A Future Friendly Digital Economy Strategy: Submission to the Government of Canada's Digital Economy Consultation — Appendix 5
Theme(s): Digital Infrastructure, Growing the ICT Industry, Innovation Using Digital Technologies
Comments of Robert W. CrandallFootnote 1
Introduction and Summary of Conclusions
I have been asked by TELUS to address questions raised by the Commission in its Call for Comments in Telecom Notice of Consultation CRTC 2009-261-7 (Notice). Specifically, I shall address the issues raised in paragraph 11 of that Notice, which asks for comments on:
- The application of the existing essential service framework on a forward-looking basis such that it provides appropriate incentives for continued investment in broadband infrastructure, encourages competition and innovation, and leads to consumer choice; and
- In the context of the discussion provided in response to A, above.
- whether the speed-matching requirement, mandating the provision of the high-speed access services under consideration, or mandating access to any new types of Internet access infrastructure does, or would, unduly diminish incentives to invest in new network infrastructure in general and, in particular, in markets of different sizes;
- whether, in the absence of the speed-matching requirement and the mandated provision of the high-speed access services under consideration, there would be competition sufficient to protect the interests of users; and
- whether the impact of these wholesale requirements unduly impairs the ability of incumbent telephone companies to offer new converged services, such as IPTV.Footnote 2
I shall address these issues by utilizing a White Paper that Debra Aron and I prepared in 2008. This White Paper was submitted by TELUS in its petition to the Governor in Council on 11 March 2009.Footnote 3 In addition, I shall refer to research and relevant information that have appeared since the preparation of our 2008 White Paper.
Our 2008 White Paper concluded that information and communications technology (ICT) investment has been the principal driver of recent economic growth in developed countries and that Canada has lagged behind the United States in ICT investment. For these reasons (among many), the Commission should be concerned that its policies do not discourage network investment in telecommunications.
In addition, we reviewed the evidence on the effect of network unbundling regulation in telecommunications and concluded that such regulation has little or no effect on broadband penetration – that is, unbundling provides for more potential vendors of essentially the same service (DSL) delivered over the same network, but this increase in the number of vendors does not result in a greater number of subscriptions. Moreover, network unbundling has a negative effect on telecom investment, a conclusion at odds with Commission's views expressed in Telecom Decision CRTC 2008-117.Footnote 4
For these reasons, we concluded that:
"It is therefore imperative that the CRTC provide an environment that does not impose additional obstacles to such investments. Among the most important potential threats to the economic viability of next generation networks is the prospect that regulators will require unbundling of these networks at regulated wholesale prices. Such regulation would be extremely damaging to investment incentives."Footnote 5
Because this White Paper report was written 16 months ago, I have been asked by TELUS to update it based on any recent developments and published research on the effects of mandated access to incumbents' high-speed network facilities. The empirical evidence that has accumulated since the preparation of the Aron-Crandall White Paper, drawn largely from international experience with various regulatory regimes, strongly supports our earlier conclusions. In particular, there are several recent studies that conclude that regulated access to incumbents' networks through unbundling discourages investment in new technologies. Moreover, the accumulating empirical evidence supports the conclusion that platform competition increases broadband penetration while unbundling regulation has no salutary effects on broadband subscriptions. While a recent report to the U.S. Federal Communications Commission (FCC) tried to recast this evidence in support of unbundling, its effort has been discredited, and an FCC official has been quoted as rejecting this analysis. Thus, the accumulating empirical literature and international experience strongly support the conclusion that the Commission's actions that were submitted to Cabinet for review would reduce investment in advanced Canadian broadband networks without any major offsetting benefits in the form of increased broadband penetration.
Given the evidence presented in the 2008 Aron-Crandall White Paper and empirical research and international experience that has accumulated since that time, I conclude that:
In response to question A, posed in paragraph 11 of the Notice:
- The Commission should not extend its existing essential facilities framework to new broadband network facilities because such a regulatory approach necessarily reduces incentives for network investment and does not generally promote competition or increased broadband penetration. The international evidence strongly suggests that an aggressive approach to unbundling new network facilities, such as that pursued in Europe, leads to less network investment than occurs in the United States, which has decided to forbear from all such regulation of new broadband network facilities. Nor is there any international evidence that treating broadband networks as essential facilities and requiring bitstream access or network unbundling of these facilities leads to systematic increases in broadband penetration.
In response to question B, posed in paragraph 11 of the Notice:
- The Commission's mandating of wholesale access to incumbents' facilities for the provision of DSL services at "speeds that match the speeds at which ILECS provide high speed Internet services" would reduce the incentives for carriers to invest in advanced network facilities. The evidence is particularly strong for investment by new entrants, who have little incentive to deploy their own networks once regulators require incumbents to lease their networks to them at regulated rates. There is also evidence that such unbundling requirements reduce the incentives for all carriers, including incumbents, to invest in advanced network facilities, such as fiber to the premises. This evidence is clearly at odds with the Commission's conclusions in Telecom Decision CRTC 2008-117.Footnote 6
- Given that there is very little evidence that unbundling regulations would increase broadband penetration, extending such rules to broadband through speed-matching requirements would not increase broadband competition in Canada regardless of the degree of competition that currently exists. The empirical literature shows that inter-platform competition between cable companies and ILECs, which is vigorous in Canada, is far more effective than the intra-platform rivalry provided by matching-speed unbundling requirements. Because of its adverse effect on network investment, unbundling of broadband facilities would actually reduce prospective competition, not increase it.
