Intervention: Intervenor 230

Document Name: 2015-134.223931.2393444.Intervention(1f@sk01!).pdf

Layton/Aalborg University CRTC Consultation 2015-134 1 July 13, 2015

Mr. John Traversy
Secretary General
Canadian Radio-television and Telecommunications Commission
1 Promenade du Portage
Les Terrassess de la Chaudière, Central Building
Gatineau, Québec *** ***

Reference: Telecom Notice of Consultation CRTC 2015-134, Review of basic telecommunications services **** Mr. Traversy,

Thank you for the opportunity to comment in this review. My key points are:

 Focusing on the supply of broadband networks will have little to no impact to bring unconnected Canadians online. There are many reasons why individuals do not adopt broadband services.

 The CRTC should focus on the actual technologies and services that users adopt, not on network speeds.

 Most of the services that clearly belong in the bundle of basic commitments – affordable access to a reliable communications platform that provides access to emergency services, essential government services and information, employment applications, and even basic e-commerce – do not require a class of service sufficient to support high quality streaming video.

 A modest basic telecommunications policy that allows people to participate in the digital economy can achieve more than a policy promising to deliver the fastest speeds to everyone. How to provision communications networks to support important social commitments requires a more critical view than has been offered to date.

Please feel free to contact me with any questions. I request to appear at the hearing, albeit remotely, if possible.

Roslyn Layton
Ph.D. Fellow

Center for Communication, Media and Information Technologies, Aalborg University Frederikskaj 12, 3rd Floor

Copenhagen, Denmark 2450

Introduction ..............................................................................................................................................................2 Critical examination of premises is required ............................................................................................................5 Faulty Premise: Essential applications and services need high speeds ...............................................................5 Faulty Premise: People in rural and remote areas have limited participation in the digital economy because they have only one telecom service provider (TSP) .............................................................................................9 Faulty Premise: Wireless technologies cannot compete with wireline ............................................................ 10 Faulty Premise: Regulators should pursue policies of 100% deployment and adoption .................................. 11 Faulty Premise: Canada will fall behind if it does not implement a particular policy ....................................... 14 Conclusion ............................................................................................................................................................. 15 Introduction

1. Determining basic service commitments in the age of convergence is no easy task. The traditional tool kit of telecom regulation includes licensing and authorizing operators, managing spectrum, and providing oversight to interconnection. That expertise does not necessarily map to a world where content, connectivity, and communication are merged and interwoven with a complex set of technologies and user preferences.

2. As such, it is wise that the Canadian Radio-television and Telecommunications Commission (CRTC) ask the set of questions posed in the Consultation, though it should come as no surprise that the answers are not simple and straightforward. As an initial matter, basic telecommunications service has historically been leveraged to support various important social commitments. Ensuring that everyone has access to some basic communications platform, so that they can contact emergency services and avail themselves of other important government and social services is an important social value.

3. The challenging questions are at what level and by what means are those commitments maintained. Some advocate that every person needs access to high speed Internet at low cost. While there is a strong argument that the CRTC should endeavor to provide every Canadian with access to some level of connectivity, it is unclear what that level of connectivity should be. Indeed, as countries have transitioned from narrowband voice communications to broadband Internet connectivity, the advocates and policy makers have consistently increased their target for connectivity.

4. Importantly, these changes in the definition of basic services have tracked changes in median (or even high-end) usage patterns, as opposed to tracking what is sufficient to provision socially necessary services.

Historically, the difficulty of determining what services belong in this set has been masked by the nature of telephone technology. The basic unit of connection – the twisted pair of copper wires – that was necessary for any service was also sufficient for most services of interest to most consumers. As a result, by requiring universal provision of the most basic services, we also facilitated the provision of more advanced services.

5. This no longer holds in today’s digital economy. One can get connected to the Internet through various means:

fiber, coaxial cable, fixed and mobile wireless data, satellite, and even still, the good old twisted pairs of copper. Each of these means of connecting to the network offers better or worse support for various services and applications. Fiber is very fast but relatively more expensive than other technologies; cable and (especially) DSL are somewhat slower, but are also somewhat cheaper; wireless is generally a bit slower still (at least as of today), a bit less reliable and often somewhat more expensive than cable – but it’s mobile. Some of these technologies are better for voice service, for video service, for downloading large amounts of data, or for playing video games.

6. Developments in the many technologies suggest that we need to take a more critical view of how to provision communications networks to support important social commitments. The historical precedent, that we would provision a connection capable of supporting nearly the full range of possible services, was a happy historical accident. It was possible in part because the basic unit of service was capable of supporting the full range of consumer-oriented communications services. And it was possible in part because the relative elasticities of demand for communications services offered a relatively efficient mechanism for funding universal service buildout.1

7. The most difficult aspect of this more critical view is that we need to think seriously about what services are included in the bundle of basic social commitments. Some advocates would argue that every Canadian should have access to low-cost Internet service capable of supporting streaming video services. That is quite an upgrade from the basic services historically provided through universal service – basic local voice communications service (long distance was available, but at substantial cost).

