On September 4th, in a speech to the startup-focused group, 1776, Chairman Wheeler gave a speech where he discussed consumer broadband deployment and competition. The Chairman seemed to be of two minds about the state of broadband competition.
On the one hand, the Chairman praised the valuable benefits that competition has yielded, in terms of spurring ISPs to deploy new, and upgrade old, networks in order to increase the speed and availability of broadband Internet to more Americans. The White House has previously recognized broadband competition as producing better networks and faster speeds.
However, after recognizing the value of these new last mile networks, Chairman Wheeler also concludes that the present state of competition is simply not adequate to ensure that consumers can realize all the benefits of these networks,
Looking across the broadband landscape, we can only conclude that, while competition has driven broadband deployment, it has not yet done so in a way that necessarily provides competitive choices for most Americans.
Chairman Wheeler began by introducing the graphic below. The chart shows, by percentage of households, the number of providers offering service at not only the FCC's currently-defined "broadband" speed (4Mbps down/1Mbps up), but also 3 additional, speed-defined, categories--10Mbps, 25Mbps, and 50Mbps.
The Chairman explained that 10Mbps was the minimum speed that a household would need to stream one HD movie, and allow for simultaneous Internet use from other devices. Wheeler also proposes changing the definition of "broadband" to 10Mbps downstream for purposes of participation in the Connect America fund.
The Chairman further argued for changing the definition of broadband, because "only wussies use less than 10Mbps/month, and the United States will not subsidize wussy Internet usage."  The 25Mbps and 50Mbps levels of service, Wheeler predicts, will quickly become the standards, as households continue their inexorable march toward dedicated, fully redundant, OCn SONET service.
Wheeler observes that, at the 4Mbps and 10Mbps tiers, most Americans have a choice of no more than 2 service providers. Moreover, the situation only deteriorates at higher speeds, "[a]t 25 Mbps, there is simply no competitive choice for most Americans." Speech at 4 (emphasis added).
The poor picture of broadband competition that the Chairman paints has created situations where "public policy" (read: FCC regulation) must intervene to protect consumers and "innovators" from firms with "unrestrained last mile market power." In these situations, he says, "rules of the road can provide guidance to all players and, by restraining future actions that would harm the public interest, incent more investment and more innovation." Speech at 5.
Title II Just Got Trickier
As most are aware, the FCC is currently evaluating public comments on its "rules-of-the road-for-broadband-ISPs" NPRM, in which the Commission is also considering whether to reclassify broadband Internet service as a "telecommunications service" under Title II. Supporters of reclassification often contend that it would not compel the FCC to impose any obligations on ISPs, beyond the general statutory duties of fair dealing imposed under Sections 201 and 202 of the Act.
Title II Is Different for Dominant Carriers. The obligations of any specific common carrier under Title II, however, depend on that carrier's classification within Title II for the relevant telecommunications service. Consistent with what Title II proponents argue, "non-dominant" carriers have few, if any, company-specific obligations.
On the other hand, carriers classified as "dominant" have to abide by additional obligations that stem from both the statute, as well as a more specific application of the general terms of the statute, because the FCC cannot assume compliance with its general statutory obligations for the dominant carrier service.
Thus, dominant carriers' rates can be regulated by the Commission, and they must file tariffs, subject to FCC review, describing their terms of service. Moreover, dominant carriers have longer review times and more stringent standards for initiating new, and retiring old, service offerings.
Finally, once imposed, dominant carrier regulations are all but impossible to get out from under. At the end of 2012, U.S. Telecom filed a Petition with the FCC, seeking to have its members (holding less than a 50% market share in a declining segment) declared "non-dominant" for voice service. The FCC has still not acted on U.S. Telecom's Petition.
Implications for Cable ISP's Higher Speed Services
The one thing that Chairman Wheeler could not have expressed more clearly is his belief that cable is the only alternative for broadband service at or above 25Mbps. Thus, if the Commission were to reclassify broadband Internet service as a telecommunications service, it would be difficult for the Chairman to explain why the incumbent cable providers are not dominant in the provision of higher speeds of consumer broadband service.