- Finally, I conclude that extending speed-matching unbundling requirements to incumbents' high-speed networks reduces the potential profitability of incumbents' IPTV services and, therefore, reduces the likelihood of extensive roll-out of new network architectures, such as fiber to the premises or to the curb. The economic returns from developing IPTV services over enhanced telecommunications networks are subject to enormous uncertainty; adding onerous regulatory requirements, such as "speed-matching" wholesale unbundling will simply add to this uncertainty and reduce the incentives to deploy the necessary facilities to deliver IPTV.
The Aron-Crandall White Paper
Our 2008 analysis focused extensively on the relationship between investment in information and communications technology (ICT) and economic growth. We summarized a burgeoning literature that establishes the importance of such ICT investment, particularly in advanced societies. We concluded that:Footnote 7
- ICT investment generates substantial benefits both to the economy as a whole and to individual citizens. The evidence is that ICT investment is a major driver of labor productivity growth. ICT investment in broadband infrastructure in particular brings to consumers the benefits of greater access to a myriad of services, including health care, education, government services, entertainment, and media, that enhance well-being.
- Canada's leadership in broadband deployment is now threatened because it has not yet deployed the extremely fast, "next-generation" networks that can deliver large amounts of data and video services – networks that are currently being deployed in the United States by Verizon and AT&T and in other countries as well.
- Canada also lags behind the United States in investment in ICT and, consequently, in productivity growth. Promoting economic growth in Canada requires that public policies be designed to facilitate efficient investment in ICT, particularly in the next-generation networks that are vital to unlocking the potential of new information technologies, and that such policies be established promptly.
- Deploying new, very high-speed networks requires massive investment in a risky market environment. The risk of deploying these networks is far greater than the risk that the telecommunications companies faced in deploying copper networks in the regulated monopoly era. It is therefore imperative that the CRTC provide an environment that does not impose additional obstacles to such investments.
- Among the most important potential threats to the economic viability of next generation networks is the prospect that regulators will require unbundling of these networks at regulated wholesale prices. Such regulation would be extremely damaging to investment incentives.
In support of these propositions, we noted that:
"The rapidly-accumulating evidence from the international arena is that investment in information and communications technology (ICT) is the principal driver of labor productivity growth and, therefore, overall economic growth in developed nations. A major component of ICT capital is the broadband infrastructure required to transmit information in the modern economy. The benefits of broadband infrastructure are not abstract—in addition to driving economic growth, they manifest themselves in providing to Canadian citizens and businesses increased availability, quality, and efficiency of a myriad of services that are central to citizens' well-being, including healthcare, education, government services, commerce, media, and entertainment."Footnote 8
Turning to the telecommunications sector, we pointed out that billions of dollars of investment are required to build advanced, high-speed networks to serve households and businesses in the current environment:
"The development of these advanced broadband networks requires huge, risky investment at a scale that can threaten a company's viability. Canadian companies must compete in the global capital market for such investment funds, and investors must find the prospects for earning a return sufficient to be willing to accept the risks. Unfortunately, Canada lags substantially behind the United States in spending on information technology and, as a result, its productivity growth has also lagged.
"Among the most important inhibitors of capital investment in telecommunications is ill-advised regulatory policy. Rational investors in telecommunication technology must be assured that the regulations that will govern the new, "next generation" networks will not undermine their investment before they will commit to funding them. In particular, investors must be assured that mandatory unbundling of these new network facilities will not be imposed, because mandatory unbundling at regulated wholesale prices deprives the owner of the ultimate control of the assets and severely reduces the returns on these very risky projects."Footnote 9
We also concluded that:
"Empirical analyses and case studies document the damaging effects of unbundling regulations on investment in the U.S., Europe, and elsewhere. The research also documents the beneficial effects of intermodal (investment-based) competition on broadband penetration, and the insignificance of intramodal (unbundling-based) competition on broadband penetration. That is, the empirical studies find that broadband take-up rates are increased by competition between network platforms such as cable and DSL, in which the providers must make their own, at-risk investments. In contrast, broadband take-up rates do not appear to be materially affected by the kind of competition engendered by unbundling."Footnote 10
In short, we strongly recommended that the Commission avoid the imposition of network unbundling requirements on new, high speed network facilities. Instead, the Commission should rely on platform competition among telecom platforms, cable television platforms, and other new technologies that are arising to deliver high-speed Internet services.
Since we completed the Aron-Crandall White Paper, several studies on the impacts of broadband regulation have appeared. In particular, the following studies are germane to this proceeding:Footnote 11
- John de Ridder, "Catching-up in Broadband – What Will it Take?" Organisation for Economic Co-Operation and Development, July 25, 2007.