8. The justification for promoting this class of Internet service as “basic” is on the grounds that such high-speed service is needed to ensure access to, for example, educational, health care, and governmental services.

However, the reality is that most (and possibly all) of the services that clearly belong in the bundle of basic commitments – affordable access to a reliable communications platform that provides access to emergency services, essential government services and information, employment applications, and even basic e-commerce – do not require a class of service sufficient to support high quality streaming video.

9. But just as communications technologies and the services that they facilitate are diverse, so too are consumer preferences. It is absolutely the case that there are basic services to which we should do our best to ensure that everyone has reasonable access. But today we need to think more carefully about what these services are.

Most important, we should resist the urge to assume that every consumer fits one ideal type.


This is because universal service has traditionally been supported by a cross-subsidy from relatively inelastic-demand services (such as business-oriented calling plans) to relatively elastic-demand services, like local calling. This is an example of Ramsey pricing. See F.P. Ramsey, A Contribution to the Theory of Taxation, 37 Econ. J. 47, 58-59 (1927).

10. Along these lines, the meaning of “basic telecommunications service” is long past a need for review. Returning to the earlier discussion of how the basic unit of transmission has changed – from a unit capable of supporting the full range of telecommunications services to a range of units capable of supporting a range of services – the central question that “basic telecommunications” (also called “universal service”) faces is what services need to be universal. There is a strong argument, for instance, that the basic service universally available should be sufficient to support access to basic news and information, health, educational, and governmental services.

11. There may be some argument that such a connection should be capable of supporting basic online video services. But there is only a much weaker argument that high-definition, or even 4K, online video needs to be universally available. Put in today’s terms, there is a world of difference between a basic service requirement that a person could watch a safety video on YouTube versus streaming Netflix’s “Marco Polo” in 4K high definition at 25 Mpbs, which also comes at an additional price from Netflix.2 The latter can hardly be termed as essential. Netflix in Standard Definition (SD) requires only 3 Mbps.3 Moreover video over cable or satellite has never been part of a universal service commitment or been part of a subsidy regime – it is not logical or justifiable that video because it appears on Internet is suddenly essential, when it was not considered as such for decades when it appeared on cable or satellite.

12. It may make sense, for instance, to reframe basic telecommunications service goals to focus on enabling certain classes of applications. Rather than define universal service as generic high-speed Internet, universal service could be defined as service sufficient to support a minimum bundle of services. That bundle may include, for instance, healthcare, education, employment, government services, news, and VOIP.

13. To think about basic service intelligently needs to take into account not just networks, but user preferences, the level of digital skills, devices, applications, and services.4 Regulators should think more broadly about defining what digital literacy means for Canadians and how specifically Canadians are to participate meaningfully. If people lack basic literacy, employable skills, and economic opportunity, then provision of basic services will not necessarily render the outcome that regulators and advocates want. There is undoubtedly a political desire to deliver economic growth through the “silver bullet” of broadband, but the role of broadband in developed countries is not entirely understood or predictive. The OECD reports only mild correlation between high speed broadband and gross domestic product (GDP).5 As will be discussed, not only does the provision of broadband not guarrantee adoption, the GDP-enhancing benefits to society do not appear to be 2

**** Newton, “How Much Is Netflix 4K Ultra HD and What Broadband **** Do You Need?,” Recombu, (November 18, 2014),


“Internet Connection **** Recommendations,” Netflix, accessed July 6, 2015,


Knud **** Skouby and Morten Falch, “Increasing Role of Public Private Partnerships in the ICT Ecosystem,” 2014,


“Fixed (wired) Broadband Penetration and GDP per Capita” (OECD, 2014),

realized until users reach at least the third and mainly the fourth and fifth levels of Information Communications Techology (ICT) usage.

Critical examination of premises is required

14. Many arguments for regulatory intervention in telecommunications markets rest upon untested assumptions.

These are often ideas that make superficial or intuitive sense – and have great political potency – but do not necessarily stand up to critical analysis. Sometimes these premises seem so obvious that we do not bother to test them. Debates over telecom policy are necessary to the well-being and prosperity of any country. Sound telecom policy can benefit users tremendously while bad ideas can be terribly costly. At its best, telecom policy can help lift the poorest and least fortunate among us to an improved quality of life, afford unparalleled access to education, health, and other essential services, and create platforms for expression and enterprise. Few, if any, other technologies or industries have the potential to create so much good for so many.