The specific Title II provisions, and Commission rules (such as the Computer Inquiry rules), that would apply to the cable companies' dominant telecommunications services would depend in large part on how the Commission chose to reverse the Cable Modem Order. Although, with respect to the Computer Inquiry rules, in particular, it seems highly unlikely that the Commission would revisit its earlier unwillingness to "in essence create an open access regime for cable Internet service applicable only to some operators." Order at � 46 (emphasis added).
Whither the Dominant Carrier ISP?
If we pause for even a second to consider the Commission's reasoning in refusing to apply Computer II to broadband over cable, it becomes very clear why Title II regulation is not the answer. The easiest way to avoid the incremental hassles that come with being "the firstest with the mostest" is don't be that guy.
What's that, dear broadband network? You could offer the fastest broadband on the market, but because some additional regulation intended to "simulate" competition means you'll earn less than you would on the "slower" speeds? Well, the easy answer would be: don't offer the higher speeds!
This almost seems like that classic case where rent control regulations have the paradoxical effect of creating artificial shortages for the regulated service, limiting access for the very people the laws were supposed to help. But, I'm sure that could never happen here...
In its Net Neutrality Comments, Free Press combines a limited number of less-than-ideal data points with a faulty methodology and a misleading narrative to claim that has "proven" its' reckless accusation that ISPs arelyingwhen they express concerns that Title II reclassification/regulation may distort their incentives to invest in network improvements.
In the previous post, we discussed some of the problems with the methodology, reasoning, and data Free Press uses to reach its conclusion. Today, we'll correct Free Press's misleading narrative "interpreting" the data with some relevant facts that you wouldn't know if you only read their comments.
Ironically, Free Press concludes its misleading presentation of capex "facts" (Comments III.B and III.C) by stating, "[w]e hope that the Commission and other policymakers learn and understand this history, for this debate cannot be a legitimate one if basic historical facts are replaced by incorrect beliefs." Comments at 111 (emphasis added). This statement would be OK (but still too preachy), if it didn't just present the FCC with a version of history so tailored for advocacy that it exists only in Free Press's comments. But, it's easy to forget . . .
Excessive Investment=Excess Capacity=Loss of Investment + Jobs
Free Press speaks of the period before the Cable Modem Order (in 2002) with a level of nostalgia that would seem more appropriate to a former WorldCom executive than a group claiming "historical facts." Free Press confidently asserts,
[common carriage], in conjunction with policies that opened up communications markets to greater competition, also was responsible for the largest period of telecommunications industry investment in U.S. history.
Comments at 90. The only hint from Free Press that this period may not have been an unqualified success is when Free Press allows that, "[m]uch of this investment . . . was a bubble ...." Comments at 111.
Perceived Bandwidth Demand Drove CapEx. Internet traffic grew at incredibly high rates in the second half of the 1990s, but the Internet was new to most people, and the subject of a lot of hype. Thus, perceived Internet traffic growth not only outpaced actual Internet traffic growth, but it was also disproportionately affecting perceptions of total bandwidth demand. But, where would people get these ideas?
Well, in a March 2000 report to Congress, then-FCC Chairman William Kennard stated,
Internet traffic is doubling every 100 days. The FCC's 'hands-off' policy towards the Internet has helped fuel this tremendous growth.
(emphasis added). Kennard's predecessor, Reed Hundt, would have none of this foolishness, and wanted people to know that "[i]n 1999 data traffic was doubling every 90 days." (emphasis added) ( Quote is from Hundt's self-congratulatory book, "You Say You Want a Revolution"at 224.)
The Reality. Not everyone at the FCC was buying (or selling?) the hype. A senior economist at the Commission, Douglas Galbi, published a paper the same year (2000), warning that total bandwidth demand was not as high as everyone seemed to think.
Growth of bandwidth in use for Internet traffic has been dramatic since 1995, butInternet bandwidth is only a small part of total bandwidth in use. . . .
(emphasis added). Meanwhile, massive fiber deployments and innovations in optical transmission equipment meant that capacity was about to explode.