- Boyle, G., Howell, B., and Zhang, W., "Catching up in Broadband: Does Local Loop Unbundling Really Lead to Material Increases in OECD Broadband Uptake?" New Zealand Institute for the Study of Competition and Regulation Working Paper, July 2008.
- Bouckaert, J., VanDijk, T., and Verboven, F., "Regulation and broadband penetration—what is required to regain speed in Belgium?" Belgium: University of Antwerp and Leuven, 2008. Available at: http://www.ua.ac.be/download.aspx?c=jan.bouckaert&n=72967&ct=68422&e=184390S.
- Friederiszick, H., Grajek, M., and Roller, L., "Analyzing the relationship between regulation and investment in the telecom sector," ESMI White Paper No. WP-108-01, 2008.
- Jung, I, Gayle, P.G., and Lehman, D.E., "Competition and investment in telecommunications," Applied Economics, 40(3), 2008, pp. 303–313.
- Wallsten, S., and Hausladen, S., "Net neutrality, unbundling and their effects on international investment in next generation networks," Review of Network Economics, 8(1), 2009, pp. 90–112.
All of these studies, except the de Ridder and Jung, et.al., studies, conclude that unbundling is an ineffective regulatory policy for increasing broadband penetration and/or that unbundling discourages network investment. However, the contrary conclusions of the de Ridder study have been shown to derive from a poorly-specified econometric model by Boyle, et.al. (2008) and by Crandall, Ehrlich, and Eisenach (2009).Footnote 12 I will return to this study and to the Boyle, et.al. study that critiques it when, later in these comments, I address the FCC's current policy process in developing a national broadband plan. Each of the other studies cited above, except for Jung, et.al., concludes that unbundling requirements for new broadband networks are largely ineffective and/or damaging to network investment.
The Bouckaert, et.al. study begins with a review of unbundling practices in Europe and a review of previous studies. Like Aron-Crandall, it concludes that "…the econometric studies conclude that intra-platform competition through mandatory unbundling has hardly contributed to, or even slowed down, the penetration rate of broadband services. Inter-platform competition, on the other hand, has positively contributed to broadband penetration."Footnote 13 Its econometric analysis of broadband penetration across 20 countries provides new evidence that inter-platform competition (from cable television platforms) is positively associated with broadband penetration while intra-platform competition through bitstream access and local loop unbundling is associated with lower broadband penetration. This conclusion is buttressed by its analysis of broadband subscriptions across the four regions of Belgium, which shows that regions with the most intense cable competition have the highest broadband penetration. Their results thus confirm the general conclusions that emerge from the empirical literature on broadband: broadband penetration is positively affected by inter-platform competition, but network unbundling does not increase competition and, therefore, the output of broadband services.
Bouckaert, et.al., could not be more emphatic in their policy conclusions. They criticize Belgian policy for relying on bitstream access and local loop unbundling:
"Belgian broadband regulation has focused too much on short term goals, such as encouraging entrants to make use of service-based access regulation. This policy has been at the expense of investment in infrastructure-based broadband competition. Promotion by the national regulatory authorities of competition in the telecommunications sector by encouraging efficient investment in infrastructure is, however, one of the main policy objectives and regulatory principles of the Common Framework Directive. A strong commitment by the regulator on its position towards fiber-based networks and a long-term vision on how to encourage alternative broadband infrastructure is, therefore, urgently welcomed. This urgency to develop a long-term vision is even more present since there is increasing theoretical and empirical evidence that the "ladder of investment" theory underlying broadband access regulation to encourage short-term competition is false."Footnote 14
Friederiszick, et.al., address network investment rather than broadband penetration. They estimate an econometric model of a firm's infrastructure (tangible fixed assets) as a function of demand, cost, competitive, and regulatory variables. They also account for the effect of various market variables on the political determinants of the regulatory regime. Thus, they correct for the "endogeneity" problem that could affect single-equation models. Their results show that as their regulatory index increases because of more unbundling, line sharing, or bitstream access requirements, entrant investment declines sharply. They conclude that "…our results suggest that the entrants would more than double their infrastructure over 5 years if they did not have regulated access to the incumbents' local loops."Footnote 15 However, they do not find a similar result for incumbents' investment. But because their data set does not include incumbent investment in next-generation networks, they caution that their failure to find an effect on incumbents' investment does not extend to the decision to invest in new network architectures.Footnote 16
The Friederiszick, et.al., results are particularly damaging to the theory that unbundling is a regulatory mechanism designed to increase competition in broadband services. They conclude that entrants are far less likely to invest and therefore migrate to their own platforms from which they could become vigorous competitors.
