sure. So, you know, T-Mobile’s example, that’s how theirs work. So the user -- Binge On is a choice that the user has and there’s no commercial cost to the choice of Binge On. So if you have one of the SIM cards in these plans Binge On is available to you. Some people are arguing whether it should have been opt-in or opt-out.

961 But, anyway, you have a portal on there and you can go to it and there’s a knob you can twist, and you can say I would like to have unlimited volume of video, I acknowledge that it will be capped at one megabit, standard definition. You can twist the knob the other way and say I would like to have unlimited quality of network, whatever -- the unlimited quality of video, whatever the network will deliver I acknowledge that that will come out of my data plan. And that’s the choice the user has. There’s no commercial incentive one way or the other. They can go one way or the other.

962 There’s a lot of plans that will work that way over history. So there’s a lot of -- there’s sort of two different types of data plan providers, the ones that have unlimited volume at limited experience. So, you know, they talk about them as exceptional use policies. We don’t give you a specific cap but as you get more -- as you use more and more data it gets slower. So that’s been one type. And then the other type is you have an unlimited cost and constant quality, which is more the style that’s common in Canada.

963 So different people have a different tolerance for one or the other. Some people don’t want any variable costs in their life and they’re willing to settle for a degradation of quality. Other people are prepared to take variable costs and have fixed quality. So that’s also an opt-in choice. In that one you opt-in by picking your plan.

964 COMMISSIONER MacDONALD: And based on your best practices, how often do people opt-in and opt-out. You opt-in and you’re always opted in until you go in and change it, or you re-evaluate on a monthly basis, or ---

965 MR. BOWMAN: Every shiny new knob in my life I’ve twisted a few times in the first month or so and then it’s forgotten. So, yeah, that’s what happens.

966 COMMISSIONER MacDONALD: Okay.

967 MR. BOWMAN: You know, people aren’t changing, and people aren’t going to say “I’m going to watch this one movie let me make it go back to high quality and back.” It doesn’t really work that way.

968 COMMISSIONER MacDONALD: You also suggest I think you called it a My Three concept, where providers -- sorry -- consumers could go in and choose the three applications that they were most interested in. Maybe for one provider it’s Netflix, Facebook and something else, and so forth. In your mind, does that remove the roll of -- or remove the concern that many have expressed with respect to service providers acting as gate keepers?

969 MR. BOWMAN: So we have experimented with that both from a congestion quality control -- there’s a phenomena called self-congestion where, you know, you may be sharing a device which has lost lots of capacity but in your home there’s congestion because there’s other people in your home that are chewing up your fixed resource. So, you know, from a -- we’ve proposed a model where people would self-prioritize and say, you know, I prefer my Skype to be prioritized over my Netflix.

970 You could do the same with zero rating. I’m not sure how well that would work commercially. Like, that would be something that somebody could explore. You know, they could put a bouquet of ones out there and say you can pick three from the list, or something like that, or five from the list. Like, that would be -- sure, that would be a fine approach, in my opinion.

971 COMMISSIONER MacDONALD: But it would still be a list generated by the service provider not ---

972 MR. BOWMAN: Well, I think at the limits of the equipment like ours there would have to be -- there is an implicit list. So our equipment we support 1,300 protocols today in our primary list and 2,200 in our secondary list. You know, there’d have to be some way of identifying it.

973 So, you know, I don’t know if it could be truly -- you know, you could -- you and your brother could write an application and put it on the list. Like, I don’t know if it would be that simple. But maybe we can get there somebody I suppose.

974 COMMISSIONER MacDONALD: If you’d ever met my brother you knew we wouldn’t be writing an application together.

975 MR. BOWMAN: If you’d met my brother you know it’s possible.

976 COMMISSIONER MacDONALD: You’ve said you’ve done some testing. Has this actually been implemented anywhere?

977 MR. BOWMAN: I’ll try to think. There are some public studies online of carriers who have done that. There was one in the UK called Plus MAT, which did provide that end user ability knobs to twist to do the prioritization. That was more of a technical prioritization. Some of the other ones we’ve done they were more proof of concepts where we demonstrated this.

978 You know, in our customers -- as you must imagine, consumer carriers there’s a point of sales complexity, so, you know, if they have infinitely complex plans then they can’t train their sales staff to sell it. There’s a help desk complexity. So they have to evaluate, you know, what’s the best technical approach, what’s the best commercial approach, and then what are the costs, and sometimes there’s a compromise in true Canadian fashion across all of those.

979 So we often propose or demonstrate things that they take under advisement and then some revised fashion of that comes out later, you know, something simpler.

980 COMMISSIONER MacDONALD: With respect to some privacy concerns, you’ve mentioned that the solutions you’ve put forward, your tech knowledge, doesn’t raise any concerns, but later you also said that -- or at least indicated that DPP’s are not possible when a subscriber uses VPN.

981 MR. BOWMAN: Well ---

982 COMMISSIONER MacDONALD: So that raises some concerns that ---

983 MR. BOWMAN: Well, I guess, there’s several different types of VPN. There’s what called split-horizon VPN’s. When I run my corporate VPN my corporate traffic goes to the company and my internet traffic still goes to the internet. So that doesn’t present any real challenge.

984 There’s VPN’s that are based around devices. So like again the Blackberry, the Blackberry Enterprise Server does something similar than that.

985 Some users turn their VPN’s on for specific purposes, whatever they want. So at that time you would send a message saying your zero rating has been temporarily suspended, or in your terms of service you would say this is not compatible with the use of that VPN technology.

986 Some VPN’s don’t provide any privacy. Some VPN’s are just ways of tunneling something over something and they don’t provide any lack of visibility. It’s a more complex question there.

987 So there’s a small number of users that use VPN’s on their devices. Usually those users aren’t the ones that you’re trying to target some of these services at, and they’d be able to make the commercial decision which would they prefer more I suppose.

988 In some cases the VPN could be how you provide the service. In the Blackberry unlimited Facebook plan was provided because they had a VPN. That’s how it was done.

989 COMMISSIONER MacDONALD: So, in your mind, you don’t feel that DPP’s are a disincentive for people to use VPN’s? You don’t think that they’re going to change their behaviour in that regard?

990 MR. BOWMAN: No, I don’t think so.

991 COMMISSIONER MacDONALD: Earlier today we heard from Canadian Media Concentration Research Project -- and I apologize, I don’t know if you were in the room -- but ---

992 MR. BOWMAN: I was here.

993 COMMISSIONER MacDONALD: --- in their opening statements they raised a potential concern and it was with respect to access providers clearly requiring greater awareness of their customers activities to engage in DPP’s then they otherwise would and raised the concern because providers would be aware of what their customers were potentially doing that there was an opportunity to use deep packet inspection as part of some of the zero rated sources. Would you like to discuss that at all and perhaps tell us whether you would share that concern or have seen such practices ---

994 MR. BOWMAN: Sure.

995 COMMISSIONER MacDONALD: --- hopefully not in Canada, but perhaps elsewhere in the world?

996 MR. BOWMAN: So deep packet inspection is one of the underpinning technologies inside Sandvine, we’re the market leader in that space. It’s used by a lot of different things today. So every time you make a voice call, you know, deep packet inspection is used in what's called your session board or comptroller to map up the flows between the two different service providers. It's used in the network address translation because we’ve run out of IPv4 space, there's a lot of different uses. So we tend to focus on the use cases. We talk about, you know, this is the use case.

997 So here you're talking about something that you would be putting in the contract with the user and charging them by. So it's been a long-standing tradition in Canadian telecom that if it’s a function of providing the service, that I collect that for the purpose of providing the service, for the function of sizing the service and for the function of diagnosing the service. And then I could use that data then later in aggregate with a third party or internally as a first party to market a better plan to the users. That's sort of the general purposes of information.