The Reckoning. Only a year after Kennard's report to Congress, CNET reported that the U.S. was in the midst of a bandwidth glut, and that prices would likely decline much further. By summer 2001, the equipment companies issued clear warnings that the unraveling was well underway. A few months later, the Enron scandal would break.
Over the next year, what followed was the largest dislocation, in terms of job loss (500,000) and wealth destruction ($2 trillion) the telecom industry has ever seen. See, e.g., this BusinessWeek article. Law professor Dale Oesterle writes that the telecommunications industry in 2002 may have been the largest, most scandal-ridden, industrial meltdown in U.S. history.Here at 1.
The Aftermath. After the telecom bubble burst, depressed Internet transport prices would continue well into the middle of the decade. If you're wondering how low
In 2006, Level 3 needed additional transatlantic capacity, so it purchased 600Gbps of lit capacity on another carrier's transatlantic fiber. At the time of this purchase, though, Level 3 was carrying 480Gbps of traffic on its own transatlantic subsea cable system; a system that was scalable to 1.28Tbps. In other words, Level 3 already owned unlit transatlantic capacity, but using its own fiber didn't make sense because wholesale prices had dropped below operational and replacement costs!
The Biggest Lie About Capital Investment
The central deception of Free Press's entire misleading capex narrative is, of course, the notion that the 2002 Cable Modem Order was the defining event for broadband Internet capital investment. As explained above, the telecom bubble had little to do with Title II, and neither did the bust. Moreover, broadband Internet services, in particular, benefited more from the bust (post Title I classification), than they did from the boom.
The cheap [below-cost] Internet bandwidth of the early/mid-2000s led to a lot of web application experimentation and new Internet companies. Consumers responded quickly, and favorably, to the new, high bandwidth Internet applications, like Myspace. Xbox, and Youtube.
This led to strong consumer broadband Internet adoption, which could not have been possible if the broadband ISPs had under-invested in their networks. The FCC data show broadband Internet services increased by a factor of about 4.5 between 2002 and 2008; from 17 million customers in 2002 (see Table 3) vs. around 75 million telco and cable broadband customers in 2008 (see Table 1).
Indeed, this 400-500% increase in demand for broadband Internet service compares favorably with total bandwidth demand growth of around 300% during last half of the 1990s. See Galbi at Table 2. In fact, the success of the broadband Internet economy (Internet companies, backbones, metro fiber providers, and broadband ISPs) from 2002-2008 would finally end the bandwidth glut, and bring back demand for new "Title II" Internet transport capacity, including transatlantic capacity.
Free Press tries to prove that broadband ISPs are lying about their concerns with potential new, and undefined, rules under a Title II reclassification. But, if the FCC is tempted to change its regime based on erroneous cause-effect propositions that ignore historical facts, then it would seem the broadband ISPs have every reason to fear the unintended effects that will accompany a new regulatory classification.
It's no secret that Net Neutrality pressure group Free Press would like the FCC to revisit the 2002 Cable Modem Order, in which the FCC classified broadband Internet service over cable as an "information service." Nor is it a secret that the largest broadband ISPs oppose such a reclassification.
The ISPs often contend that a reclassification of broadband Internet service as a Title II, or "common carrier" service, would open the door to a range of regulations that could dampen or distort their incentives to invest in network improvements. But, in its comments on the FCC's Net Neutrality NPRM, Free Press intends to conclusively vanquish the "investment fear" arguments of the ISPs once and for all.
Free Press believes it can "debunk" the "myth" that Title II discourages regulated firms from investing in their networks if it can show that the broadband ISPs invested heavily in their networks at a time when the ISPs' broadband services were (pretty much) subject to Title II classification. Free Press relies on revenue and capital expenditures from the annual reports of a cross-section of large, publicly-traded, telecom and cable companies to tell the Commission a fairy tale.
Perhaps everything that could be wrong with Free Press's facts and theory about ISP network investment over the last 20 years is wrong--starting with the theory itself. This blog will focus on the problems with Free Press's theory, and its limited set of "facts" in support of its theory. Tomorrow, we'll explain what really happened (using Free Press's data, along with other relevant historical facts), and why Free Press's narrative is so misleading.