Jung, et.al., examine investment by U.S. incumbent telephone companies in 1997-2002 as a function of population, household income, revenue per line, price-cap regulation, and competition from competitive local exchange carriers (CLECs). They find that an increase in CLEC market share has positive short-term effects on incumbent investment, but that these effects do not persist. That is, CLEC competition does not have a sustained effect on incumbent investment. Moreover, their results are based on data from a period when U.S. CLECs were using unbundled network elements (UNEs) largely for offering traditional telephone service, not for offering broadband Internet services. Their competition variable is the CLEC's share of traditional telephone access lines, not the share of broadband lines; hence, their results are less applicable to the issue facing the Commission in this proceeding.
Wallsten and Hausladen use European data for 2002-07 to estimate the effects of unbundled DSL loops on the deployment of fiber to the home (FTTH) by incumbents and entrants. They find that the number of unbundled loops per capita and the number of loops provided through bitstream unbundling per capita are inversely related to the deployment of fiber by both incumbent companies and entrants. Moreover, the number of unbundled lines per capita is also inversely related to the number of broadband lines provided by cable television platforms, wireless local loops, and DSL over the entrants' own facilities. In short, Wallsten and Hausladen find that unbundling reduces competition in broadband by reducing the incentive of incumbent telephone companies and other communications carriers to invest in fiber and in alternative platforms.
These recent studies – with the exception of Jung, et.al., which does not bear directly on broadband investment – support the conclusions reached in Aron-Crandall. Broadband unbundling does not increase penetration, but it reduces investment incentives. Thus, the accumulating empirical evidence directly contradicts the conclusions reached by the Commission in Telecom Decision CRTC 2008-117.Footnote 17 Unbundling does not increase competition in the broadband market. If it did, one would expect to find that jurisdictions that require network unbundling would, all other things equal, have lower broadband prices and, therefore, higher broadband penetration. The studies do not find such an effect. On the other hand, there is substantial evidence that unbundling requirements reduce network investment by incumbents and entrants.
The U.S. FCC Broadband Proceeding
The Federal Communications Commission (FCC) renounced wholesale regulation of broadband networks in 2005 after a lengthy legal battle over its jurisdiction over cable modem services was resolved by the courts.Footnote 18 The 2009 U.S. Economic Stimulus package required the FCC to devise a broadband "plan" and associated broadband regulatory strategies.Footnote 19 The broadband plan was to be announced in February, but the deadline has now apparently been deferred until March. As part of its inquiry, the FCC commissioned two studies, one from Harvard's Berkman Center and one from Columbia University's Columbia Institute for Tele-Information (CITI). In this report, I focus only on the Berkman Center Report,Footnote 20 which provided strong policy recommendations; the CITI Report, while providing very useful empirical summaries of capital spending, did not advance major policy conclusions.
The Berkman Center Report was delivered in October 2009, and it generated substantial controversy because, surprisingly, it supported a return to a vigorous unbundling policy as a mechanism to promote increased broadband penetration and improved broadband services.Footnote 21 The Report examined broadband policies in a number of countries and concluded from this review and an econometric exercise that unbundling can increase broadband penetration. The problems with its review of international experience are many, as I have pointed out elsewhere. For the sake of brevity, I provide only a brief review of this critique in these comments. For the more exhaustive analysis, I refer the reader to the full analysis I and two co-authors provided in a declaration filed with the FCC.Footnote 22
The international comparisons provided by the Berkman Center Report are simply not convincing. For instance, the Report refers approvingly to the alleged effects of unbundling in several Nordic countries and in the Netherlands without pointing out that platform-based competition from cable television and government-provided fiber networks (in Sweden) generally provide many more broadband connections than unbundled copper loops. It cites France, which relies heavily on network unbundling, as a "success" and Germany which relies much more heavily on platform competition from cable as a "failure" despite the fact that the two countries have similar broadband penetration and essentially no fiber to the premises. And the Berkman Center Report reflects approvingly upon the United Kingdom's "success" with network unbundling and functional separation without pointing out that since the UK adopted its new, more aggressive policy, broadband growth has slowed substantially. Before the change in policy, UK broadband subscriptions were growing more rapidly than connections in the rest of the EU-15; since the new policy, they have grown more slowly.
The Berkman Center Report also claims that Japan has been the beneficiary of considerable investment in fiber to the premises despite the fact that the Japanese regulator (MIC) has required the unbundling of this fiber.Footnote 23 If this were true, it would certainly contradict most of the available empirical evidence on the adverse effects of unbundling regimes on network investment. But while it is correct to say that MIC has nominally required its incumbent carrier, NTT, to unbundle its fiber facilities, there has been little such unbundling in Japan because the regulator apparently has set the wholesale price of fiber very high and does not or cannot require unbundling of NTT's individual fiber connections. As a result, NTT apparently has not been forced to unbundle its new fiber to the premises despite the professed public position of its regulator. This has allowed NTT to invest heavily in fiber to the premises with little fear that it will have to lease the fiber to its competitors.
