The Way Forward: Preserving the Openness of the Internet
In this section, we discuss the lack of competition in the telecommunications market within the context of service discrimination. We argue that the industry’s traditional means of securing profit clash with user freedom and subsequently destroy what users value most about the Internet. Next, we argue that a neutral, affordable Internet is key to the success of innovators and entrepreneurs. And lastly, we highlight how Canada has fallen behind in broadband development globally and argue that Canada’s weak digital infrastructure is a key challenge to investment. Subsequently, we argue that Canada needs a digital policy agenda that guarantees the neutrality and affordability of the Internet in order to encourage investment and advance economic growth more broadly.
The Specificities of the Telecommunications Market: Why Open Access Must Be Preserved
While the term ‘discrimination’ has negative connotations, in purely economic terms, ‘discrimination’ or differentiation is not necessarily a bad thing. Future Shop, for example, may choose to offer Apple products at a discount while ordering less Microsoft. In a functioning market, when there is demand for Microsoft, one of Future Shop’s competitors will fill that demand at competitive pricing. Discriminatory traffic management practices (ITMPs) that slow one type of traffic (P2P) in order to give Internet users better access to another (YouTube) are similar in character.323 In fact, ISPs bill P2P throttling as a ‘benefit’ to their customers.324 So, why is discriminatory traffic management so bad?
First, the telecommunications industry is not nearly as competitive as most other industries. Starting a new ISP is far more expensive and onerous than opening a new computer store to sell Microsoft.325 Indeed, since each ISP will need to build a direct physical line to each customer’s home in order to serve them, competition under such a model is not very efficient — it would be like asking a flower delivery company to build its own roads to deliver flowers to its customers’ homes. For this reason, other than a very small number of ISPs, most competitors (wholesalers or secondary ISPs) must use the infrastructure of the main ISPs in order to reach their customers. What this means is that if a primary ISP (Bell) chooses to throttle P2P on its networks, it can and will apply the same throttling to any secondary ISPs on its networks.326 So, in the absence of enlightened digital policy, Internet service markets offer less choice than others.
Second, application-specific throttling is not necessarily an issue that competition may fix. This is because, as noted by Marsden:
ll network owners have incentives to stop traffic flowing over their networks that is low value, high volume and for which it is technically infeasible or uneconomic to charge — notably non-network affiliated content including user-generated and transmitted content.327
Even though ISPs benefit indirectly from P2P and similar traffic, the direct value of such content is to ISP customers and is, in fact, the very reason these customers are willing to pay for Internet access. The incentive for each individual ISP will always be to throttle newly emerging high bandwidth applications.328 That said, it is difficult to know how a newly emergent service will develop — for example, five years ago, did anyone suspect YouTube would become what it is today? For the Internet to retain its power as a driver of innovation, developers must be free to create without the need to rely on an ISP choosing to employ non-discriminatory practices.329 If incumbent ISPs successfully secure their traditional lines of revenue by discriminating against competitors’ offerings (while hurting rates of invention at the same time), they will destroy the aspects of the Internet that make it so valuable to inventors, subscribers, and society, all for the sake of maintaining their own traditional means of profitability.330
Preserving the Open Internet for Innovators and Entrepreneurs
A neutral and affordable Internet is key to the success of enterpreneurs, innovators, and the emergence of new businesses. There are inherent risks involved in entering a competitive marketplace, but the level technical playing field offered online alleviates some of these risks and allows innovators to start small businesses without being well resourced from the outset. In turn, these businesses can enjoy the ability to reach global markets without high physical infrastructure costs. Internet traffic discrimination would therefore significantly hinder startups, service providers such as VOIP companies, and technological innovation more generally.331 Venture capitalists investing in these businesses want assurances that broadband providers won’t throttle their application in favour of an affiliate’s services and applications.332
Where larger firms can typically afford higher quality, often more scalable, services, a non-neutral Internet would compound obstacles for small businesses and entrepreneurs by forcing them to pay for a particular level of Internet service. If a startup or a small business has to pay extra for usage or a more reliable connection, they face a greater degree of difficulty coming to market with applications and services that require a higher level of Quality of Service.333 That is to say, a tiered and discriminatory Internet model makes it very difficult for a startup or a small business to gain market share in the online realm, particularly if they are developing bandwidth or latency-intensive products and services. This small business disadvantage adds to the lack of scalable infrastructure made possible by self-owned server infrastructures that entrenched market competitors often possess, as well as other nontechnical resources incumbents can routinely bring to bear (e.g. organizational expertise, advertising buys, cut-throat product costing, etc.).