Investment Itself Is Never an Appropriate Regulatory Goal
Free Press seems to equate periods of rising capital investment as a "good" outcome, and periods of falling investments as a "bad" outcome. However, regardless of whether the investment was efficient or not (it isn't), the FCC should never try to assume the role of central economic planner. The FCC's only interest in investment should be to make sure that consumer interests are served in the manner that least distorts company investment incentives.
CapEx from Financial Statements Doesn't Show What Free Press Thinks It Does
Even if stimulating investment was the right focus for the Commission, the capex information Free Press presents does not prove that Title II is the answer. If Free Press is trying to show that the regulatory classification of consumer broadband service affects how much a firm invests in that service, then aggregate, firm-wide network investment wildly overstates mass-market broadband investment in any period.
In Fig. 1 (Comments p.100), Free Press tracks capex for a number of telecom carriers over time. But, by using aggregate enterprise capex, Free Press is primarily tracking capex for Title II services in all relevant periods. Notwithstanding the regulatory classification of one residential service, the majority of the revenue produced by these firms' networks still comes from Title II services (e.g., both AT&T and CenturyLink reported record numbers of residential broadband customers in 2Q 2014, but this service only comprised ~16.5% of total firm revenues for both firms).
A more accurate estimate of the capex devoted to the Title I service would focus on correlations between significant broadband subscriber growth and increased (decreased) capital investment over the same period of time. For example, CenturyLink has tripled its broadband subscribers (from ~2m to ~6m) over the last 5 years; during this same period, capex has grown at a CAGR of over 60%. See here (figures are from 2013 Annual Report, and the 2Q 2014 Earnings Supp. spreadsheet). A more careful review of the companies' data, however, may not support the story that Free Press wants to tell.
Investment and Revenue Figures from the Late '90s Are Not Entirely Accurate
Even if we accept that "investment" is a worthy regulatory goal, Free Press paints a misleadingly "rosy" picture of the era. Free Press concludes its recasting of the "golden age of investment under Title II" by simply stating that, "the 2001 recession and the economic impact of the September 11th attacks took their toll on the U.S. economy, and the telecom sector wasn't spared." (Comments at 101)
Free Press neglects to mention the devastating accounting scandals that would surface right after 9/11, or the massive layoffs, bankruptcies, and distress sales that would follow, and cascade through the industry over the next two years.
On October 16, 2001, Enron announced it would have to restate its earnings for the prior 2 years. This statement, and the subsequent SEC investigation, would uncover widespread accounting fraud throughout America's largest companies.
When you look at this list of the accounting scandals that were exposed in the 11 months after 9/11, don't focus solely on the telecom and cable companies. Keep in mind that every energy company on this list also owned significant telecom network assets. (See this 2002 study at p. 21/38).
Bandwidth trading. If you're wondering why most of the accounting scandals involved telecom or energy firms, that's because they had a common thread. Most of the telecom-related accounting fraud was related to "bandwidth trading." If you don't know what bandwidth trading is, just listen to Enron explain it.
The idea of bandwidth trading was just a few years ahead of its time. In practice, it would take BitTorrent and The Pirate Bay to make using someone else's capacity while they were sleeping a reality.
Early bandwidth traders, like the modern P2P thieves users, did not actually exchange money. Rather, if you had bandwidth on one route, and another company had capacity on a route you wanted, you could just swap capacity--but that's boring. Instead, each party would "pretend pay" the other for the prevailing value of the capacity (which still seems kind of dull).
The real fun came with the accounting. Both parties could record each other's pretend payment as real revenue, and record the capacity they were giving up as a capital expenditure; winning! For more, see this 2002 Wall Street Journal article. Oh, and when I say "could record," I mean literally; not legally.
As you can see, Free Press makes a number of mistakes in its attempt to prove that their Net Neutrality opponents could never justifiably fear Title II regulation--from trying to prove that a fear of undefined future regulations is unwarranted, to a misunderstanding of what their data actually show. Tomorrow, we'll explain what actually happened in the golden age of Title II and why Free Press's narrative is so deceptively misleading.