The leading DSL competitor in Japan, Softbank, has found it impossible or uneconomic to use NTT's fiber and has therefore retreated, allowing its market share of broadband to decline substantially. In its most recent annual report, Softbank said that is pulling back from the residential market and is not investing in broadband with the result that: ". . . the [broadband] segment's profitability and cash flow are showing marked increases. This reflects a lower depreciation from the end of major capital expenditures, combined with a structural transformation toward businesses."[emphasis supplied]Footnote 24 In this annual report, Softbank also showed that its total broadband lines in service have been declining since 2007, as subscribers shift away from its DSL service delivered over unbundled NTT loops to the fiber-based services of NTT and other platform-based competitors. Thus, Softbank apparently is not investing in residential broadband, nor is it using NTT's fiber facilities to offer higher-speed services to residential consumers. Instead, it is quietly allowing its position in the residential broadband market to erode.
Another competitor, KDDI, has recently decided to expand its own footprint by purchasing a 38 percent interest in a large Japanese cable company.Footnote 25 This acquisition allows KDDI to gain access to half of Japan's households over its own cable facilities, an important consideration because of the extremely high charges required for gaining access to NTT's fiber facilities, according to KDDI's President, Tadashi Onodera.Footnote 26 Broadband competition in Japan is thus intensely focused on platform competition among cable companies, power companies, and NTT. Investment of fiber has occurred not "despite" network unbundling of fiber, but precisely because such unbundling is not a realistic alternative for competitors. As a result, platform-based competition has been so successful that fiber-to-the-home subscriptions now far exceed DSL subscriptions, and intra-platform (DSL) competition is rapidly receding.
In its econometric analysis of the effects of unbundling on broadband penetration, the Berkman Center Report is particularly misleading because it re-estimates the deRidder econometric model, referred to above,Footnote 27 using an arbitrarily altered measure of unbundling. deRidder had used the number of years that unbundling had been in place (GUYRS in Table 1) to measure the potential impact of unbundling on penetration. Boyle, et.al., criticized his econometric model on several grounds, including the fact that his unbundling variable serves as a proxy for the number of years since broadband was first introduced in many countries. Given that broadband penetration grows over time in a traditional "S-curve" fashion, it is important to separate the effect of the time since broadband was introduced from the time that unbundling began. When Boyle, et.al., do this, the statistical significance of the unbundling variable disappears.Footnote 28
The Berkman Center Report attempts to rescue the deRidder model by making changes to the key policy variable, GUYRS, the number of years since unbundling began. I reproduce the changes in Table 1 below. In most cases, the Berkman study simply discards years of unbundling in which it believes that regulators were insufficiently aggressive or incumbents were not cooperative. In the case of the United States, which has had unbundling since 1996 and had line sharing for four years (1999-2003), the Berkman Report simply replaces deRidder's GUYRS measure of 10 years of unbundling prior to 2005 with a zero. It increases the number of years of unbundling for Korea despite the fact that Korea's incumbent telephone company did not face an unbundling requirement until 2001. It reduces the number of years of unbundling for the Netherlands because the regulator did not allow entrants access to "naked DSL" for several years, and despite the fact that by the end of 2005, deRidder's terminal year, Netherlands' entrants used unbundled lines to reach fully 28 percent of the country's DSL customers. Finally, the Berkman Report drops Switzerland completely, apparently because Switzerland is an outlier that had relatively high broadband penetration and no unbundling as of 2005, a result that is in conflict with the Berkman Center's conclusions. Once the GUYRS variable is arbitrarily modified in this fashion, the Berkman Center Report finds that it can obtain a statistically-significant positive coefficient for it in its broadband penetration regressions. But this result derives from the arbitrary adjustments of the data, not from an underlying positive effect of unbundling on broadband penetration.Footnote 29
For all these reasons, the Berkman Center Report should not be relied upon by regulators. Indeed, after the FCC received the Report, the head of the FCC's broadband task force, Blair Levin, was quoted in a newspaper interview as saying that:
"As to (line-sharing rules), the courts threw that out, and we're not that terribly interest(ed) in moving toward things that will just freeze capital investment and have long, drawn out court battles…that doesn't strike me as that productive"Footnote 30 [emphasis supplied]
The Effect of the European Union's Unbundling Policy
European regulators have been strongly encouraged by the European Commission in Brussels to promulgate aggressive unbundling policies to allow intra-platform competition in DSL services. In the last two years, the Commission has even promoted "functional" separation of incumbents' retail and wholesale activities, and the outgoing EC Commissioner in charge of telecommunications openly suggested that structural separation may be required in some cases.Footnote 31 Although the EC has not finalized its regulatory approach to new, fiber-based networks, it is considering the extension of unbundling to these new networks as well.
These aggressive regulatory policies have had little effect on broadband penetration, as Aron-Crandall and some of the studies cited above demonstrate. However, they are surely inhibiting investment and are thus likely responsible for Europe's slow deployment of fiber to the home, as shown in Figure 1. Note that these data show that barely 1 percent of European households subscribe to fiber-based connections while North American (largely, U.S.) connections are more than four times as great. In the EU-15, there were only 1.27 million fiber-based household subscriptions in the first quarter of 2009 according to ECTA.Footnote 32 Of these, more than half were in Sweden where municipal governments have constructed the underlying fiber networks, not private companies. Another one-quarter of EU-15 FTTH subscriptions were delivered over entrants' facilities in Italy, largely over Fastweb's network. Outside of Italy and Sweden, there were fewer than 300,000 household fiber connections and just 28,500 over incumbents' facilities.