998 So our equipment is capable of calculating for example the volume of Facebook that you do. It doesn’t know the specifics of it, it doesn’t the posts that you do or the websites you go to, but it does know, you know, how many minutes for day of Facebook, how many minutes per day of Twitter you use for example or how many bytes. So, if that’s what being zero rated on your behalf, then we will start collecting that. And then that will show up on your bill, you know, 412 megabytes of Facebook at $0 would be what you would see. That information will then be available to size the network, diagnose it with you, and then used later in aggregate.

999 So from our standpoint, that doesn’t create any additional privacy concerns than the traditional CDRs would in invoice or call data records, you know, your long distance calling, they kept who you called for how long, et cetera. And those are the same purposes that it was there for.

1000 COMMISSIONER MacDONALD: Okay, thank you. Just a couple more questions and then I’ll hand you back over to my colleagues. But some intervenors have suggested that we should perhaps consider introducing requirements for ISPs that offer zero rated services, to enable their subscribers to be able to track their zero rated data usage and what the cost would have been associated with that. Do you think such a provision adds any value to the end consumer? And if so, should we impose it as a condition of providing the service?

1001 MR. BOWMAN: I don’t think it’ll add value to all users, it will certainly add value to some users. I don’t see it as being particularly negative to anyone, so I don’t have a problem with this certainly.

1002 COMMISSIONER MacDONALD: And another suggestion was put forward with respect to transparency that all ISPs that are entering into a zero-rated arrangement with content providers or various applications and in situations where there may be an aspect of sponsored data, that the ISPs be required to disclose all of those details on their website, including any financial arrangements and terms and conditions. What are your thoughts on that? Is that information that customers should have and want to have?

1003 MR. BOWMAN: I struggle to think that the consumers would use it, but I suppose that industry advocates and watchdogs would probably use it in aggregate, so it’s probably good overall for society. I don’t think the average end consumer would use it though. So I don’t have a problem with either, but I just ---

1004 COMMISSIONER MacDONALD: Okay, thank you. Those are my questions.

1005 THE CHAIRPERSON: Thank you. I believe those are all our questions for you. Thank you very much for being present. Thanks.

1006 We’ll take a short 10-minute break till 3:10, and hear from the final intervenor before we break for the day.

1007 Thank you very much.

--- Upon recessing at 2:59 p.m.

--- Upon resuming at 3:11 p.m.

1008 LE PRÉSIDENT: À l’ordre, s'il vous plaît. Order, please.

1009 Madame la secrétaire.

1010 THE SECRETARY: Thank you, Monsieur. Chairman.

1011 We will now proceed with the presentation by the Canadian Network Operators Consortium Inc. Please introduce yourselves and you have 20 minutes for your presentation.

1012 Thank you.

PRESENTATION

1013 MR. TACIT: Thank you, Madam Secretary.

1014 Good afternoon, Mr. Chair and Commissioners. My name is Chris Tacit, I’m counsel to CNOC. Seated to my right is Christopher Hickey, CNOC’s Director of Industry Affairs; seated to my left is Matt Stein, Vice-Chair of CNOC. As you may be aware, while Matt is also CEO of Distributel, he’s appearing here today on behalf of the CNOC membership as a whole. Mr. Stein will deliver the first part of our opening statement.

1015 Thank you.

1016 MR. STEIN: Thank you, Chris.

1017 This proceeding is meant to examine differential pricing practices, or DPPs, related to internet data plans. As in all hearings, the Commission must balance the interests of various stakeholders. In this case, those stakeholders are consumers, ISPs and content providers.

1018 We believe that striking an appropriate balance must preserve the flexibilities that ISPs need to innovate and offer the services that consumers demand, so as to benefit them to the greatest degree possible.

1019 In fact, DPPs offer real benefits. These include opportunities for innovation, increasing choice and enhancing access to online content, as well as increasing affordability that consumers demand.

1020 Yet, at the same time, DPPs also have the potential to result in undesirable and dangerous outcomes such as: the anti-competitive impacts of an ISP applying differential pricing practices only to its own content in order to give itself an undue preference; ISPs engaging in gatekeeping that can prevent end-users from accessing like-for-like content on an even playing field; content providers being harmed if they are excluded from a differential pricing practice; and market distortions due to end-users having an incentive to choose content based on a DPP, and not on the merits of the content itself.

1021 Throughout this proceeding, we strive to put forward a proposal that strikes a proper balance to address these benefits and associated concerns.

1022 There are two other issues that have arisen during the course of this proceeding that are of particular interest to our members, and so we will be commenting on them as well.

1023 Those items include proposals by some parties for: restrictions on the use of economic traffic management practices that fall within the scope of the ITMP framework, and the prohibition on the use of data caps on retail plans for Internet access; and the imposition of technical requirements on ISPs to advise end-users in real time when they’re moving out of a zero- rated zone to a non-zero-rated content.

1024 CNOC opposes these proposals and will explain its reasons for that opposition, but first Chris Hickey will explain our proposal and how it can operate in practice.

1025 MR. HICKEY: Thank you, Matt.

1026 Our proposal would establish a DPP regime that is premised on certain principles designed to address the concerns that Matt described. A key element of the regime is that it must be applied by all ISPs in a transparent manner. Complaints regarding whether certain practices or schemes conform to the regime would be addressed by way of applications brought pursuant to Part 1 of the Telecommunications Act.

1027 This is precisely the approach that the Commission has adopted in its ITMP framework, which has worked very well to date because it has enabled ITMP practices to evolve with changing network economics and technologies, rather than cementing them based on the circumstances that were in place when Regulatory Policy 2009-657 was issued. By replicating this approach, an equally robust and flexible framework can be applied to DPPs.

1028 Our regime would not allow the sponsoring of contents in a manner that allows an ISP to obtain a financial benefit from the owner or other right holder. Therefore, our DPP regime focuses on the zero rating of contents.

1029 The principles that would apply to our DPP regime are as follows: first, ISPs should be permitted to apply zero rating to broad categories of contents, or to all contents of broadly defined categories uniformly, but not to specific content alone; an ISP must clearly identify the broad service category that it intends to make subject to a zero-rating program, and make those categories public for both the benefit of content providers and end-users; the terms and conditions for a content provider’s content to be accepted into a defined zero-rating program must be fully disclosed and available on the ISP’s website, with the exception of any technical or operational provisions that are of a confidential nature in the normal course of trade; an ISP must include, in any given zero-rating program, all contents of a defined broad service category so long as the content provider is willing and able to adhere to the ISP’s posted terms and conditions; ISPs must identify the content that is included in the zero-rating program in a manner that is transparent to actual and potential end-users; the zero-rating of tools and resources that allow end-users to manage their data consumption or their accounts should be permitted.

1030 And finally, a zero-rating program must not involve any activity that would violate section 36 of the Telecommunications Act.

1031 We believe that ISPs should be -- should present clear terms and conditions to content providers regarding DPPs. Reasonable commercial terms dealing with technical and operational matters should be allowed.

1032 I would like to stress that it is important for the Commission to allow ISPs to require content providers to adhere to reasonable terms and conditions in order to be included in a DPP scheme. The administration of DPP programs and the addition of content to them from time to time can require significant technical and operational efforts. It is, therefore, important for ISPs to have the flexibility to manage such programs in a manner that recognizes these realities and does not interfere with or disrupt their business operations. This approach will enable ISPs of all sizes to implement such practices.

1033 We also believe that ISPs should be allowed to require content providers to certify that content is legal in order to be included in a DPP category. Consequently, we are also asking the Commission to authorize ISPs in advance to deny or remove any content that is not legal from a DPP program.

1034 The principled approach we are proposing would ensure that vertically integrated entities cannot prefer their own content, since such practices, if allowed, would interfere with the launch or uptake of other potentially new and innovative services and distort competition in the market for the online distribution of contents.

1035 The transparency inherent in this proposal would enable any interested party to have the evidence required to challenge any DPP scheme should it believe that the scheme contravenes section 27(2) or section 36 of the Telecommunications Act, thereby addressing concerns relating to anti-competitive behaviour.

1036 By keeping content categories reasonably broad, eliminating the sponsorship and vertical integration practices already discussed, and applying a transparent approach to the entire regime, concerns regarding possible gate keeping of contents by ISPs would be addressed.

1037 For the same reason, the exclusion of content from DPPs, which could harm the corresponding content provider, would be held in check.

1038 Importantly, this approach would also avoid market distortions caused by end-users having an incentive to choose content based on a DPP and not on the merits of the actual contents.

1039 This principled, flexible, forward-looking, and transparent framework will enable ISPs to operationalize DPPs in a manner that is readily compliant with the Telecommunications Act, and for parties to challenge any non-compliance that may arise.

1040 I will now turn it back to Matt to address two related issues that have come up in this proceeding.

1041 MR. STEIN: Thank you, Chris.

1042 Before discussing the specific issue of data caps, I wish to stress that we do not think that any changes need to be made to the Commission’s ITMP framework as a result of the adoption of a DPP-related regime. Regulatory Policy 2009-657 already established a robust and flexible ITMP framework that is working well.

1043 Practices that treat all content the same, such as usage-based billing or off-peak unlimited, do not raise any of the concerns associated with DPPs. Therefore, there is no need for DPP policy to cover those activities.

1044 However, some parties are arguing particularly strenuously for a mandated elimination of retail data caps. The reality is that caps constitute an ITMP that can be used to address network congestion. Thus, requiring the elimination of data caps could cause significant problems for network operators and potentially degrade end-user experience.

1045 In addition, the mandated removal of data caps would be harmful to competition, because it reduces choice.

1046 The vast majority of CNOC members already offer both capped and unlimited data plans, and both types of plans are in demand in the workplace -- excuse me, in the marketplace. Where consumers are used to having the ability to choose a lower cost plan that includes a data cap, removing data cap options just removes the choice of selecting the lower cost option. Since users can choose either capped or uncapped plans, such a requirement would amount to significant interference with the operation of competitive markets and would be inconsistent with paragraph 7(f) of the Telecommunications Act, as well as the Policy Direction.

1047 We believe that an ISP that chooses to implement a DPP should be subject to transparency requirements to ensure that all customers are aware of what services and content are subject to differential pricing practices.

1048 Some have suggested that technical requirements should be imposed on ISPs to advise end-users in real time when they are moving from a zero-rated to a non-zero-rated zone. This is an inappropriate measure. It will interfere with the online activities of consumers, resulting in poor end-user experiences. It is also unnecessary because, under our proposal, an ISP will be required to transparently identify the broad service categories and services that it intends to subject to a zero-rating program.

1049 Chris?

1050 MR. TACIT: Thank you Matt.

1051 This concludes our presentation. Thank you for your attention. We would be pleased to answer your questions now.

1052 THE CHAIRPERSON: Thank you very much. Commissioner Vennard will start us off.

1053 COMMISSIONER VENNARD: Good afternoon, and congratulations to making it to the -- almost the end of the day. I hope you enjoyed the hearing, got a lot out of it so far.

1054 You provide us in your intervention and your presentation that you just give us -- gave us with a lot of different things that I think we need to look at how do they all kind of come together. There's quite a few things there that I think are different than what we've heard, and in my mind, a lot of these pieces don’t really go together very smoothly either.

1055 But to start with, what I'd like to do is just something on kind of a practical level. I understand that CNOC has about 30 members, competitive ISPs? Of your membership, tell me -- tell us about if there are any DPPs that are being employed by any of your members now. Have you run into a problem with it? What's the landscape like now among your membership?

1056 MR. STEIN: I'm not familiar with any DPPs haven't been rolled out amongst our members.

1057 COMMISSIONER VENNARD: Okay. Have there been any attempts to do that, do you think, or that you're aware of? Here I'm wondering if you’ve encountered any barriers to doing that?

1058 MR. HICKEY: No. When we were looking at this proceeding, the -- as we were putting together our position and having member input, no member raised that they had been trying to bring a DPP to market, other than of course, any type of -- unless you assume differential pricing such as Sandvine was talking about with respect to unlimited peak -- or sorry, unlimited off-peak and those types of practices. Our members have brought those types of practices to market, but in a DPP such as zero-rating or sponsored data, those types -- or you know, that definition of DPP, our members have not been experimenting.

1059 COMMISSIONER VENNARD: Okay.

1060 MR. HICKEY: They (inaudible).

1061 COMMISSIONER VENNARD: Have they -- has anybody tried and encountered any barriers in that respect, or ---

1062 MR. STEIN: No, not that we're familiar with.

1063 COMMISSIONER VENNARD: Do you know if anybody's thinking of it?

1064 MR. HICKEY: No, we do not. Those -- that's participated and responded to the Commission interrogatory requests did not mention that they were preparing to experiment and no members have made clear that they're also planning to experiment either.

1065 MR. TACIT: What I can add is that members have certainly expressed that they would like to see the certainty flowing from the outcome of this proceeding before they commit themselves to certain paths, but like the other -- like my other two colleagues, I'm not aware of any who haven't actually implemented or vigorously attempted to implement anything at this point in time.

1066 COMMISSIONER VENNARD: Okay, that's good. It was just for a record. I wanted to see whether or not there had been any barriers that we needed to talk about here. But I see we can just move on for that.

1067 Now, do you see any reason why small competitors would not be able to execute a DPP like unlimited music? Do you see any reason for that? It's just not part of the plan so far?

1068 MR. STEIN: Just not part of a plan so far. No, I don’t see any reason that they couldn't.

1069 COMMISSIONER VENNARD: Okay. So you don’t see any technical reasons why that would not -- your membership just isn't really doing this yet or thinking about it? Okay.

1070 MR. STEIN: Correct. No, there's no technical barrier there.

1071 COMMISSIONER VENNARD: Okay.

1072 MR. TACIT: So let me just say that we are thinking about it. They're just not doing it yet. As I said, I am very serious about the fact that we were trying to get some regulatory certainty around this first.

1073 COMMISSIONER VENNARD: Yeah, I get it. Okay, and on that note, let's have a look here at what sort of framework you think would bring about some certainty for your segment of the whole internet market. And here we can start by looking at your broad categories, and maybe you can just expand on what some of those might be.

1074 You suggested quite a few categories in there. They seem to be very very large.

1075 MR. HICKEY: So I think when we’re looking at defining categories in our response to a Commission request for information we had put together effectively a principle that would be use to define categories and ensure that they’re not overly narrow. And it’s basically to try to identify like for like services.

1076 So flowing from those broad categories we think that the broader the category, the less concern with respect to discrimination from keeping like for like parties out of that category would occur.

1077 So effectively, if you’re looking at video streaming or you’re looking at audio streaming and you’re looking at those types of broader categories, those are the type that we’re thinking are broad and would lead to less discrimination concerns rather than very narrow or targeted type definitions for ---

1078 COMMISSIONER VENNARD: Okay. When I look down the list you've got non real-time video, non real-time audio, non real-time gaming, web browsing, audio-based communications, video-based communications, non multi-media based communications, email, Google, hangouts and so on; Cloud storage, file sharing, and social networking. These are your suggested categories.

1079 Who should -- who do you think should be defining what should go in one of these categories? How do you see that working?

1080 MR. STEIN: The ISPs themselves. So these are just examples that we cited, and you’re right. We listed a great number of examples because as we started talking about it a great number of examples came up.

1081 THE CHAIRPERSON: Right.

1082 MR. STEIN: Our membership is pretty diverse with over 30 Canadian ISPs. A lot of examples came forth pretty quickly.

1083 But like I said, the ISP themselves would select the category, propose the category, fill in that category, and so on.

1084 MR. TACIT: I just want to sort of maybe add a nuance which I think is present here but perhaps wasn’t present in our earlier submissions, and that is that when we were tying to look at how to balance especially the subsection 27(2) concerns we thought that the broadest categories possible would have the most minimal chance of infringing that provision, whereas dealing with one individual piece of content or application would raise all of the concerns that we discussed in our presentation and in our submissions.

1085 That is not to say that it’s not a continuum; it is a continuum and it’s certainly possible to have other broad categories that aren’t as broad as those specific ones that were cited in our evidence. It is a graduated scale and the -- you know, with time and experience ISPs will determine a) what’s marketable; b) what works in terms of the regulatory obligations and so on.

1086 So it wasn’t meant to be crystallized as just those categories, but those are examples of some of what we thought were the broadest possible categories that would probably not raise any subsection 27((2) concerns.

1087 COMMISSIONER VENNARD: Maybe you could help me out where with trying to figure out what would regulatory certainty look like with all of these very very broad categories? Where would those boundaries be in terms of regulation?

1088 MR. TACIT: So the idea here is to apply some principles and as with a lot of undue discrimination provisions it’s not completely possible to create sort of a water tight list or an absolute list of things. But can discuss a few things that would be helpful, I think.

1089 So one is if you’re looking at a category, however we define it -- how many potential members could there be in that category? Are we talking one or two, or are we talking 10, 5, 20? Obviously, the larger the number of possible participants who would experience no barriers, no substantial barrier to joining the lower the concern would be.

1090 The other would be, okay, how many -- what sort of the market share of -- so it’s -- a lot of it is built around sort of traditional competition economics considerations. And another key element is the degree of substitutability. You know, if I say I’m going to have a category of one social media and one video streaming, that’s not really helpful in terms of trying to address the subsection 27(2) concerns because they’re not even partially substitutable with each other.

1091 So I think when you dig down into this it’s the same sort of economic criteria that the Commission is used to applying when it looks at other types of situations where these types of -- so you know, you've got to look at what sort of the market is, so to speak. And the market is the group of services in question. How substitutable are they? How many people have -- how many participants have the opportunity to be included? Are there any that are disproportionately large that will swamp everything else out so as to make the category meaningless?

1092 Those are some of the considerations that would apply when one is looking ex post at what the category looks like and whether it might infringe that section or not.

1093 COMMISSIONER VENNARD: Yeah. Let’s just pause for a moment on that idea of substitution. To what extent -- I mean, my concern here is that with your suggestion that those categories do end up being kind of meaningless because everything is in it. So would you, for example -- if we use music as an example -- would you consider radio to be in that category as well?

1094 MR. STEIN: Well, the idea would be -- the idea that we’re proposing would be that radio would be a substitution for radio -- other kinds of radio would be a substitution for radio, whereas a library based music service where rather than just experiencing a stream perhaps with or without a DJ and so forth you get to go and select what you’re looking for -- that might be a different category.

1095 But again, that would fall to the ISP. The ISP, when setting up their categories may choose to include both together that allows for flexibility between them.

1096 COMMISSIONER VENNARD: Okay. What role would CRTC have in -- would you be thinking of like an approval role for something like this? Or how do you see the market forces working? I’m not sure where we would fit into these categories in establishing and possibly even oversight on these types of categories.

1097 Maybe you could just expand on that.

1098 MR. HICKEY: Of course, I can walk through. I think a lot of our inspiration came from the ITP framework. At that time there was many different practices, technological, economic, and ways that we’re talking about managing network congestion at that time. And what the Commission did at that time was set out a principled framework that set the bounds of appropriate practices.

1099 On one end of the spectrum was you couldn’t use a technological practice that ended up blocking contents or that would be considered a breach of 36. On the other end there was no (inaudible) rules as to here’s what’s acceptable and here is what -- you know, here is the economic or specific technological practice that’s acceptable.

1100 I think we’re looking for a similar boundary. So on one end we know that based on past decisions the Bell Mobile TV issue where you can just zero rate your own contents to the exclusion of all others, represents one end of the spectrum.

1101 The other end of the spectrum that if you were to zero rate all services, like for like within a category, that would represent the other end. And within that, should a service provider define a category then if someone thinks they’re being excluded they can bring a complaint under Part 1 and the Commission would then make a decision.

1102 So the Commission’s upcoming decision based on substitutability of unlimited music and radio would make a principle that would then apply to that framework. So everyone knows that that’s the scope of substitutability. And we build on that framework and obtain certainty.

1103 COMMISSIONER VENNARD: So the regulatory certainty would then be enshrined in a set of principles that would be developed?

1104 MR. TACIT: That’s correct. Much like the ITMP framework where are general principles and then it; up to individual entities to complain if they think that the principles as being breached. We’re not suggesting that there be a flood of applications to the Commission to approve each one of these categories, just as there isn’t any application to approve any new or existing form of technical or economic ITMP that a particular ISP uses today.

1105 And I might add that when the decision came out initially there was concern all around that this would lead to a whole ton of Part 1 applications for undue discrimination and so on. And the reality has been quite different. The industry adjusted actually quite quickly to that reality and found ways to work within that. And at the end of the day, the Commission ends up being a decision-maker of last resort in those cases because even where commercial parties have disagreements, they often are able to actually resolve those disagreements on their own without recourse to the Commission.

1106 So this should not, in our view, take up a lot of oversight resources.

1107 COMMISSIONER VENNARD: Okay. We’ll return to some of that, but there was one thing I just wanted to mention and ask you to at this point expand a bit on that. When you’re saying that -- in number 18 of your presentation:

1108 “We also believe that ISP should be allowed to require content providers to certify that the content is legal.”

1109 This is not really in line with the common carrier principles and those concepts that we’ve heard. Are you thinking differently in that sense or would you want to ---

1110 MR. TACIT: Here’s our concern. We’re concerned that right now, you know, Section 36 sort of prevents the cutting off or even sort of saying that something doesn’t qualify for a program. Once it’s in -- the problem is this, once something is in the basket, if it turns out to be illegal, how do you make sure the ISP isn’t seen as promoting that content anymore?

1111 Whether it’s copyright infringement, whether it’s child exploitation material, whatever it might be, the ISP’s here are looking for a mechanism which would enable them, which the Commission would say is okay, to say okay you may qualify in terms of the basket but if -- you know, if your video is unlawful video we have the right not to necessarily zero rate that and make it look like we’re promoting that content.

1112 COMMISSIONER VENNARD: So ---

1113 MR. TACIT: So that’s really what this is about.

1114 COMMISSIONER VENNARD: So with the idea that the common carrier principle shielded the -- originally the telecommunications but now in the -- if we think of the internet or we think of the network as being a common carrier, no more than a platform really, then the ISP’s are shielded from liability. Are you -- how does that ---

1115 MR. TACIT: Well, there’s still ---

1116 COMMISSIONER VENNARD: How does that fit into that and how does that ---

1117 MR. TACIT: There’s still an issue -- you know, there’s still a live issue, and this is perhaps a broader issue than just the DPP issue, but I think there’s still a broader legal issue that remains a little bit unaddressed as to how and when a common carrier can actually cut off somebody because they’re providing content that’s illegal when a common carrier is supposed to generally be content diagnostic. It’s a much bigger issue than this.

1118 COMMISSIONER VENNARD: Would you like to comment on it? Where do you think that is? Where is CNOC on that?

1119 MR. TACIT: Well, I mean, certainly, you know, if you’re looking at things like, for example -- the best example that I think probably everybody in this room could agree with, child pornography or child exploitation material, it would be wonderful if the Commission could come out with an app priori rule that says this material under Section 36 yes you can block it, because right now there is no such ruling under Section 36 per se and so the -- you know, the status of telecommunications common carriers that want to actually stop such content from being transmitted is a little bit legally up in the air.

1120 But again, you know, I know we’re sort of digressing from this.

1121 COMMISSIONER VENNARD: Yeah.

1122 MR. TACIT: In this context, we’re not even saying block it we’re saying don’t make it look like the ISP’s promoting this content. You know, whether it then gets blocked is a separate Section 36 question that I was just discussing.

1123 COMMISSIONER VENNARD: Yeah.

1124 MR. TACIT: But here we just want to make it clear that something may have fit into a category yesterday but if it’s illegal the ISP through its zero rating -- putting it on a zero rating list doesn’t want to look like it’s promoting whatever that is.

1125 COMMISSIONER VENNARD: Yeah.

1126 MR. TACIT: Then the question of blocking is another separate issue that would have to be addressed.

1127 COMMISSIONER VENNARD: Yeah. But I hope you understand that where I’m -- I’m looking at this from the point of view of how to develop some principles on this and from a regulatory perspective what would this look like. Because sometimes it’s -- sometimes, you know, we come up with ideas but when the rubber hits the road it’s a lot more complicated then what we ---

1128 MR. TACIT: Yeah, understood. And maybe, you know, it would be helpful too if the Commission could deal with the broader Section 36 question and the lawfulness. That might actually give the answer for the purpose of this regime as well, because it would be implicit here. But since we don’t have that ruling that’s why we thought it was important to at least try to deal with that piece here.

1129 COMMISSIONER VENNARD: So does CNOC actually look at the internet as being a common carrier as such, the network?

1130 MR. TACIT: I think ISP’s generally do look at it that way that they are -- they’re telecommunications service providers. Some are common carriers. Some would not meet the technical definition of telecommunications common carriers but they’re still telecommunication service providers and as such are agnostic to the content typically.

1131 COMMISSIONER VENNARD: To the content. Yeah, I was a little surprised to see that in there, but that’s okay.

1132 Let’s get back to your regime that you have suggested. So we have the categories and there’s a lot of thinking that would have to happen around those categories and categories in terms of the definition, what’s in, what’s out, who’s in, who’s out, do you have any ideas that you’d like to share with us with respect to inclusion or exclusion, how that would actually work?

1133 MR. STEIN: Sure. From an exclusion standpoint, we’re suggesting that if it meets the definition of the category it’s included. But there are some practical situations, such as the application being developed that day and so forth, that that may not be caught but that when an ISP sets up a category that they would take every reasonable measure to actually make sure that they’ve covered off everything that should fit within that category and that they would not be able to exclude or prefer their own and so forth.

1134 COMMISSIONER VENNARD: Okay. Let’s talk a little bit about that one. You’re saying that your regime does not allow for any financial benefits. As I was reading through this and listening to you, I started to wonder about the money part of this, because if you follow the money then you -- money has to come into it somehow. Who pays for this and how do they pay for it? Is this just out of the plans or -- how does -- put money in the picture for me.

1135 MR. STEIN: Well, keep in mind, when we talk about zero rating we’re not saying it’s free, it’s rated as zero, meaning it’s not included in the cap but it is probably included in the base fee. So if today you’re used to paying for home internet for however many dollars per month and you have a cap of a certain size it’s that first -- that monthly fee that is covering the cost of whatever else is getting zero rated. So that’s where the payment is.

1136 COMMISSIONER VENNARD: So do you see an increase in the internet bills or communication bills? I mean, this would -- it seems to me this would be pretty expensive.

1137 MR. STEIN: Well, I -- if I may, I think every category will work differently and every ISP is going to look at how they package their internet services differently and so they’re going to have to make decisions. I’m not so sure it would be very expensive to do, depending on which categories you choose. Some categories could be quite costly, some categories not.

1138 MR. TACIT: So we just had the -- we heard the previous witness for Sandvine testify that usage of the internet increases usage of the internet.

1139 COMMISSIONER VENNARD: M’hm.

1140 MR. TACIT: I think the point of zero rated schemes isn’t to create an immediate pot of money from that content. The idea is to create a market differentiator for ISP’s that will attract additional consumers and that will attract additional usage from existing consumers, and that’s where the economic benefit is. But there’s no requirement for anyone to pay anything, and, in fact, under our proposal we wouldn’t want content providers to be able to pay to be listed because that can distort the market and create all sorts of competitive issues.

1141 So this goes to why would an ISP do this. That’s fundamentally the question.

1142 COMMISSIONER VENNARD: Exactly.

1143 MR. TACIT: It does it to grow the market and to grow usage generally but not to increase prices for existing users nor to charge content providers for it.

1144 COMMISSIONER VENNARD: But if you take it to -- a little bit further and consider what could end up happening, there could end up being a lot more network capacity or spectrum capacity required, and in your submission you say that doesn’t come for free, so money does actually have to enter into it at some point.

1145 MR. HICKEY: Well, I think the concepts there are a little bit distinct. So right now when we’re talking about differential pricing practices, that we look at as distinct from data caps and usage based billing.

1146 So when you’re talking about increased price due to the use of a differential pricing program, when I look at our membership pretty much all of them have unlimited pricing today in the markets. So I don’t know how implementing a zero rating could increase the rates for those users when they already have access to unlimited services.

1147 COMMISSIONER VENNARD: M’hm.

1148 MR. HICKEY: When we’re talking about usage based billing with respect to internet traffic management practice, there is a intersection between how much capacity you can buy, what rates you can charge to consumers to ensure that they were willing to respond to your offer, and then ensuring that the price capacity and usage amounts make the service that you offer fair to your end users. I think that's a different issue than price following any type of differential pricing practice.

1149 MR. TACIT: I think what we're really saying here is we're -- ISPs are going to do this if it makes sense for them. If the costs end up to be too big, they're not going to have DPP programs.

1150 COMMISSIONER VENNARD: Okay.

1151 MR. TACIT: If it's going to cause network congestion that's going to make the experience of their customers so poor that they lose customers, they're not going to do that. So it's a market-driven solution. The idea is to increase choice in the marketplace. It will actually increase usage. It will increase usage of the internet and it will bring new ISPs to the market that can offer new value propositions. That's the point of it.

1152 But of course, all of these entities hopefully will make it. Some may not. Some may experiment and fail, and that's the nature of the marketplace. We're not -- nobody's seeking a guarantee for any individual ISP. We're just saying, "Let the ISP community do its best to serve its end users and try to innovate in these ways as best as they can," and then do that within a framework that provides the protections, the content related, the competition related, and social related protections that the Commission would expect in that framework.

1153 COMMISSIONER VENNARD: Okay. Yeah, I get it. It's just, in your submission, you say that the capacity isn't free to the ISPs, so that's the reason I brought it up. That was written in -- actually, in your submission.

1154 So let's go back to your regime. So there's no financial benefit to the vertical integration part of things; it’s not something that fits into your proposal?

1155 MR. TACIT: Well, the way we handle vertical integration is to say a VI entity is free to add its own content to a bundle as long as that is a real -- or is -- or to that program as long as that -- the basket of content is broad enough so that it isn't the only one that's benefiting or primarily benefiting from it. So if there is other like-for-like content or other substitutable content in that basket in that category that's subject to that DPP, we wouldn't have any objection to the VI entity also including its own content.

1156 So in that way, it's actually kind of flexible because normally, you know, there are a lot of constraints on VI entities and for good reason.

1157 COMMISSIONER VENNARD: M'hm.

1158 MR. TACIT: But in this case, this framework has a safeguard that actually allows them to have that benefit as long as it's done in a way that doesn’t distort competition and give them an undue preference.

1159 COMMISSIONER VENNARD: So you're saying that they can offer their own and -- but they must offer others as well?

1160 MR. TACIT: In that category.

1161 COMMISSIONER VENNARD: Okay. Just hang on.

1162 So can you comment a little bit about the benefits and the potential harms, the balance that needs to be struck in your particular proposal?

1163 MR. HICKEY: Yeah, so I think with respect to the benefits, we certainly see that there will be opportunities for innovation as in certain cases making more access more affordable. And you know, if you're looking at -- a customer may want a capped package but would respond to an offer that provides them with greater access to certain services or some, you know, innovative offerings that we're not even thinking of at the moment but that are likely coming -- to come down the road.

1164 But I think when we're looking at concerns, one of the largest that we have is obviously gate keeping within a category. That's what the largest piece of our proposal with respect to zero rating is trying to address, that I can say, "Here's one or two zero rates. Everyone else that's with content and application providers are not permitted within that category." So we're certainly trying to address that piece.

1165 And then with respect to the no financial benefits, we're more talking on that front as sponsored applications. So some kind of revenue benefit for me to push an application to my end user that maybe my end user will only respond to because I'm getting revenue to promote that, even -- but it's not on the merits of that application. So there may be better ones out there that don’t get used just because that new entrant doesn’t have the financial resources to afford that type of program.

1166 MR. TACIT: The principle we're operating under is a rising tide floats all boats. And what I mean by that is, if you have a well-constructed, lawful category of services, there may be a bunch of services in there that people are already familiar with, but there will be some new ones and they could be very innovative things that would might not have otherwise gotten market exposure.

1167 So those things suddenly are going to be accessible to the public now in a way that they haven't been before because, hey, people can try them for free as part of a -- you know, as part of some other package in terms of transmission. There may be charges for the actual app or whatever, but from a transmission perspective, they don’t have to pay for the data.

1168 So it actually provides an opportunity, not just to give consumers what they're already familiar with, but also expose them to new content as well ---

1169 COMMISSIONER VENNARD: And it ---

1170 MR. TACIT: --- which is very consistent with either requirements when you look at the undue preference rules to take into account the Broadcasting Act objectives and the diversity requirements of subsection 3(i) of the Broadcasting Act.

1171 COMMISSIONER VENNARD: And so in that sense, there's -- your proposed regime would actually encourage innovation?

1172 MR. TACIT: Exactly so.

1173 COMMISSIONER VENNARD: Including from edge providers?

1174 MR. TACIT: Exactly so.

1175 COMMISSIONER VENNARD: Okay. What about the -- are there other benefits that you would like us to have on record that you perceive or that you ---

1176 MR. TACIT: Well, as I say, it's helpful overall to the increase, not just to exposing new content, but I think it's good for the ISP market as well, because it is likely to make the competition among retail ISPs more vigorous and innovative in a way that, you know, they're not limited to just, you know, who's got the lowest price, necessarily, but they can get more creative about that. And that will -- that's good for consumers. It attracts more players into the market, which also is good for consumers. It's more choice.

1177 COMMISSIONER VENNARD: Okay. You mention also that DPPs have the potential to result in undesirable and dangerous outcomes, and then you go on to list four of them, of which we've talked to some of them. Can you explain to us how your proposed regime would mitigate against some of these? Maybe just briefly on all four of them, for the record.

1178 MR. STEIN: Sure, I'd be happy to. So in paragraph 6, we talk about four issues. The first one, just to reiterate quickly, the anti-competitive aspect; this was -- and what we're addressing here, this is a case where an ISP might have said, "I also have this radio service or this streaming music service. I'm going to offer my streaming music or I'm going to offer my streaming music and maybe one other."

1179 In our proposal, we're saying that if they're going to offer streaming music they have to offer all of that streaming music being as they're -- being the same zero-rated DPP.

1180 The second is a variant of that, ISPs engaging in gatekeeping and therefore perhaps offering their own or their favourite music, again, using the music service. Here are the two music services we like for whatever reason; maybe for no reason at all but because we chose them. They were popular; we were familiar with them. And in doing that gatekeeping, it -- and by zero rating, the problem is that it takes all the other music services and puts them on an uneven playing field.

1181 The third, content providers being harmed if they're excluded, so again, that being the content providers that were excluded from that package.

1182 And then lastly, market distortions due to end users having an incentive to choose. This is again offering that free service or the ISP choosing which one to give free and therefore sort of incenting a user to try that one just because it's free as opposed to because it's -- it has become popular for good reasons: great features, great functionality, great quality, and so forth.

1183 COMMISSIONER VENNARD: Okay. I'd like to talk a little bit about transparency on -- in that sense. Now, how do you see that working within your proposed regime?

1184 MR. STEIN: Each ISP when they create the category and make it available would have to also make available a tool or a method, so that all the customers could really see exactly which services are included in that category so that there's no question. The examples that were used this morning of social media listed a lot of social media, but there should be an ability for the user to go check, “Is the social media that I want in this?” And it shouldn’t be subject to sort of, you know, “All social media? Sorry, not that one.” We’re not here to play gotcha kind of thing.

1185 So there’d have to a method on the website similar to how in the ITMP framework it has to be published and explained. Similarly, there’d have to be on the ISP’s website a space where you could go and look and understand exactly what's included in my zero-rated package.

1186 COMMISSIONER VENNARD: Okay. And my final question has to do with exemptions and restrictions. According to your intervention, you don’t think that applications -- even such as those associated with social needs -- should be treated differently. Would you ---

1187 MR. TACIT: So I think what we ---

1188 COMMISSIONER VENNARD: --- like to comment on that?

1189 MR. TACIT: --- said if I remember is that sometimes it might be difficult to determine what is social and what isn't. And I think we gave the example of YouTube where, you know, it could be perhaps used in an educational way, but it could also be used for cat videos. So would you want to classify YouTube as a social necessity? Probably not.

1190 However, things like video relay service or things of that nature that only have a social purpose, we certainly wouldn’t object to having zero rated if it's clearly for that purpose. It's where there's potential multiple uses we -- I guess we didn’t want to start attracting a bunch of regulation to provide zero-rated services for marginal cases that aren’t really -- that’s not the core focus. And I think that’s what we were trying to say, not that we object to providing zero rated to truly social services.

1191 COMMISSIONER VENNARD: Okay, thank you.

1192 Is there anything that -- any final sort of comments on your proposed regime that you’d like to include in the record?

1193 MR. TACIT: Well, as I said, I think we tried to strike a balance, and really our -- a lot of the focus of our proposal is on making the market work as much as possible within the require -- the framework of subsection 27(2) and 36, maximizing transparency which is very important, making it clear what's in there for both potential parties who might want to add content to a category, and for end users who might want to decide whether to sign on to this.

1194 We, you know, we tried to make sure that we were fair to everyone, so even VI entities can participate if they do it in an appropriate fashion. So it is a very inclusionary scheme, it is a very market-driven scheme which we think is a good thing. We think it's consistent with both the policy objectives and the policy direction. So, I guess I would leave you with that thought.

1195 And as I mentioned earlier, it provides opportunities for innovation and discoverability of new content and applications that might otherwise find a harder time becoming visible in the marketplace.

1196 COMMISSIONER VENNARD: Okay, thank you.

1197 Those are my questions.

1198 LE PRÉSIDENT: Monsieur le conseiller Dupras, s'il vous plaît.

1199 COMMISSIONER DUPRAS: Yes, good afternoon.

1200 Wireline services, I mean they have higher caps and they offer unlimited options. I mean aren’t DPP of lesser value on wireline than on wireless?

1201 MR. TACIT: Well, isn't that something for the consumers to decide though? I guess that’s the answer. It’s, you know, it’s new for our members. The reality is they're going to be experiments, some will succeed, some will fail. The ones that succeed will benefit consumers and that's the point. If it can work in a way that is -- that meets the regulatory framework requirements, it seems to us that that’s a question the marketplace should answer.

1202 COMMISSIONER DUPRAS: M'hm. And do you think it could be an incentive for wireline to reduce data caps or even the -- get rid of the unlimited option eventually if DPPs are a profitable idea?