A non-discriminatory Internet sets the conditions for small businesses to develop a global user-base, but such efforts may be disproportionately affected by high transit fees and tiered services. As noted in the “Special Section on E-Commerce,” convenience is a key enabler of e-commerce. If potential customers are hindered from accessing a businesses’ website(s) due to discriminatory traffic handling, these practices reduce the convenience of the transaction (e.g. delaying instant gratification) and thus impact the commerical success of the transaction.334 By threatening small businesses’ ability to market to niche and underserved markets, online discrimination makes the global community of prospective buyers less available, limiting online commercial opportunities.335 Thus, discriminatory network practices dilute the advantages that e-commerce companies, particularly smaller businesses, receive from conducting business online: these practices erode convenience, delay gratification, and undermine their capacity to reach niche markets.
Stifling online competition through usage fees or access discrimination has numerous detrimental effects, such as delayed product and service innovation, and higher retail prices. In the face of an increasingly discriminatory communications system, the barriers that small companies and startups face in trying to acquire market share may become so prohibitive that entering the market turns into an unreasonable economic proposition.
Investing in Canada: Preserving Openness to Restore Canadian Competitiveness
In a major 2009 report, World Bank researchers Qiang, et. al. found “a robust and noticeable growth dividend from broadband access in developed countries” since.
All else equal, a high-income economy with an average of 10 broadband subscribers per 100 people would have enjoyed a 1.21 percentage point increase in per capita GDP growth.336
As the researchers further point out, the “availability, quality, and affordability of broadband services are now important factors for international investors when deciding whether to invest in a specific country.”337
Canada has fallen behind on several key broadband measurements compared to other OECD countries, which makes keeping pace with the rest of the OECD pack highly unlikely for Canada’s ICT-related productivity and efficiency metrics. This is not surprising given that the current regulatory environment in telecommunications does little to address the need for transparency and fair Internet access pricing, and thus inhibits ICT investment. As well, Canada’s telecom industry has failed to understand the fact that retail Internet pricing is most efficient when it treats the Internet as inseparable from the content and applications available online (see Figure Four).
Usage-Based Billing is Economically Inefficient
The OpenMedia.ca/CIPPIC submission to the CRTC for its most recent proceedings on usagebased billing (CRTC 2011-77) includes economic analysis that demonstrates that retail Internet pricing is most efficient when it treats the Internet as inseparable from the content and applications available on it. With that in mind, we note how economists predict that flat rates will lead to more optimal pricing. That is to say, users pay higher rates and receive higher value from flat Internet access rates than they do under a usage-based billing (UBB) regime.
Excerpts From The Submission:
The reasons [usage-based billing is economically inefficient]...are essentially two fold. First, the value of an Internet access service to a customer does not correlate to ‘usage’. Rather, it correlates to the myriad online services the customer can enjoy via that access. In this sense, Internet access is essentially a ‘bundling’ of a variety of online services. Economists predict that, where marginal costs are low, bundling becomes optimal. Further, recent economic models have developed that demonstrate this holds even where varied usage scenarios (read ‘bandwidth hogs’ and ‘low bandwidth consumers’, as well as individuals with low disposable income) are factored in. Second, the much discussed network effects which result from the type of increased network usage that accompany flat rates benefit ISPs as much as any other entity in the telecommunications ecosystem.
It is...the online services that customers value in such transactions, not the number of GB per month that they use. "The economic efficiency of Internet access pricing is best analyzed as a ‘bundling’ service.338 Nabipay et. al. explain:"In telecommunications, flat rates can be viewed as a form of bundling a very large number of goods, such as access to hundreds of millions of websites or phone calls to potentially billions of people."339
The economic literature on bundling demonstrates that, where marginal costs are low, flat rate pricing will, in fact, provide higher revenues for ISPs. Nabipay et. al. conclude from their mathematical model that selling ‘bundles’ will permit a seller to “come close to capturing the maximal possible profit... but separate sales never capture more than half the maximal possible profits.”340 The model concludes that with low marginal costs, optimal ISP revenues will occur in many scenarios even if high and low cost users are charged the same flat fee.
Find the full OpenMedia.ca/CIPPIC (CRTC 2011-77) CRTC submission here: http://openmedia.ca/submission- crtc-reply-comments-ubb-proceeding
In Canada specifically, the Information and Communications Technology (ICT) industry is critical. The ICT services sector contributed over $59 billion to the Canadian GDP in 2009, accounting for approximately 5% of the total GDP.341 The open Internet has allowed businesses to communicate with customers, organize and mobilize internal resources, and penetrate new markets at relatively low costs.