The impact of network unbundling regulation on fiber deployment, captured econometrically by Friederiszick, et.al., above, is best illustrated by Fastweb's recent decision to cease expanding its fiber deployment in Italy. This decision was taken after the Italian regulatory authority streamlined its unbundling policy, allowing Fastweb to gain access to Telecom Italia's unbundled loops to deploy DSL. This has allowed Fastweb to deploy broadband at a lower cost, albeit with lower potential speed, to additional subscribers. The availability of unbundled loops, therefore, directly discouraged further network investment and innovation in fiber-based facilities.
In previous proceedings, I have shown that the European Union's attachment to network unbundling and functional separation have resulted in less investment by their large incumbent telecom companies than one observes in the United States and Canada. This disparity continues. In 2008, the latest year for which I have data, Canadian incumbents (TELUS and BCE) invested 17.7 percent of their annual revenues, and U.S. incumbents (AT&T, Qwest, and Verizon) invested 16.5 percent of revenues. By contrast, the twelve EU-15 incumbents for which I have capital expenditure data, invested an average of only 14.5 percent of revenues.Footnote 33
The Prospective Effect of Unbundling Requirements on IPTV
Cable companies have successfully invaded the ILECs' traditional voice telephony business through a variety of VoIP offerings. However, the ILECs have a much more difficult task to reciprocate by invading the cable companies' traditional video business because their copper-based facilities, as currently configured, are largely inadequate to deliver video services. A major route to such entry is the deployment of fiber to the home (FTTH). In the United States, Verizon has moved aggressively to deploy fiber in its region, and it has had a modest degree of success in doing so, attracting between 20 and 25 percent of households to its FiOS video service and its faster FiOS broadband Internet service. Unfortunately, this success has not been reflected in its financial returns.
From the very outset, Wall Street analysts were openly skeptical of Verizon's plans to deploy FiOS. For instance, Bernstein Research opined two years ago:
"There is little doubt but that Verizon's FiOS is a terrific product… for consumers. But for shareholders, the benefits are less clear. Against a weakening macro-economic backdrop, and deteriorating fundamentals for the Consumer Wireline segment at the major TelCos, FiOS will provide welcome growth… but it may actually dilute Wireline margins. And FiOS's growth will come at very high cost…
"Unfortunately, there is no simple answer to the all-important returns question. Considered from the beginning of the project (in 2005), we believe Verizon's enormous $23B investment in FiOS project will almost certainly fail to return Verizon's cost of capital – and by a wide margin."Footnote 34
Bernstein Research's conclusions derive in large part from its view that FiOS could not attract sufficient subscribers to generate cash flows that would offset the $23 billion in investment. Verizon has assumed the risk of this investment, hoping that it can attract sufficient subscribers to amortize its huge investment. But it did not begin to deploy FiOS until the U.S. regulatory picture had cleared, and it had thus been assured that the FCC would not require the unbundling of next-generation fiber. Thus far, however, Verizon's common equity price has lagged substantially behind those of the other major North American ILECs –AT&T, Bell Canada, and TELUS.Footnote 35
Given Verizon's struggles with the economics of its FiOS service, even with regulatory assurance that unbundling of fiber will not be mandated, investors will surely be reluctant to underwrite a Canadian ILEC's potential foray into fiber to the home if such an investment is encumbered with a requirement that it be unbundled and offered to competitors at cost-based rates. Such an investment will require substantial video (and triple-play) revenues from the households that value large numbers of high-definition video services and high broadband speeds. If entrants were able to target these high-valued households and pay the ILEC only a cost-based rental on the fiber, the economics of such an investment would be severely impaired.
Since the completion of the Aron-Crandall White Paper in 2008, additional empirical evidence that confirms our analysis has continued to accumulate. It is therefore worth repeating one of our most important conclusions from the 2008 White Paper:
"It is clearly possible for the CRTC to maintain its existing essential facilities policy towards existing network facilities while exempting new next generation networks from mandatory unbundling. This is precisely what the FCC has done in the United States, and the result has been remarkable. Immediately after the FCC announced its policy of exempting these new networks from wholesale unbundling requirements, Verizon and SBC launched massive new investments in fiber-to-the-home and fiber-to-the-node networks."Footnote 36
The Commission would be making a major mistake if it extended "speed-matching" unbundling requirements to high-speed broadband networks. By doing so, it would discourage network investment and gain little or nothing in return in terms of short-term competition in broadband services. Moreover, by discouraging investment in a key part of the ICT sector, the Commission's action would place downward pressure on productivity growth in Canada, thereby potentially reducing the future standard of living for Canadian citizens.