1203 MR. STEIN: I don’t see it that way. I don’t think that that would happen. I can't see that.

1204 COMMISSIONER DUPRAS: Okay.

1205 MR. HICKEY: I guess -- if I could add, I think that, you know, during the time when usage-based billing -- it was our members that continued to offer when the only other, you know, usage-based billing was all the rage, it was our members in 2009-2008 that continued to offer unlimited service offerings, that’s kind of been a core. I don’t see that just disappearing because there's an extra way to price your -- to price your service.

1206 MR. TACIT: I know there's been a lot of concern expressed about data caps and we understand that. But I think what's important for people to realize is that that is just one option in the marketplace. It’s not like the whole marketplace is structured only with data caps. Even among CNOC members, the vast majority of them -- and I mean vast majority of them offer both unlimited and capped services, so people have a choice.

1207 What -- what a capped service does, is it actually provides a lower cost choice for that segment that wants to pick it. It doesn’t force them there. They can also go to an unlimited plan if they want. And there are all sorts of other tiers based on usage; it’s not just one tier. It’s not one speed and it’s not all just data caps.

1208 So I think one of the things that the Commission’s been doing that’s actually been very helpful, is creating that website that allows people to find out who the service providers in their area are. I think when people start using that a little bit more; they're going to see the choices they have. And that if they click on a CENOC member link, they're going to see not just the capped services, but the unlimited services that are available as well.

1209 COMMISSIONER DUPRAS: So you see more DPPs for wireline maybe for your entry level services with lower data caps to which you could add some free additional zero-rated content possibly?

1210 MR. STEIN: That’s absolutely possible. DPPs could come in many forms, whether they're targeting a very starter package or a very high-end package. That's really what we’re talking about as differentiation, it's one more tool to differentiate packages.

1211 COMMISSIONER DUPRAS: Okay. Thank you very much.

1212 THE CHAIRPERSON: So, I understand your point that there’s different data offerings in the system, but doesn’t a lower data cap just create a price point for a higher offering? It’s a bit an artificial construct; isn't it?

1213 MR. STEIN: I’m not sure if I -- I’m not sure if I would characterize it as an artificial construct. Many customers do seek that low price point, and use it and enjoy it and stick with it for a very long time. But also, some customers start there and move their way up to something either faster or with more capacity, but that's just more options.

1214 THE CHAIRPERSON: Your categories -- your broad categories, I take it they're loosely based on Sandvine’s categories? Are there any market differences between the two?

1215 MR. HICKEY: We looked at two reports, there was the Sandvine and the Cisco instead of by era. We looked at those reports and we took -- we did our best to kind of package those. And amongst discussion to ensure that what we were proposing made sense from a category and actual usage in place today, because we were trying to say, “If you were defining today, here’s the types of things you could set categories based on.”

1216 THE CHAIRPERSON: So you're not seeing -- you don’t see a major difference between that? You didn’t do a choice and say, “Well, the Sandvine was too much or too little of something.” You essentially adopted theirs and Cisco’s or were -- are there important differences?

1217 MR. HICKEY: I think those we looked at as -- if you're -- we didn’t go to all end users and ask them to pull a whole bunch of data about what's those -- our members about end user data. We thought the best way was to say, “Well, we’re...” ---

1218 THE CHAIRPERSON: I'm just talking about the categories, the broad categories.

1219 MR. STEIN: If I may, these were just examples. We took ---

1220 THE CHAIRPERSON: Okay.

1221 MR. STEIN: You're right. We looked at those examples. We had some conversations and said, “These are some good examples.” But I just want to clarify, we aren’t suggesting that these are the right categories or that these are best practice categories or that these even ought to be the categories. We’re proposing that each ISP has to choose their own categories. These were just the examples that we raised.

1222 THE CHAIRPERSON: Right. Although you're also advocating for regulatory certainty, so ---

1223 MR. STEIN: That’s right.

1224 THE CHAIRPERSON: --- I'm not sure if that provides you a lot of certainty ---

1225 MR. STEIN: Well ---

1226 THE CHAIRPERSON: --- when everybody can decide their own.

1227 MR. STEIN: Well, everybody can decide their own provided that they still adhere to the same principles that we’ve laid out and we hope would be enshrined in that regulatory certainty, that being that everything in the category, that all the substitutions are all included in that category, the example that we used earlier about music, or you could use an example of video or any of these things, and that people don’t promote their own and receive a specific value or really game the category to achieve ideal value for themselves, that it is broad.

1228 THE CHAIRPERSON: What would you say to somebody that says that the proper category ought to be all entertainment streaming services and that it’s artificial to make the distinction between music and video?

1229 MR. TACIT: Well, there’s got to be some flexibility for ISP’s to make marketplace decisions. That’s the point of this, is the only way that a DPP framework makes real sense is if ISP’s do have an app ability ---

1230 THE CHAIRPERSON: There’s a certain arbitrariness making a distinction between music and non-music streaming sources.

1231 MR. TACIT: And maybe some ISP will want to have a category that has all of it but maybe another one won’t, and that’s the point. Maybe one will only want to do audio ---

1232 THE CHAIRPERSON: We have to put it through its 27(2) lens. I’m trying to get your views as to why cutting it off at music is appropriate for Section 27(2).

1233 MR. TACIT: Because if there is a big enough sample of those, such that there isn’t necessarily any one service that’s getting unduly preferred, if they’re substitutable, if in total they represent a significant ---

1234 THE CHAIRPERSON: You see, that’s the point, substitutable -- I mean, ---

1235 MR. TACIT: What you ---

1236 THE CHAIRPERSON: --- entertainment is entertainment is entertainment, isn’t it?

1237 MR. TACIT: Well, I don’t know about that. I mean, I think there are a lot of broadcasters who would say that there are different classes of entertainment, different types of entertainment, different categories and so on, and they -- you know, we have packages priced based on -- you know, in the distribution world based on the perceived value to the consumer of the content of those packages. The idea here is to let the consumer make those decisions as much as possible while respecting the framework.

1238 THE CHAIRPERSON: Okay. Well then let’s build on that. Why make a distinction between streaming services and radio services that are high in music content and let’s say a podcast, which isn’t a streaming service but a subscriber type service, which may actually contain a lot of music?

1239 I mean, we know from our CMR that -- it actually strikes me every time I see it -- in how younger generations are consuming a great deal of podcasts. So that’s their choice as well.

1240 MR. TACIT: You’re right. And an innovative ISP may have just gotten the idea from you today to have a podcast DPP, but -- and somebody else may include a podcast ---

1241 THE CHAIRPERSON: I put it to you though that the ISP’s will rather -- in creating that, rather piggyback on named brands that are known, like Spotify and others. I mean, they won’t create -- that’s what -- they’re trying to harvest, I put it to you, the marketing of these popular music streaming services to attach themselves to it.

1242 MR. TACIT: I’m sorry, Mr. Chair, but I really disagree with that characterization. It comes back to what I said earlier, which is that a rising tide floats all boats.

1243 Sure, the fact that you may have some leading well-known applications or content in a package may actually help sell that zero rated category to the public, but if you accept our premise that you can’t turn away other ones that are lesser known, then those will benefit and they might get the discoverability that they would have otherwise not had absent being included in that package.

1244 THE CHAIRPERSON: I get back to your point about ex post and ex ante and the need for regulatory certainty. I’d like to quote -- and it’s a bit long but bear with me -- paragraph 27 of the Indian regulator’s analysis. And I realize that particular decision inflows, but I think it raises some interesting points. They write:

1245 “Intuitively the case-by-case approach may seem reasonable. However, this approach creates substantial social costs.”

1246 And then they cite a study by Barbara van -- I’m not going to pronounce it right -- Schewick I guess -- maybe not.

1247 “First, a case-by-case regime will fail to provide much needed certainty to industry participants. In the absence of a clear setting out of the permissible and impermissible business practices, service providers may refrain from deploying network technology. This will be due to the fear that their conduct may subsequently be construed as being discriminatory as per the case-by- case analysis. Second, it will create high cost of regulation on account of the time and resources that will be required for investigating each case. It will also lead to further uncertainty as service providers undergoing the investigation would logically try to differentiate their case from earlier precedents. Third, there is also the concern that this approach provides a relative advantage to well-financed actors and will tilt the playing field against those who do not have the resources to pursue regulatory or legal actions.”

1248 And, in fact, we heard testimony that smaller players don’t come to our hearings because of that barrier.

1249 “This may include end users, low cost innovators, start-ups, non-profit organizations, et cetera.”

1250 And then the authority -- it goes on to say that this was a significant concern.

1251 Regardless of where they go with that, it seems to me it points out some problem with the ex post ---

1252 MR. TACIT: Well, if that were true then shouldn’t the ITMP framework be revisited because it is also an ex post framework. There are situations when an ex post framework makes sense, and we think that this is one of them and we think that it would be not just a marketplace innovation but also a regulatory innovation to give that a try and see how it works.

1253 THE CHAIRPERSON: I would have thought an organization like CNOC would have been more concerned about regulatory gaming that occurs in Part I applications on an ex post basis.

1254 MR. TACIT: Well, look, we’ve been there and, you know, we have issues with Part I applications and so on and then the need to bring those, but the reality here is we’re being very transparent about what we’re doing. We think that ISP’s generally will understand the principles that the Commission articulates for how to conduct themselves if you adopt the principles that we’ve suggested, and that based on that ---

1255 THE CHAIRPERSON: See, the point is it’s not just an ISP issue.

1256 MR. TACIT: Right.

1257 THE CHAIRPERSON: It’s about Canadians.

1258 MR. TACIT: Understood. But this ---

1259 THE CHAIRPERSON: How will they have the resources to bring these cases forward?

1260 MR. TACIT: Well, they do have voices today. It’s not like -- the same Canadians, you know, can complain about ITMP’s today if they need to. Who complains about ITMP’s if not Canadians? It is an analogous situation.

1261 THE CHAIRPERSON: We heard testimony earlier with the video game industry and Rogers’ behaviour how long that took to get it back on track. The Commission doesn’t have limited resources.

1262 MR. TACIT: Understood. And it may be that there may be one or two cases -- you know, one or two test cases that come up that further define how this works. That’s the nature of our jurisprudential framework. You know, the common law framework and so on where through subsequent decisions the conduct becomes clearer, but that doesn’t mean don’t give it a try at all because there might be a few tough cases along the way.

1263 I mean, again, I go back to the ITMP framework, the Commission found a way to do it there and it’s worked very well. Have there been some complaints? Yes. Have there been a huge number? No. Is the marketplace working reasonably well on the ITMP side? I would say yes, from what I’ve seen. We think this is worth a try for the benefit of consumers to give them access to new content that they might not otherwise see and to increase the number of ISP’s in the marketplace which again will benefit consumers through increased choice and more ---

1264 THE CHAIRPERSON: Oddly, it would seem in this hearing that the vast majority of consumers are concerned about DPP’s.

1265 MR. TACIT: And I understand the concern, but, you know, that’s why ---

1266 THE CHAIRPERSON: So you’re saying we should dismiss those consumers that have appeared before us in favour of some theoretical consumer that you think you’re serving?

1267 MR. TACIT: Well, you know -- no, we’re not saying that. We’re saying accept those. But there are a whole bunch of other consumers -- and I recall, for example, the Bell surveys in their evidence where they said that there’s a high number of consumers that want to see zero rating as well.

1268 THE CHAIRPERSON: There are always ways of phrasing questions and surveys to get the answer you want.

1269 MR. TACIT: But that cuts both ways. That cuts both ways.

1270 THE CHAIRPERSON: Except that the Commission’s approach in this proceeding is not a survey where you ask perhaps a leading question. We’ve actually asked Canadians to come to our hearing, not a survey ---

1271 MR. TACIT: No.

1272 THE CHAIRPERSON: --- and they’ve expressed a view.

1273 MR. TACIT: And I understand that. And you’ve asked Reddit users, which are a particular subset of users -- I don’t think they represent necessarily all Canadian users. So ---

1274 THE CHAIRPERSON: But they do represent a ---

1275 MR. TACIT: They ---

1276 THE CHAIRPERSON: --- significant portion of people that use the internet.

1277 MR. TACIT: They do represent a segment that uses the internet, but there are an awful lot more people, including me, who haven’t signed up for Reddit. So and that doesn’t make me an unknowledgeable Internet user.

1278 THE CHAIRPERSON: With respect to Videotron’s unlimited music and based on your proposed criteria, are you of the view that it would be -- using your filter, your proposed filter, would Videotron’s unlimited music be acceptable?

1279 MR. TACIT: I guess it depends on how transparent they are about the rules for joining and how good a job they’ve done in allowing others to join that group, and not just gone out and sought out ---

1280 THE CHAIRPERSON: Well, there’s been a significant record on that asking them what the rules are and so forth so we have a very fulsome record. Based on all that, do you think it’s acceptable ---

1281 MR. TACIT: If they complied with the way that we suggest in terms of their behaviour then, yes, I think their category would pass the test.

1282 THE CHAIRPERSON: But we’ve asked a series of questions.

1283 MR. TACIT: And I understand that. It’s not about the questions, it’s about ---

1284 THE CHAIRPERSON: Do you think ---

1285 MR. TACIT: --- their conduct going forward. It’s about do they disclose their terms under which someone can sign up on their website. Do they ---

1286 THE CHAIRPERSON: You see, the point is here we are.

1287 MR. TACIT: --- explain it clearly to end users?

1288 THE CHAIRPERSON: Here we are in a Part 1 after the fact, and you’re still saying you’re not sure whether they’re in ---

1289 MR. TACIT: No, I’m saying that it ---

1290 THE CHAIRPERSON: Hence the difficulty of an ex-post factor approach.

1291 MR. TACIT: I think what I’m saying is, to the extent that their proposal needs to be tweaked to comply with our ---

1292 THE CHAIRPERSON: How so?

1293 MR. TACIT: --- proposal.

1294 THE CHAIRPERSON: How so?

1295 MR. TACIT: Like I said, make sure they post the terms of how you join as a content provider clearly on their website. And, you know, what they have to do and who qualifies; make that clearly posted. With that sort of tweaking, their category is fine. We would have no problem with it.

1296 THE CHAIRPERSON: That’s your only caveat?

1297 MR. HICKEY: Well, I think that when they introduced I think that we would like to have seen obviously a wider scope of included providers. Our position is that provided you are treating all service providers and content providers within a category equally that raises less concerns.

1298 Ideally, we would like to see a broader and wider scope of providers that are in the service to begin with. And then if you’ve missed someone, provided you’ve done your best faith and they’ve come and asked, you can include them in your zero-rated service.

1299 I would say that picking those -- a smaller selection of the most popular would likely raise a concern, in my view, under Videotron’s practice. If they had come out with a much wider scope, I think that would have addressed that concern based under our proposal.

1300 THE CHAIRPERSON: Do I take it though that based on their behaviour since then you think that they’re now in compliance?

1301 MR. HICKEY: I would say based on the evidence we have, yes.

1302 THE CHAIRPERSON: Okay. Thank you very much.

1303 Those, I believe, are all our questions. And thank you very much, gentlemen.

1304 MR. HICKEY: Thank you.

1305 THE CHAIRPERSON: And we’ll adjourn until nine o’clock tomorrow morning. Thank you.

--- Upon adjourning at 4:19 p.m.

REPORTERS

Sean Prouse

Mathieu Bastien-Marcil

Nadia Rainville

Lyne Charbonneau

Marie Rainville

Patricia Cantle

Lise Baril

Jacqueline Clark

Janice Gingras

Suzanne Jobb

Mathieu Philippe

Date modified:

2016-10-31