In their 2010 consultation paper on Canada’s digital economy strategy, the Government of Canada noted that Canadian businesses have been reluctant to innovate or quickly adopt new ICT systems and platforms, rightly identifying the “state of our digital infrastructure…[as a] key challenge.”342 The Government also recognized that Canadian firms’ underinvestment in ICTs has been linked to slower than expected economic growth more broadly.343
Investing in broadband and providing greater regulatory and platform certainty (through net neutrality and price stability) will enable Canadian businesses to more rapidly, and more significantly, integrate rich ICT systems into their routine practices. Such integration will allow businesses to realize efficiency and productivity gains, therefore improving their success in the Canadian and global economy.
Network investment can increase supply and avoid the neccessity to restrict customer access to the network, but it is expensive — thus, there is an incentive for ISPs to find other ways to decrease traffic. In fact, some have pointed out that the greatest benefit of throttling is its ability to “save millions of dollars in capital expansion costs that would [otherwise] be necessary in order to meet growing bandwidth demands.” 344 The CRTC also recognizes that network investment is a superior method of dealing with network congestion, followed by economic measures and, as a last resort, technical measures (e.g. throttling).345
We know from “The Technical Case for Openness” section that traffic growth in Canada has stayed relatively stable. The question is, have Canadian ISPs fully utilized network investment to deal with that stable growth and to stave off potential network congestion? One way to determine if network capacity investment has been reasonably exausted is to examine the network infrastructure capital expenditures both in absolute terms and in relation to revenues, while at the same time comparing both of these measures to patterns found in other comparable nations. We found that Canada’s dominant telecommunications companies and ISPs invest less in network infrastructure compared to other industrialized countries (see Figure Five). Thus, network investment is far from exhausted as a means to deal with Internet congestion and Canada’s falling status in key Internet metrics.
How Much Do Canada's Dominant Telecoms and ISPs Invest in Digital Infrastructure
Almost all OECD countries saw a huge surge in capital investment in their telecommunications networks during the Telecom-Media-Technology (TMT) bubble years (1996-2000), followed by a steep decline until roughly 2003, and a steady rise since. Canada saw a similar pattern, with network investment peaking by 2000-2001 as the telecom and cable companies invested between one-fifth and one-quarter of their total revenues back into the public network, followed by a sharp drop thereafter. The difference between Canada and most other OECD countries, however, is that investment levels have stayed relatively flat throughout most of the last decade, with no significant upward swing in recent years to make-up for the ‘lost years’ following the collapse of the TMT bubble.346
Since 2001, investment in the telecoms network infrastructure in Canada as a percentage of revenue has hovered between 16 and 20% and consistently lagged behind the OECD average.347 In contrast, the OECD points to cable operator ONO in Spain, with “the highest level of investment relative to revenues in 2007 at 33%” as well as to Telstra (22%), Time Warner Cable (22%), Telenor (21%), and Comcast (20%) as examples of other companies with high levels of investment relative to revenues.348
The same applies with respect to three other conventional measures of investment: as a percentage of gross fixed capital formation (GFCF); per total communication access path; and on a per capita basis. And in each case, the story of the Canadian telecoms industry is the same: a couple of years during which investment exceeds the OECD average, but on average, investment levels that have failed to keep pace with the OECD average in six or seven of the past ten years for which data is available.349
A stable innovation framework will enhance the growth of the ICT sector as a whole. This framework requires a digital agenda that includes a provision guaranteeing that the Internet remains neutral. Past innovation depended on the fair treatment of data as it flowed across the networks of Canadian ISPs. If innovation is to fluorish, Canada must safeguard the neutral Internet. Further, as Figure Five illustrates, Canada must encourage the telecommunications industry, and especially incumbent players, to invest more in network infrastructure.
We believe that a stable, non-discriminatory Internet ecosystem is essential to both economic growth and revitalizing and re-invigorating investment in Canadian ICT. Policymakers should quickly reorient regulators and industry alike in order to enable and encourage Canadians to innovate and fully take advantage of the economic and social benefits of the Internet.
323. CRTC. (2009, July 14). Transcript of Proceeding. Retrieved from http://www.crtc.gc.ca/eng/transcripts/2009/tt0714.htm See line 6825: “in designing and implementing ITMPs to improve the overall experience for the vast majority of users.”
324. See Bell. “Network Management” In Troubleshooting FAQ here http://service.sympatico.ca/index.cfm?method=content.view&content_id=12119. Bell explains throttling to its customers as such: Bell is using the latest, state-of-the-art technology to improve the customer experience for a vast majority of our customers’ favourite applications (such as Internet Browsers, E-mail, Instant Messaging, Streaming Video, etc.) as required during peak periods on the Internet, while ensuring all customers receive fair use of the network when there is heavy Internet traffic.
325. Noam, E.M. (1992). Network Tipping and the Tragedy of the Common Network: A Theory of the Instability of Public Telecommunications Systems. In C. Antonelli (Ed.), The Economics of Information Networks, (107-127). North-Holland: Elsevier. Retreived from http://home.himolde.no/~molka/in765/network-tipping-noam.pdf
326. CRTC. (2009). Telecom Regulatory Policy CRTC 2009-657. Retrieved from http://www.crtc.gc.ca/eng/archive/2009/2009-657.htm at paras. 74-76.
327. Marsden, C.T. (2007). Net Neutrality and Consumer Access to Content. SCRIPTed, 4(4), 416. Retrieved from http://www.law.ed.ac.uk/ahrc/script-ed/vol4-4/marsden.pdf
329. Benkler, Y. (2003). The Political Economy of the Commons. European Journal for the Informatics Professional 6, 4(3), 8. Retreived from http://www.benkler.org/Upgrade-Novatica%20Commons.pdf
330. Van Schewich, B. (2010). Internet Architecture and Innovation (p. 317). Cambridge, Mass.: The MIT Press.
331. Handler, C.E. (2009, February 23). The Struggle Over ‘Net Neutrality’ - New developments heat up this vital Internet and e-commerce issue. Law Technology News. Retreived from http://www.law.com/jsp/lawtechnologynews/PubArticleLTN.jsp?id=1202428474927
332. Werbach, K. (2010, August 18). Why Network Neutrality Is Good for Business. Harvard Business Review Blog. Message posted to http://blogs.hbr.org/cs/2010/08/why_network_neutrality_is_good.html
333. Internet Business Law Service. (2007, May 21). INTERNET LAW - ‘Net Neutrality’ Laws Will Cause $70 Billion Loss. Retrieved from http://www.ibls.com/internet_law_news_portal_view.aspx?id=1760&s=latestnews
334. Grewal, D., Iyer, G.R., & Levy, M. (2004). Internet retailing: enablers, limiters and market consequences. Journal of Business Research, 57, 703-713.
335. eBay Main Street. (2010). Public Policy for the E-commerce World - Net Neutrality, eBay. Retrieved from http://www.ebaymainstreet.com/policy-papers/net-neutrality
336. Qiang, Z., Rossotto, C., & Kimura, K. (2009). Information and Communications for Development 2009: Extending Reach and Increasing Impact. The World Bank (p. 45). Retrieved from http://issuu.com/world.bank.publications/docs/9780821376058
337. Ibid., p. 43.
338. Nabipay P., Odlyzko A., & Zhang, Z. (2011). Flat Versus Metered Rates, Bundling, and ‘Bandwidth Hogs’ (p.1). Retrieved from http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1829784
339. Ibid., p. 6.
340. Ibid., pp. 1‐2.
341. Industry Canada. (2010, November). Information and Communications Technologies Statistical Overview. Retrieved from http://www.ic.gc.ca/eic/site/ict-tic.nsf/vwapj/0105864eng.pdf/$FILE/0105864eng.pdf
342. Government of Canada. (2010). Consultation Paper on a Digital Economy Strategy for Canada. Retrieved from http://de-en.gc.ca/consultation-paper/consultation-paper-6
344. Arbor. (2009, February 23). Comments to Telecom Public Notice 2008-19, Response to PN question 2(d). Retrieved from http://www.crtc.gc.ca/partvii/eng/2008/8646/c12_200815400.htm
345. CRTC. (2009, October 21). Review of the Internet traffic management practices of Internet service providers. Retrieved from http://www.crtc.gc.ca/eng/archive/2009/2009-657.htm
346. Organization for Economic Cooperation and Development (2009). Communications Outlook 2009. Paris: OECD. pp 115.
347. See Ibid. pp 141; and Canadian Radio-television and Telecommunications Commission (CRTC). (2010). Communications monitoring report. Retrieved from www.crtc.gc.ca/eng/publications/reports/policymonitoring/2010/2010Monito...
348. Organization for Economic Cooperation and Development (2009). Communications Outlook 2009. Paris: OECD. pp 116.
349. Ibid. pp. 141-143