|Country||Original GUYRS||Alternative value||Justification for alternative value|
|Australia||6||1||Reflects the resistance by Telstra, and the high prices until the competion notice issued in March 2004 by ACCC, which reportedly led to lower prices.|
|Austria||7||5||Reflects that Austria Telekom failed to comply with the 1999 Order that seems to be the basis of the 7 year designation here; instead it was in 2001 in response to the EC directive that Telkom Austria did some more unbundling, and when UPC actually was able to enter using unbundling.|
|Belgium||5||0||Reflects continuing complaints, to this day, about absence of LLU; all entrants who are not facilities based, about 22% of the market, are resellers. This suggests no real competition on network management, which may explain the data cap anomaly.|
|Canada||9||5||Reflects initial sunset and high LLU rates, followed by the 2001 decision to extend LLU indefinitely and reduction in service charges.|
|Czech Rep.||2||2||No change.|
|France||5||4||Reflects delayed implementation between formal adoption in 2001 and actual implementation following EU action in 2002.|
|Germany||8||0||Reflects DT resistance and BnetzA lack of capacity|
|Greece||5||0||Reflects the fact that only in May of 2007 was LLU regulation changed, and a new framework put in place that seems to have resulted in a shift of some of the wholesale/carrier pre-selection model operators shifting to more unbundling|
|Hungary||4||4||Insufficient data to form an opinion|
|Iceland||5||0||Eirecom litigated extensively to delay the regulator, and succeeded in delaying implementation of unbundling throughout the relevant period.|
|Italy||5||5||Data is ambiguous about quality of implementation; insufficient data to argue that there was in fact no real LLU uptake|
|Japan||9||6||Reduces the period to reflect the weakness of MPT until the passage of the new requirements by MIC in 2000|
|South Korea||4||9||Reflects the fact that Thrunet entered over access to incumbent facilities—cable owned by Kepco.|
|Luxemburg||5||0||Justified by claim that incumbent effectively only offers an unregulated wholesale product to competitors, with no effective enforcement to the contrary|
|Netherlands||9||5||Reflects exclusion of competitors from naked DSL until 2001.|
|New Zealand||0||0||NZ changed only after 2005|
|Poland||3||0||Reflects the fact that Netia only enters unbundling in 2007, and Multimedia Polska still has had to get an ad hoc decision from the regulator as late as August of 2007 on its unbundling arrangement with incumbent TP Group. That a regulator still needs to rule on a negotiation is a good sign that LLU has not been implemented seriously.|
|Portugal||5||3||2002 citation by EU of Portugal as a particular site of lack of compliance with unbundling directives.|
|Slovak Rep.||0||0||Should remain 0. LLU never passed, only promised. Still not passed.|
|Spain||5||4||Reflects the April 2002 new reference offer reducing LLU rates by 30%.|
|UK||5||0||Reflects unobserved BT resistance that led to massive jump in unbundling loop usage after functional separation implemented in 2005|
|US||10||0||Reflects 2001-02 FCC decisions to shift to intermodal competition|
Source: Crandall, Ehrlich, and Eisenach (2009)
- 1 Non-resident Senior Fellow in Economic Studies, The Brookings Institution, Washington, DC. These comments reflect the view of the author and should not be attributed to the Brookings Institution, its Trustees, or its other staff members. (back to footnote reference 1)
- 2 I have not listed and do not address the Commission's item B. (c). (back to footnote reference 2)
- 3 Debra J. Aron and Robert W. Crandall, White Paper: Investment in Next Generation Networks and Wholesale Telecommunications Regulation (White Paper or Aron-Crandall White Paper), filed in March 2009 by TELUS in its Petition to the Governor in Council requesting that the Governor in Council vary Telecom Decision CRTC 2008-117 and rescind Telecom Order CRTC 2009-111. (back to footnote reference 3)
- 4 Telecom Decision CRTC 2008-117, Cybersurf Corp's application related to matching service speed requirements for wholesale Internet services, 11 December, 2008, § 22. (back to footnote reference 4)
- 5 Aron and Crandall, p. 42. (back to footnote reference 5)
- 6 As cited in fn. 4, supra. (back to footnote reference 6)
- 7 These and other, related conclusions may be found on pp. 41-43 of Aron and Crandall's White Paper. (back to footnote reference 7)
- 8 Aron and Crandall, Executive Summary, p. i. (back to footnote reference 8)
- 9 Aron and Crandall, Executive Summary, p. ii. (back to footnote reference 9)
- 10 Id. (back to footnote reference 10)
- 11 The de Ridder study was published slightly before our report, but we did not address it. A review of these and earlier studies may be found in Carlo Cambini and Yanyan Jiang, "Broadband investment and regulation: A literature review, Telecommunications Policy, 2009. (back to footnote reference 11)