15. It is not surprising that these assumptions tap into deep currents in the popular psyche. The questions at issue in telecommunications policy reflect values at the core of democracy, social commitments to equality and universal access, and concerns about the control of information. The intuitive appeal of these arguments ensures that they find substantial support among well-intentioned legislators, regulators, and much of the public. But intuitive appeal often leads analysis astray. Policymakers need the intellectual courage and fortitude of a scientist when it comes to testing assumptions.

16. More narrowly, the important ideas in these debates require greater nuance than they are ordinarily afforded.

In some cases this lack of nuance yields a false dichotomy, such that consumers (it may be asserted) either need or do not need a given service; in some cases it results from incorrect technical understandings or overly-simplified models; and in other cases it results because the ideas that we debate are really implicit proxies for other political or policy views. In any of these cases, however, the result is that participants in telecom policy debates often talk past each other and adopt entrenched, self-reinforcing, positions.

Faulty Premise: Essential applications and services need high speeds

17. One idealized notion is that if everyone paid the same price for telecommunications, adoption would equalize, and the digital divide would be closed. This is flawed economic reasoning. Demand for broadband is not for the network itself but the services it delivers, also called “derived demand.”6 For example, people do not purchase an airline ticket because they like the flying experience, but rather they need to do some activity at another destination.

18. The idea that high-speed broadband is necessary for education, healthcare, and other social and government services, is related to the assertion that everyone needs low cost access to high speed. This premise is problematic both because it is factually dubious, and also because its power is based in an implicit appeal to 6

**** J. **** et al., “Factor Prices, Factor Substitution, and the Relative Demand for Telecommunications across US Industries,” Information Economics and Policy 5, no. 1 (January 1993): 73–85, doi:10.1016/0167-6245(93)90031-B.

inherently emotional issues. It creates a sense that the only way to support high-quality education, provide access to healthcare and employment opportunities, and address concerns about the digital divide is to support extensive regulatory intervention (subsidies, obligations, or industry supported contribution programs) for high-speed broadband.

19. **** – especially “high-speed” – is not the only or most important metric to consider when provisioning these services. It is unfortunate that certain advocates abuse the indisputable importance of services such as healthcare and education to buttress their argument for government intervention in the high-speed broadband market. At best, this represents a misunderstanding of these services’ actual requirements. It may also represent a desire on the part of broadband advocates to assert their idealized view of how the Internet should be used over the needs of those who actually will rely on these services. At worst, it is a deliberate tactic, being used as an emotional appeal to advocate for a preferred policy that is not otherwise supportable by technical requirements.

20. To the extent that some argue that only fiber will have enough speed and capacity, it is important to note a more fundamental point: engineers optimize – that is they design around the simplest and least costly constraints. This means, that if bandwidth is cheap and plentiful, programmers will design applications that make use of that bandwidth. If, on the other hand, bandwidth is costly, programmers will design applications that make less use of data – and consumers will demand such applications. Indeed, we see this in the mobile wireless industry, in which engineers design applications to minimize bandwidth requirements because mobile bandwidth is relatively expensive. For example, as more users access Facebook with mobile devices, Facebook re-engineered its mobile platform, decreasing average monthly data use from 14MB/mo to 2MB/mo.7 Not only does this lower long term operating costs for Facebook, the lowered data requirement of the platform encourages users to access it more.

21. Consider environments where bandwidth is scarce, for example **** and parts of Asia, and Africa. In these areas engineers and entrepreneurs conceive applications from the beginning as needing to function within strict bandwidth constraints. Due to the limited bandwidth available in these regions, video conferencing and streaming video applications frequently need to be delivered on a less than 1 Mbps connection. As such, design technologies can make more efficient use of bandwidth than does engineering in economies where bandwidth is cheaper and greater.8 Innovation is the result of optimization within constraints, and these emerging country solutions are examples of “world class” telecom engineering.


**** Zuckerberg Comments at Mobile World Congress in 2014 at 28:20, YouTube, (February 24, 2014),


N. Narendra et al., “MobiCoStream: ****-Time Collaborative Video Upstream for Mobile Augmented Reality Applications” (Advanced Networks and Telecommuncations Systems (ANTS), 2014 IEEE International Conference on, IEEE, 2014), 1–6, doi:10.1109/ANTS.2014.7057254, and Dhiman Chattopadhyay et al., “Adaptive Rate Control for H.264 Based Video Conferencing over a Low Bandwidth Wired and Wireless Channel” (Broadband Multimedia Systems and Broadcasting, 2009. BMSB '09. IEEE International Symposium on, IEEE, 2009), 1–6, doi:10.1109/ISBMSB.2009.5133834.

22. Another example is in education, namely the system requirements for video conferencing applications, including programs routinely used for distance education and MOOCs (“Massive Online Open Courses”). Here the requirements are only in the 1-2 Mbps range. The developers of these applications recognize that their products need to work even in low bandwidth environments, so they design their applications to work without high-speed broadband. Adobe Connect, for instance, only requires 512 kbps connection for classroom participants. Coursera, a popular MOOC platform developed by ****, Princeton, the University of Michigan, and the University of Pennsylvania and that today comprises a consortium of over 100 universities, has a mobile-optimized app that allows students to view recorded class sessions on their mobile devices.

Similarly, Adobe Connect has a mobile application that allows for real-time video participation.

23. More bandwidth is of course preferable, but typically is not required for basic operation. In technical terms, it is important to recognize that most of the video delivered in the MOOC setting is highly compressible. Unlike television or movie content, most of the frame is generally static, with relatively simple background settings.

Such video is readily and substantially compressible. Moreover, because MOOC software needs to support the typical student’s computer hardware (e.g., a moderate resolution monitor displaying both in-class video and other class-related materials on a single screen), the typical resolution of video in the online teaching environment will be far below that of HD streaming video services.9

24. Additionally, and perhaps counterintuitively, MOOCs with their large enrollments generally require less bandwidth than smaller online teaching settings. The large class sizes mean that most video will be delivered one way, from the instructor to the students – due to the large number of students – interactivity will be achieved through non-video means (such as quizzes or written questions moderated by an in-class assistant).

In such a setting, the user experience will be less sensitive both to bandwidth and latency variations.

25. This reveals another often overlooked aspect of broadband policy debates: bandwidth is not the only or even the most important, metric. Latency (the time it takes a packet of data to traverse the network), jitter (the change in latency between packets), and packet loss (the percentage of packets of data that never make it across the network) are very important metrics, especially for applications in education and health care – applications where the user may need to interact in real time with a teacher, classmates, or healthcare professional. Substantial or irregular latency and packet loss can lead to jumpy, broken, or lost audio and video – it is far preferable to have a lower resolution but consistent-quality audio and video than high-resolution but unreliable audio and video.

26. The idea that latency and packet loss can be as important as bandwidth is not new. But it is one that plays little role in contemporary policy debates. The failure to appreciate the importance of these metrics is a serious flaw 9

See also **** King, “Many-to-One vs. One-to-Many: An Opinionated Guide to Educational Technology,” AEI, **** 12, 2012, (arguing that the more fundamental change to education enabled to technology is many-to-one teaching through adaptive textbooks, rather than the massive one-to-many model of teaching facilitate by MOOCs).

in these policy discussions. It is akin to having a transportation policy that focuses on miles of highway constructed but pays no attention to whether those highways actually decrease commute times or accidents.10 27. Indeed, where education, healthcare, or other services require high-performance Internet service, one important alternative to provisioning high-speed Internet service in high-cost areas is to rely instead on quality of service (QoS) and prioritization techniques to ensure sufficient performance over lower-speed links. This would not allow a service requiring an average 2 Mbps throughput to operate over a 1 Mbps link – but, where such a service may not function well on even a 3-4 Mbps connection, prioritization could allow it to operate satisfactorily over a lower-speed (e.g., 2 Mbps) link. To make sure this paragraph’s suggestion is clear: lower-speed links that do not adhere to “network neutral” routing may often be able to support the same services that would require a higher-speed (and higher-cost) connection on a neutral network.

28. Another important, and often overlooked, metric, is adoption. In recent years survey evidence, such as the Pew Research Center’s study on Internet and American Life,11 has made clear that availability and price are not the primary reasons that people do not have Internet access. Rather, low adoption results from concerns about usability, relevance, and online harms.12 These concerns are particularly salient among older adults – those who would be most likely to benefit from (or even need) Internet-based healthcare, government, and other services. Indeed one leading researcher of broadband adoption, Pew’s John Horrigan, asserts that spurring demand among non-adopting populations entails comprehensive solutions that require first and foremost partnerships at the community level and training resources. Targeted cost relief is secondary.13 29. Other issues with the idea that high-speed broadband is necessary for these services become clear when looking at each service individually. In the case of health care, for instance, it is unlikely that residential users would have any need for the sort of telemedicine devices that require high-speed connections.14 Rather, consumer-grade healthcare applications are more likely to be used for monitoring and reporting – applications that either send occasional large bursts of data or send consistent, possibly latency-sensitive, small packets of data, and that in either case do not require particularly high-speed connections. The greater challenge for these applications is likely to come from the multiplicity of such devices – the so-called Internet of Things, 10

See, e.g., **** Steinberg and **** I. Zangwill, “The Prevalence of Braess’ Paradox,” Transportation Science, ****1, 1983, doi:10.1287/trsc.17.3.301. It demonstrates that adding roads to a transportation network can actually increase traffic congestion.


“Broadband Technology Fact Sheet,” Pew Research Center: Internet, Science & Tech, accessed July9, 2015, ; **** Smith, “Older Adults and Technology Use,” Pew Research Center: Internet, Science & Tech, accessed July 9, 2015,; **** Smith, “Statement of **** Smith – Broadband Adoption: The Next Mile,” Pew Research Center: Internet, Science & Tech, accessed July 9, 2015,


John B. Horrigan, “The Training Difference: How Formal Training on the Internet Impacts New Users,” **** 31, 2015,


Consumer- and patient-oriented devices are unlikely to require substantial bandwidth. Rather, they are more likely to require a reliable connection of almost any speed. Devices that require high bandwidth, such as MRIs and other advanced imaging devices, are generally housed in institutional settings (e.g., hospitals).

where dozens of devices in one home or millions of devices on larger networks. There is concern that millions or billions of devices, each sending small bursts of data, will overwhelm networks. In such cases, even if the network provides sufficient bandwidth, it may not be able to handle the multiplicity of connections. A network transmitting 100 million small packets per second will be far more congested than the very same network transmitting 10 million large packets per second, even if they are both transmitting the same total amount of data.15

30. The idea of universal service may have grown out of ensuring basic telephone access, but is it worth questioning whether it necessary or even desirable that every broadband technology –whether fiber, coaxial cable, wireless voice, fixed and mobile wireless data, satellite, and even copper, must comply with such requirements. Moreover, emergency, employment, health, government, and e-commerce applications do not require high speeds. Indeed ensuring that an application and content are designed efficiently not only improves user experience and lowers operating costs for content and application providers, but it increases the chances that those technologies can be accessed today on whatever kind of network is available.

31. Such pronouncements of a mandated low price for high speed frequently ignore the important consideration of cost. For example some assert that telecommunications is a basic human right, like water. We can observe however, that while we agree for clean water for everyone, we accept that water should available for human activities but not necessarily for filling the swimming pool. That is to say, there are limits. There is a world of difference between a requirement of 5 Mpbs and 25 Mbps, and the choice can have a material impact on cost.

Emotional arguments in favor of certain telecom policies—that certain policies must be implemented regardless of the cost—is a manipulative technique often used to avoid debate and discussion. However, if any service is as inelastic as advocates purport, then it is more likely to be provisioned by the market anyway.

32. Therefore, if a funding mechanism is implemented for broadband services, then it must be carefully crafted with appropriate checks and balances in order to keep the funding from spiraling out of control, as has happened in the US. The FCC has come under pressure to address the failure of its Universal Service program and set up a “Strike Force” to address its “waste, fraud, and abuse.”16

Faulty Premise: People in rural and remote areas have limited participation in the digital economy because they have only one telecom service provider (TSP) 33. There is no reason to assume that because people live in rural and remote areas and have just one provider that they will inherently be limited in their participation. Indeed being in a remote location can increase likelihood of adoption. A relevant example for Canada is ****. Here is a country where digital health and 15

See, e.g. “Bandwidth, Packets Per Second, and Other Network Performance Metrics,” Cisco, accessed July 9, 2015,


**** Grace, “FCC Chairman **** Announces Universal Service Fund Strike Force” (FCC, July 14, 2014),

education services have near universal adoption and have been delivered for almost a decade across low bandwidth. The user-friendly design of the health application helps support its adoption.17 34. Not only does the country’s geography make broadband deployment difficult, Greenlanders rely heavily on the Internet for health, education, and other social services. Because of the country’s harsh climate, one cannot necessarily go to a school or hospital at will. There is no calling 911 and waiting for an ambulance in most of ****. Recognizing these challenges, TELE **** has made the digital delivery of educational and medical services a top priority. This requires the support of telemedicine and online education platforms that can work with a 1 Mbps connection, such as the Pipaluk telemedicine platform launched in 2008.18 Enabled by an ordinary broadband connection, the device is an all-in-one examination workstation with camera, stethoscope, blood pressure meter, ECG, and spirometer. The device is simply designed so that any person can operate it.19 Pipaluk, meaning “little helper,” can diagnose a long list of of ailments and illnesses20 and is enabled in some 70 locations across the country. Similarly, the ATTAT platform21 delivers an intranet for ****’s schools, teachers, and students to communicate and share knowledge.

35. It turns out that people have different derived demand depending on where they live and even suggests that people in rural areas would pay more for broadband than in urban areas. A study in New Zealand demonstrated that there was a higher uptake of broadband in suburban and rural areas than in cities. The authors surmised that Internet communications (VOIP and email) were cheaper than telephone calls, which benefitted the companies with customers at the farthest distance.22 These two examples demonstrate that the need for services, not the network type or its speed, is what drives demand and adoption.

Faulty Premise: Wireless technologies cannot compete with wireline

36. In the mid- to long-term, as wireless moves into millimeter-wave bands accessing many GHz of capacity, wireless may well supplant cable in terms of throughput. In any case, it’s important to recognize that different users may value the technologies differently, and it is by no means a fait accompli that a basic set of services can only be realized on one kind of technology.


Roslyn Layton, “What the US Can Learn from Broadband in ****,” Tech Policy Daily, May 22, 2015,


“Telemedicin I Grønland - E-Sundhedsobservatoriet,” Dansk Center for Sundhedsinformatik, accessed July 6, 2015,

“Pipaluk - Telemedicin I Grønland” (Mindmakers), accessed July 6, 2015,


Jaaku Lyberth, “Pipaluk - En Succeshistorie I Sundhedsvæsenet,” Kalaallit Nunaata Radioa (KNR), **** 27, 2012,

“Homepage of ATTAT,” accessed July 6, 2015,

Roslyn Layton, “What the US Can Learn from Broadband in ****,” Techpolicy, (May 22, 2015), **** Boyle, Bronwyn E. ****, and Wei ****, “Catching Up in Broadband Regressions: **** Local Loop Unbundling Really Lead to Material Increases in OECD Broadband Uptake?,” July 28, 2008, doi:10.2139/ssrn.1184339.

37. While the characteristics of mobile devices – that they are small and mobile – means that they will not be able to reap these benefits to the same extent as fixed wireless networks, they too stand to see marked improvements in performance. Here, consumers have been voting with their wallets in ways that demonstrate the value of mobility. This is a fundamental point that those who assert wireless cannot compete with wireline broadband have yet to confront: evidence shows that for many consumers wireless does compete.23 Wireless broadband subscription growth is outpacing wireline by more than double in OECD countries.24 Faulty Premise: Regulators should pursue policies of 100% deployment and adoption

38. It is understandable that a telecom regulator would like to achieve a success of 100% deployment and adoption and attempts to pursue such a policy accordingly. But no country in the world has complete adoption. Even if the fastest broadband network were available to everyone, not everyone will adopt. There is a growing literature on why people fail to adopt the Internet even when it is available and affordable.

39. One of the leading scholars of the diffusion of innovation,25 **** Rogers, described that adoption is a social, not economic process. His research was conducted over decades in the American mid-west and focused on why certain farmers adopted new technologies and others did not. Rogers found that even though new equipment and methods were widely available and affordable, certain farmers did not use them. This phenomenon of non-adoption emerged consistently in his studies (approximately 16% of the population of farmers), and Rogers called the segment “laggards” in his bell curve of technology diffusion. However Rogers did discover that when farmers were introduced to technologies by their peers, they were more likely to adopt.

40. Canada, like a number of developed nations, has likely reached the limits of its basic services policy with the focus on the supply side of the equation for broadband. The lesson for Canada is that focusing on a supply-side requirement to address the remaining 13% of Canadians who are not online, whether in the form of faster network technology or lower prices, will not necessarily result in increased adoption. To close the remaining digital divide, the CRTC needs to focus on demand.

41. To address the digital divide, **** Milliard of Denmark’s Technical Institute, has researched and tested a model of Information Communications Technology (ICT) usage for at least a decade. His model appears on the following page.26


In Denmark 7 percent of the population has chosen to rely solely on 3G or 4G mobile connectivity. Mobile-only broadband subscribers outnumber FTTH subscribers by 100,000, even though 100 mbps connections are available to 70 percent of the population. See Roslyn Layton, “The European Union’s Broadband Challenge,” AEI, February 19, 2014,


“Historical Fixed (wired) Broadband Penetration Rates” (OECD, 2014 2003),


**** M. Rogers, Diffusion of Innovations, 5th Edition, 5th edition (New ****: Free ****, 2003).


Kim Andreasson, ed., Digital Divides: The New Challenges and Opportunities of E-Inclusion, Chapter 1 (Boca Raton: CRC ****, 2015).

42. The first level is Access, the proliferation of technologies such as the Internet and broadband networks, computers, mobile devices, online content and services. This is the area of supply that regulatory policy has traditionally tried to promote by ensuring basic services. The second level is the Socioeconomic characteristics of the user; the education, occupation, gender, and age. This is a set of factors that proves highly predictable in whether a person goes online, for example having high education level is highly correlated with going online.

Improving Level 2 requires long term policy initiatives. The third level is Skills. This is the the motivation, opportunities, and needs the user feels in relation to ICT products and services. This level may be easier to acquire than Level 1 or 2 and is not necessarily related to the user’s characteristics. Even if a person does not have the skills, it may be the case that a friend or family member goes online on their behalf, giving the society the benefit of the adoption even if the user does not directly experience it. The fourth Level is the Beneficial use of ICT, i.e., whether and how the supplied ICT products, services, and content are appropriately used to provide benefits for the user. This is important for the CRTC to note, for even if a user has levels 1-3, if he/she does not perceive a benefit to use technology and does not go online, the society does not realize the benefit of that user’s participation. Level 5 is Participation and co-creation of ICT, whether and how the user is actively engaged in contributing to or developing ICT products, services, and content. In summary, basic service policies tend of focus on Access, Level 1 or the supply side of ICT usage, but to have any more impact, governments need to address Levels 2-5, or the demand side of the issue.

43. To give a relevant perspective of the model for Canada, consider the 2014 report “Surveying the Digital Future”27, the longest running studying of Internet use in the world conducted by the Center for the Digital Future. This report focuses on the United States, but is informed by input from some 30 countries including Canada. For the last 12 years, this report has surveyed why people do not go online and “Internet drop-outs”, people who have gone online, but stop going online. In the 2014, the most-cited reason for not going online was lack of knowledge, 22 percent of non-users, up from 17 percent in 2012. The second-most cited reason for not going online was that the Internet was perceived as not interesting or useful – 18 percent of non-users, 27

**** Cole et al., “The 2014 Digital Future Report Surveying The Digital Future Year Twelve” (University of **** California, 2014),

down from 24 percent in 2012 and decreasing for the last three years. Similarly the percentage of non-users who say Internet access was “too expensive” continues to fall –11 percent in 2009, 7 percent in 2010-2012, and 5 percent in 2014.

44. It should not be downplayed that many users feel unsafe online, whether for concerns about privacy or security, or perhaps past bad experiences such as bullying, harassment, or unwanted sexual attention.28 In particular some older adults may feel a sense of fear and vulnerability about the technology and themselves and a fear of breaking equipment or the technology if they do something wrong.29 Age is a factor in non-user’s reluctance to go online, and it underscores the need for human support and training to get this group to adopt the Internet. To put it in other words, if this group does not feel comfortable with the Internet today with a DSL connection, upgrading the connection to fiber does not make it more attractive to them.

45. Fortunately, the leading reason for not going online, that is non-users not having the necessary skills (Level 3), is theoretically the least costly area to address, at least nominally. Deploying digital skills training at libraries and community centers entails the cost of salaries and marginal overhead for rent, utilities, and equipment, theoretically less expense than the deployment of networks in rural areas, ceteris paribus. However, this issue is probably best addressed by other government policies rather than additional regulation on networks.

46. In any case, empirically, Canada’s broadband policy is a success, with 87 percent of households connected to the Internet, an 80 percent increase from the last 5 years.30 Even the world’s most connected nation, Falkland Islands, does not have 100 percent connectivity. Looking at the world’s most connected countries by percentage of households online, Canada is in the league with the best and is by far the largest country geographically to perform so well. If it were not for a few city-states with higher ranking, Canada would be in the top ten in the world and certainly “world class.” Moreover, the Canadian Internet Registration Authority (CIRA) echoes the OECD finding that Canada’s average advertised download speeds equal that of **** Korea, 68.33 Mbps.31

47. Beginning in 2002 and through 2007, Canada led the G7 in fixed broadband penetration (subscriptions per 100 people).32 This means the country was a first mover and grew fast. Today Canada is second only to the UK in the G7 for Internet penetration.33 CIRA34 offers insight to key trends for Canadians based upon analysis and 28


Rowena ****, **** R. ****, and **** E. ****, “Older Adults’ Experiences and Perceptions of Digital Technology:

(Dis)empowerment, Wellbeing, and Inclusion,” Computers in Human Behavior 48 (July 2015): 415–23, doi:10.1016/j.chb.2015.01.062.


“Internet Access, ****, **** and Usage,” CIRA, accessed July 6, 2015,


“Fixed (wired) Broadband Penetration, G7 Countries” (OECD, 2014 2002),


“Homepage of Canadian Internet Registration Authority (CIRA),” CIRA, accessed July 6, 2015,


“Internet Access, ****, **** and Usage,” CIRA, accessed July 6, 2015,

surveys, revealing that more than 80 percent of respondents were satisfied with the speed of their connection and more than 75 percent with the amount of data.

48. Additionally comScore reports that Canadians are second in the world for the amount time they spend online, more than an entire day per month (33 hours), and about half of that is with mobile devices.35 More than three quarters of Canadians have a smartphone, and at least four percent will only use a mobile device to go online and that number appears to be increasing.36

49. The premise of 100 percent adoption should also be viewed from an economic perspective. Basic service policies are frequently justified on general notions of network effects, the notion that as more users come online, the value of the network increases. However, once the network has achieved critical mass, adding additional users may give only diminishing value to the network. This can be said more simply: 100% adoption is not necessary to the maximum benefit of the network. Many networks have far less adoption, but are still considered vital and valuable.

50. To be sure, it is politically popular to say that all Canadians should be online. There is no doubt that the non-user can realize an individual benefit, but it cannot necessarily be extrapolated that adding the remaining 13% who are not online will be as economically valuable as any other tranche of Canadians. This is simply to say that the CRTC needs to be honest and realistic about the cost, challenge, and difficulty about pursuing 100 percent adoption. This is all the more reason to pursue a more modest policy that is grounded in the reality of the actual needs of Canadians evidenced by services they use. A policy that focuses on digital skills (Level 3) will likely have greater effect than simply building new networks or upgrading existing networks to fiber.

Faulty Premise: Canada will fall behind if it does not implement a particular policy 51. The phrase “We are falling behind in ____ (fill in the blank)” can be heard in nearly every developed country in telecom debates. The assumption equates the country as the sum of a single measure, but it begs the question as to better for what and for whom and to what end. Broadband rankings can be created to “prove” that almost any country is the best or the worst. It’s a tool of political grandstanding that releases political leaders of responsibility. It is relatively easy to improve the numbers on discrete, isolated measures. The greater challenge is to ensure that broadband has a productive use in society, something that is far harder to achieve.

Rankings do not in themselves constitute appropriate and sufficient evidence for decision-making.

52. The salvo of “falling behind” is a common refrain for the policy crise du jour where emotion and fear overrule analysis and rigor. The myopic focus on broadband, as simply the sum of discrete measures such as speed or price, miss important nuances about how broadband creates economic and social value. Simply put, broadband is not an end in itself but an enabler. There is no value in being the “best” in any broadband metric 35

“49% of Time Online Now Spent on Mobile Devices, comScore Says - Business,” The Canadian ****, October 10, 2014,


if it does not deliver social benefit. Indeed the OECD Council’s principles for internet policy embrace a range of outcomes, but no one metric of speed or network type.37


53. Services are what people adopt; not a network speed or type. To answer the question as to what should the basic telecommunications service policy be for Canada requires real work in the real world. It requires investigation to what Canadians do online and why they do not adopt even when the networks are available to them at reasonable prices. It also requires a proper inventory and accounting of the essential services and activities for which online access is needed: checking health information, applying for a job online, participating in online education etc. A modest basic telecommunications policy that allows people to participate in the digital economy can achieve more than a policy promising to deliver the fastest speeds to everyone. How to provision communications networks to support important social commitments requires a more critical view than has been offered to date.

54. Crafting the correct policy will likely require different skillsets and tools—anthropological and sociological inquiry and analysis—compared to what telecom regulators have deployed traditionally. There is no doubt that conducting the proper sociological analysis as well as regulatory impact assessments increases the time to decision making and can introduce facts that make the case for any one policy outcome less robust. However it guards against the emotional argumentation and manipulation that certain actions must be done regardless of the cost. **** telecom policy is rarely simple. As such the CRTC should resist temptation to make binary interpretations of the world where more critical views can ultimately deliver better social outcomes.

55. The points I have shared are based upon work prepared for the International Institute of Communications and are available in longer form a forthcoming issue of the Journal of Information Technology & Privacy Law.38 56. To answer any additional questions, I request to appear at the public hearing, ideally remotely. Thank you for the opportunity to participation in this consultation.

***End of Document***

OECD Council Recommendation on Principles for Internet Policy Making, 2011,


See the original article prepared by **** (Gus) Hurwitz and Roslyn Layton, Debatable Premises in Telecom Policy, CMI Working Paper No. 5 (Center for Communication, Media and Information Studies, Aalborg University, 2014), See also Layton “Testing Telecom Assumptions”, InterMEDIA, **** 2015 Vol 43 Issue 1, magazine of International Institute of Communications.

Intervention: Intervenor 230

Document Name: 2015-134.223931.2397214.Intervention(1fdp@01!).html

I have provided my comments in the attached document.Raisons pour comparaitre / Reasons for appearanceComparative studies of basic telecommunications service policies across countries is one area I research. I have written a number of articles on this topic, and I find that the topic needs more critical analysis than it is normally afforded. I would like to give the CRTC the benefit of my learning on the topic, though I request to appear via teleconference if possible.