- 12 Declaration of Robert W. Crandall, Everett M. Ehrlich and Jeffrey A. Eisenach Regarding the Berkman Center Study, in Federal Communications Commission, In the Matter of International Comparison and Consumer Survey Requirements in the Broadband Data Improvement Act, GN Docket No. 09-47; A National Broadband Plan for Our Future, GN Docket No. 09-51; and Deployment of Advanced Telecommunications Capability to All Americans in a Reasonable and Timely Fashion and Possible Steps to Accelerate Such Deployment Pursuant to Section 706 of the Telecommunications Act, GN Docket No. 09-137, 2009. (back to footnote reference 12)
- 13 Bouckaert, et.al., pp. 33-34. (back to footnote reference 13)
- 14 Id., p. 58. (back to footnote reference 14)
- 15 Id., p. 24. (back to footnote reference 15)
- 16 "It has to be highlighted though that the data used for the analysis does not cover investment in next generation access networks. To the extent that the investment in next generation networks is qualitatively different from upgrading the current infrastructure of incumbents, this result cannot be extrapolated to future investments." Id., p. 25. (back to footnote reference 16)
- 17 Telecom Decision CRTC 2008-117, §19 and §22. (back to footnote reference 17)
- 18 FCC, Report and Order and Notice of Further Proposed Rulemaking, CC Dockets Nos. 02-33, 95-20, 98-10; WP Docket Nos. 04-242, 05-271, August 5, 2005. (back to footnote reference 18)
- 19 FCC, International Comparison and Consumer Survey Requirements in the Broadband Data Improvement Act, GN Docket No. 09-47; A National Broadband Plan for Our Future, GN Docket No. 09-51; and Deployment of Advanced Telecommunications Capability to All Americans in a Reasonable and Timely Fashion and Possible Steps to Accelerate Such Deployment Pursuant to Section 706. (back to footnote reference 19)
- 20 Berkman Center for Internet and Society, Next Generation Connectivity: A review of broadband Internet transitions and policy from around the world, October 2009. (back to footnote reference 20)
- 21 The Berkman Center for Internet and Society, Next Generation Connectivity: A review of broadband Internet transitions and policy from around the world, October 2009, Draft. (back to footnote reference 21)
- 22 See fn.12, supra. (back to footnote reference 22)
- 23 Berkman Center Report, p. 85. (back to footnote reference 23)
- 24 Softbank, Annual Report 2009, p. 41. (back to footnote reference 24)
- 25 Daisuke Wakabayashi, KDDI Deal Sets up Telecom Skirmish in Japan, The Wall Street Journal, January 26, 2010, available at http://online.wsj.com/article/SB10001424052748703808904575024770373801704.html?mod=WSJ_latestheadlines. (back to footnote reference 25)
- 26 Id. (back to footnote reference 26)
- 27 John de Ridder, "Catching-up in Broadband – What Will it Take?" Organisation for Economic Co-Operation and Development (July 25, 2007). (back to footnote reference 27)
- 28 See Boyle, et.al., Table 2. See also the estimation of the model in Crandall, Ehrlich, and Eisenach, fn.6, supra. (back to footnote reference 28)
- 29 For further criticism of the Berkman Center Report's econometric analysis, see Crandall, Ehrlich, and Eisenach (2009) and Bronwyn Hall, Comments to the Federal Communications Commission, Washington D.C. on Broadband Study Conducted by the Berkman Center for Internet and Society, NPB Public Notice # 13; GN Docket Numbers 09-47; 09-51; 09-137, November 13 2009. (back to footnote reference 29)
- 30 http://blogs.wsj.com/digits/2009/12/29/fcc-eyes-average-internet-speeds-for-rural-areas. (back to footnote reference 30)
- 31 Viviane Reding, EC Commissioner, opined in a 2006 speech: "I believe that the policy option of structural separation could answer many competition problems that Europe's telecom markets are still facing today." See http://europa.eu/rapid/pressReleasesAction.do?reference=SPEECH/06/422. (back to footnote reference 31)
- 32 These data are available in the Broadband Scorecards on the ECTA web site, available at www.ectaportal.com. (back to footnote reference 32)
- 33 All data are from the companies' public financial reports. The EU-15 carriers are Telekom Austria, Belgacom, TELIA-Sonera, France Telecom, British Telecom, KPN, Deutsche Telekom, Telecom Italia, TDC, Telefonica, Portugal Telecom, and OTE. (back to footnote reference 33)
- 34 Bernstein Research, Verizon (VZ): Project FiOS… Great for Consumers, but What About Investors? January 14, 2008. (back to footnote reference 34)
- 35 For historical values of each company's common equity, see http://finance.yahoo.com/. (back to footnote reference 35)
If the following document is not accessible to you, please contact the person below for assistance in obtaining the documents in the appropriate format.
Industry Canada | Industrie Canada
300 Slater Street, Ottawa ON K1A 0C8 | 300, rue Slater, Ottawa ON K1A 0C8
Telephone | Téléphone 613-990-6456
Facsimile | Télécopieur 613-952-2718
Teletypewriter | Téléimprimeur 1-866-694-8389
This document is better viewed and printed in PDF format:
A Future Friendly Digital Economy Strategy: Submission to the Government of Canada's Digital Economy Consultation – Appendix 5 (PDF, 535 KB)
Information on Downloading a PDF Reader
To access the Portable Document Format (PDF) version you must have a PDF reader installed. If you do not already have such a reader, there are numerous PDF readers available for free download or for purchase on the Internet: