stLiberal Rules for a 21 Century
The U.S. Experience with Convergence
Thomas W. Hazlett
Professor of Law & Economics
George Mason University, USA
The obsolescence of traditional Postal Telephone and Telegraph (PTT) monopolies has forced
liberalization in the telecommunications sector. This transition challenges policy makers with two
visions of convergence, however, one implying the emergence of a new monopoly where diverse
applications are provided via a single high-capacity network, the other implying the growth of
diverse networks that encroach upon rivals. The former generally implies mandatory network
sharing regulations; the latter is associated with deregulation. The selection of these distinct policy
paths can now be informed by experience. Ironically, U.S. competition in fixed-line phone service
succeeded only after the sharing mandates of 1996-2004, designed to facilitate competitive entry,
were over-turned by federal courts.. Residential broadband markets yield further data suggesting
that deployments accelerate when open access requirements are eliminated. These experiments in
liberalization can inform and advance pro-consumer, technology-friendly policies in
Liberal Rules for Telecoms page 2 Thomas W. Hazlett
1. Convergence and the Transition to Competition
The devolution of the separate telecommunications silos of the 1970s is now a familiar story. As early as
1987, Peter Huber‟s masterful “Geodesic Network” framed the policy debate as one of adjustment to
convergence, removing artificial barriers that kept potentially rival networks quarantined in monopolistic
fiefdoms. Even then, a generation ago, the picture on the horizon was coming into focus: local voice access,
long distance transport, multi-channel video, and data were not productively supplied under “regulatory apartheid”.
What was a “vision” then is a reality today. Liberalization in telecoms has swept the globe; the PTT
monopolies of the 1980s are now largely museum exhibits. But the ways in which communications
markets have been opened differ across regimes. This can be said to reflect rival assumptions as to the full
meaning of the term “convergence.”
The “Convergence A” view is that the maturation of communications networks has produced a natural
competitive rivalry in the marketplace. What U.S. regulators called the “two wire to the home” policy in
the 1980s – one owned by the local telephone carrier, the other by the cable TV operator – spontaneously combusts. Given modern technologies and new business plans, head-to-head network rivalry emerges.
The phone wire expands to carry video; the cable wire expands to carry voice. What were two monopolies
in distinct product markets merges into a market with overlapping networks battling for market share in
either. Integration into broadband richens the rivalry and brings customers “triple play” (voice, video, data) options. The addition of wireless carriers extends the competitive continuum.
In this view of things, the role of the government is to unleash the market, eliminating barriers to naturally 1rivalrous behavior.
In the “Convergence B” framework, existing networks do not expand onto rivals‟ turf (Perhaps this owes, 2in part, to the stunting of their initial growth due to economic or regulatory constraints.) And now the day of the competitive entrant has passed, precisely because the “convergence” implies economies of scale:
services that once required their own delivery platform are now efficiently supplied via facilities supplying
a vector of services. Rival platforms are now replaced by a communications network with abundant
capacity to provide triple plays. Of course, organizing all communications within a single firm has costs as
well as benefits; the loss in market rivalry is typically met by regulatory interventions that seek to offset
some of these social costs. The resulting markets may be deemed “open,” in the sense that non-network application providers utilize the underlying facilities on “reasonable terms and conditions,” but basic network infrastructure is supplied on the regulated monopoly model. Convergence B delivers us not to
competitive networks but to the PTT model whence we came.
This paper does not undertake the task of global comparison, an important and ongoing element in 3discerning the optimal policy path. It sets its sights on more specific target, describing the policy path in
1 These include such competitive impediments as local government video franchises that deter telephone
carriers from entering cable TV markets. See Hazlett 2007, Cable TV Franchises as Barriers to
Competition, 12 VIRGINIA JOURNAL OF LAW & TECHNOLOGY 2. It also includes liberalization of spectrum
policy, providing more bandwidth for operators to expand services (e.g., into high-speed data) and to adopt
innovative technologies. See Hazlett 2008, Optimal Abolition of FCC Spectrum Allocation, JOURNAL OF ECONOMIC PERSPECTIVES (Winter). 2 In some countries, cable TV networks have been deterred by regulatory barriers. In others, cable TV has
not developed throughout the market due to the presence of abundant broadcast TV channels (e.g., Italy) or
the relatively high costs of constructions (e.g., the U.K.). 3 For contrasting analyses, see, e.g., OECD 2001, The Development of Broadband Access in OECD
Countries, Working Party on Telecommunications and Information Services Policies (Paris: Organization
for Economic Co-operation and Development; Oct. 29). And National Research Council 2002, Broadband: Bringing Home the Bits, Committee on Broadband Last Mile Technology, Computer Science and
Liberal Rules for Telecoms page 3 Thomas W. Hazlett
the United States where the “A” view has come to dominate – not exactly by strategic vision or formal consensus but by the rough-and-tumble reality of policies tried, failed, and then reformulated.
The short story – and a summary of the paper‟s main points – goes like this. The 1996 Telecommunications Act laid out a policy reform on the vision of “Convergence A,” but with a
schizophrenic twist: the transition to competitive networks would be managed by a mandatory network
sharing regime. This was a temporary expedient to assist the development of new networks, not a long-run
“B” view of convergence as leading to dominant platforms with sufficient capacity to efficiently re-monopolize the market.
The transitional regime, however, proved administratively complex and economically disappointing. It
encouraged extreme contentiousness in the drafting of network access (wholesale) pricing rules, and
deterred network development by both incumbents and entrants. It collapsed in mid-2004, overturned by
courts as violating the instructions in the 1996 Act to facilitate the construction of competitive systems.
But the demise of regime did not prove a set-back to competition in the marketplace. Inter-modal
competition is today supplying the competitive constraints that the network sharing regime, effectively
ended just four years ago, could not.
Cable television systems – having made local phone service available to under 15% of U.S. households in
2003, the product of seven years of investment since being awarded the right to supply voice services in the
1996 Act – today pass over 85% of U.S. households. Broadband competition between cable operators and
telephone carriers has also accelerated since FCC deregulation of DSL networks in the first quarter of 2003;
DSL subscribership was about 60% higher at year-end 2006 than it would have been under the pre-
deregulation growth rate. These marked gains in competitiveness are buttressed by trends in wireless
phone markets. Yet, by standards used to determine “effective competition” in other communications
markets, local phone services are today “effectively competitive” even disregarding mobile telephony as a
substitute product. The recent market trends, highly successful in realizing the stated goals of the 1996
Telecommunications Act, appear – at best – unrelated to the “policy-induced” rivalry supported via network sharing mandates. Indeed, the data suggest that the sharing mandates designed to provide
“stepping stones” to rivalrous communications networks deterred such deployments.
In sum, the U.S. experience offers lessons from policy experiments suggesting that convergence can deliver
market outcomes advancing consumer welfare. By an iterative process, the U.S. has discovered much st Century telecoms. This may be of interest about how to craft, and how not to craft, liberal policies for 21to policy makers elsewhere as a case study revealing how the Schumpeterian forces now shaking the
communications sector can most effectively be channeled.
2.Fixed Line Telephone Competition: the U.S. Experience
The central economic concern of the 1996 Telecommunications Act was the promotion of competition in
local telephone service. This was to be accomplished with a two-pronged policy attack. First, the Act 4eliminated entry barriers to new entrants in local telecommunications markets. Previously, state and local franchising laws had created legal prohibitions explicitly or implicitly limited competitive entry. Second,
to assist entrants who might emerge, the Act required that incumbent telephone companies:
5(a) interconnect with entrants‟ networks 6(b) provide customer connections, at wholesale prices, to new rivals to resell at retail prices 7(c) provide “unbundled” pieces of their network to alternative providers
Telecommunications Board, Division on Engineering and Physical Sciences, National Research Council
(Washington, D.C.: National Academy Press). 4 47 U.S.C. ?253(a). 5 47 U.S.C. ?251(c)(2). 6 47 U.S.C. ?251(c)(4). 7 47 U.S.C. ?251(c)(3).
Liberal Rules for Telecoms page 4 Thomas W. Hazlett
The basic interconnection obligation was not controversial. However, the mandates requiring incumbents
to share their networks proved highly contentious. In the use of “resale” or “UNEs” (unbundled network
elements), rates charged and terms of wholesale access were regulated by state regulatory commissions
under guidelines crafted by the Federal Communications Commission. These rules imposed a TELRIC
(Total Element Long-Run Incremental Cost) determination of wholesale network access pricing that sought
to approximate the costs of a new, efficiently-sized, state-of-the-art network. This tended to under-
compensate investors for the risks of technological changes , placing zero value on the options extinguished when capital was sunk to create infrastructure.
According to later decisions by U.S. courts, the regime tilted decidedly in favor of low access prices and
generous terms for competitive entrants leasing incumbents‟ networks. This, in turn, was found to bias the
“rent v. buy” decisions faced by such entrants, and to depress the incentives of incumbents to fend off
emerging rivals by investing in new, upgraded facilities. Federal courts found that this undermined the
plainly stated goals of the 1996 Act in promoting the creation of competitive telecommunications networks.
Mandates to share existing infrastructure as a “stepping stone” to facilities-based competition  were
called for in the Act, but this cause-and-effect sequence was undermined by an unbalanced approach that
sought only to promote resale of incumbents‟ services. As the D.C. Circuit Court of Appeals wrote, in overturning the network sharing rules, the Commission‟s “entire argument about expanding competition and investment boils down to the Commission‟s expression of its belief that in this area more unbundling is 8 better.” 9A March 2004 appellate court decision effectively ended this regime and the two major telephone carriers
reselling incumbents resale services, AT&T and MCI, were soon (in 2005) acquired by local exchange
carriers, SBC and Verizon. The experiment in “policy-induced competition”  was over, and by
consensus a failure . There was no stepping stone to facilities-based competition, but a stifling of
investment in new network facilities by both entrants and incumbents . While about 32 million lines
(out of about 180 million total fixed connections, or about 18% market share) were provided by
competitive entrants in June 2004, the resale model appeared to be crowding out entry via “competitor-
owned” networks, including (most importantly) those provided by cable TV operators. Unbundled 10Network Element-Platform (UNE-P) lines grew, on net, from under a million in Dec. 1999 to nearly 16 million in June 2004; cable CLEC lines grew much less (to under 4 million) in June 2004, and had virtually
no growth in the 2002-2004 period when UNE-P wholesale rates were declining and UNE-P lines were
growing most rapidly. See Fig. 1.
8 United States Telecom Association v. FCC, 290 F.3d 415. (D.C. Cir. 2002). 9 United States Telecom Association v. FCC, 359 F.3d 554 (D.C. Cir. 2004). Uncertainty about whether the decision would be appealed to the U.S. Supreme Court delayed a resolution of the policy even alter the
March 2004 verdict. In June 2004, however, both the FCC and the Bush Administration announced that
they had decided not to ask the highest court to overturn the decision and the policy was essentially settled. 10 UNE-P lines were supplied by incumbent local exchange carriers (ILECs), but sold to retail customers
by competitive entrants who paid wholesale prices determined by aggregating costs for each “network
element.” This resale model dominated entrants‟ choices (a “total service resale,” or TSR, model was also
available) because the price was relatively low. TSR wholesale prices (the price paid by the entrant to the
incumbent network) were discounted about 20% from retail levels; UNE-P prices were discounted, on
average, about 50%. See Hazlett (2006).
Liberal Rules for Telecoms page 5 Thomas W. Hazlett
IG. 1. UNE-P CLEC LINES V. CABLE CLEC LINES F
As of the policy regime shift in 2004, the lack of competition in local telephone service was particularly
troubling. When the Telecommunications Act passed, more than 95% of U.S. households had access to
two wires, and the Act permitted the second of these – owned by a cable TV operator delivering video – to
supply telephony. It was known that the costs of upgrading the cable network were relatively modest –
$400-$900 per household . A U.S. cable operator, Cox, would soon document that high-quality circuit-
switched phone service could be overlaid on an existing video network with capital costs of just $610 per
new circuit-switched telephone subscriber, with costs falling to $527 by 2004 (Cox 2004).
Upwards of $65 billion had been expended by CLECs, largely for marketing resold lines, only to see the
policy fail to promote new networks or to sustain the reseller CLECs that had emerged, which largely
perished in a wave of bankruptcies [7, p125]. Meanwhile, build-out of a competitive local phone network
was relatively inexpensive: at $610 per new phone customer and twenty-five percent penetration
(assumptions used in Cox 2003), the U.S. market could have „wired for competition‟ via fixed-line cable
11telephony for just $15 billion. Not only was this a small fraction of what was burned by CLECs that
ended in bankruptcy, it was far less than what cable TV operators themselves expended to upgrade systems
for digital video, a response to satellite TV competition that drove them to sink approximately $66.4 billion
in network upgrades, 1999-2003. These capital expenditures were more than twice the previous rate. See
Fig. 2. The cable industry was an aggressive investor in new technology that enabled delivery of digital
video and two-way Internet access at broadband speeds (i.e., cable modem service). But the opportunity to
invest incrementally more in telephony was not appealing.
11 This assumes 100 million U.S. households, the approximate size of the market in 1996. In 2007, there
were approximately 113 million U.S. households.
Liberal Rules for Telecoms page 6 Thomas W. Hazlett
12IG. 2. CABLE TV CAPITAL EXPENDITURES, 1996-2007 F
So, as the unbundling regime was collapsing, cable TV systems were rapidly expanding in scale and scope
– but not in cable telephone capacity or subscribership. The U.S. situation has been compared unfavorably
with that existing in the U.K. market where, for many years, cable telephony has been a popular service (in
the one-half of the national market where cable TV lines pass homes):
The United States provides a less compelling case for the importance of cable television
networks in promoting competition in local access. In 2003, cable television firms
supplied 2.5 million residential lines, whereas there were 73.783 million basic cable
customers and 102.9 million passed homes (Brito & Pereira 2007, p. 298).
Hence, the quandary: with the 1996 Act liberalizing rules such that cable operators were permitted to
compete, why so little cable telephony in 2003? Over those seven years, cable operators had elected to roll
out telephone service to only about 15% of U.S. households, garnering just 2.5 million subscribers out of
over 100 million homes passed by cable (for video service).
That was about to change dramatically, however. Cable telephone subscribership in 20073Q equaled 11.5
1314million households, a 350% increase over year-end 2003. In the third quarter of 2007, in fact, cable
telephone subscribers grew by 1.14 million, nearly one-half total net growth 1996-2003. More importantly,
perhaps, is the change in market coverage. By 20073Q, 86% of U.S. households were offered cable
telephony, nearly a 500% increase over 2003. While it took seven years to achieve availability in 16
million homes; it then took four years to extend cable telephony service to the next 77 million. See Fig. 3.
Whatever level of subscribership is obtained for the competitive entrant in voice service, the presence of
the second physical network will itself deliver the benefits of competition. To stem substitution into a rival
12 National Cable Telecom. Assoc., www.ncta.com/statistic/statistic/infrastructureexpenditure.aspx. 1313 Cable telephony subscriber and coverage data are from Leichtman Research. FCC data for 2007 are
not yet available. 14 It should be noted that cable telephone service is largely available to households, as video networks are
built-out in residential markets rather than in business districts. Recently, however, cable TV operators
have been extending networks into business markets in order to obtain high-speed data and telephony
subscribers from enterprise customers.
Liberal Rules for Telecoms page 7 Thomas W. Hazlett
product, incumbent carriers must lower prices and improve services. If one were to employ the standard
established for cable TV (multi-channel video) rivalry by the U.S. Congress in the 1992 Cable Act, an
interesting result is achieved. Under that standard, a market is deemed “effectively competitive” (and,
hence, deregulated) when a second provider offers service to more than 50% of total households in the
market, and serves (through subscriptions) at least 15%. In cable telephony, the first threshold was crossed
in 2006 (see Fig. 3), while the latter is surmounted by adding five million non-cable “CLEC owned” subscribers to the 20073Q cable fixed-line total. As of Dec. 31, 2006, the FCC reported 4.4 millon non-15 in addition, there are millions of residential VoIP customers who opt out of cable CLEC owned lines;16ILEC voice service via a cable modem subscription. In any event, with cable telephony subscriber
additions of over one million households per quarter for cable telephony, the 15% threshold – if not met by year-end 2007, which it likely was – will be easily passed by year-end 2008.
17FIG. 3: RESIDENTIAL CABLE TELEPHONY COVERAGE, 1996-20073Q
The pattern that emerges, then, is pronounced. Cable telephony grew very slowly after the opening of local
telecommunications markets in 1996. Given the new right to build-out video networks to compete in voice
services, cable TV operators largely demurred. This was despite massive investments made to upgrade
infrastructure, capital expenditures that dwarfed the scale of those needed to deliver local
telecommunications access via cable systems.
This suddenly changed, however, in 2004-2007. Cable operators began deploying telephone services at a
feverish pace, such that the U.S. is today essentially wired for two competing fixed-line telephone systems.
The “two wires to the home” – long in place -- now deliver converged voice services. The build-out clearly occurred in two stages: the sotte voce phase, 1996-2003; the second a vocal call-out, 2004-2007. Local
15 FCC, Local Telephone Competition: Status as of Dec. 31, 2006, Tables 3 and 5. 16 The largest stand-alone subscription VoIP service in the U.S. is Vonage, which reported 2.4 million
customers in 2007. W. David Gardner, Embattled Vonage Gets Subscribers To Stick With The Firm,
INFORMATIONWEEK (May 10, 2007); http://www.informationweek.com/howArticle.jhtml?
articleID=199500846. Most of these are likely to use cable modems for Internet access, given the market
share of cable and DSL in residential broadband markets. 17 Source: Leichtman Research Group, www.leichtmanresearch.com.
Liberal Rules for Telecoms page 8 Thomas W. Hazlett
fixed line competition, essentially „left for dead‟ in 2003, now appears to be a standard part of the
telecommunications marketplace. What accounts for the reversal?
Two explanations are available. The first explains the spurt in fixed-line telephony as a product of declining costs of supply, specifically those associated with the maturation of voice-over-Internet (VoIP) technologies. The second attributes cable telephony‟s recent expansion to policy reforms, specifically the 18 collapse of the network sharing mandates for local voice service.
The explanations are not mutually exclusive, and it is asserted here that both factors played important roles. VoIP had long been anticipated as a cost-reduction gift to cable operators seeking to supply voice services, and the 2003-2004 period saw important progress in VoIP applications used in cable and elsewhere. Global VoIP use increased sharply during this recent period.
Cable operators took full advantage of these marketplace trends and shifted from circuit switched to VoIP technologies in the services they offered their customers. White Papers produced by the most aggressive (early) cable entrant into voice, Cox, which was building out cable telephone services since 1997, documented the favorable cost trend. In 2003, Cox found that a circuit-switched cable telephony customer yielded positive net present value and relatively quick pay-back . The company urged other firms to expand into voice, arguing that the circuit-switched option then available was sufficiently profitable as to compensate for the option value in waiting for VoIP.
By 2004, Cox‟ strategic assessment had changed. It then calculated that VoIP could be provided for just
$267 per voice customer by cable operators, assuming the same penetration profile for a high-quality service that included an independent power supply (i.e., the phone works even when the household‟s
electric power is cut off, as often happens in a storm, e.g.). This, in the Cox analysis, contrasted favorably with a cost for circuit-switched telephony, which was estimated to have fallen to $527 per voice subscriber. The implication is that the decline in VoIP network costs did encourage cable build-out. Yet, the delays in serving voice markets were purchased at a cost by cable operators; profitable circuit-switched phone services were available to cable operators in the years before the VoIP cost decline, as Cox also 19strategized. The quick embrace of cable telephony in 2004 begs for further explanation. Incentives
created by the creation and elimination of the UNE-P program provide it.
While cable telephony networks were not directly regulated under network sharing mandates, the
competitive voice network was. As wholesale rates for ILEC access were politically determined, and sharply declining in the 1999-2003 period, when UNE-P lines rose rapidly, an overhang on cable telephony investment was imposed. The competitive entrant‟s investment is appropriated, if indirectly, by the same
regulatory obligations opposed by incumbent phone carriers.
That appropriation threat was dealt a fatal blow in the federal appellate ruling issued in March 2004, a death that was formally recorded with the Administration‟s decision not to appeal the verdict to the
Supreme Court in June 2004. Within weeks, the major UNE-P carrier (AT&T) announced its withdrawal from the local access markets . The ensuing three years resulted in the national build-out of a fixed-line telephone competitor. Technology and policy reform combined to (finally) produce this fortuitous result.
18 A third possible explanation is that market conditions improved about 2004, leading cable operators to invest more heavily in an expansion of their product mix. This can be rejected on the evidence that cable operators spent very heavily in the 1999-2003 period to digitize their systems and to expand bandwidth. These investments were undertaken to deliver more video channels and higher cable modem speeds. Cable telephony upgrades were notable by their absence, despite being relatively affordable according to assessments taken by cable operators at the time . 19 Cox (2003) estimated that a cable telephone subscriber would generate $600 in annual fees (local and long distance revenues) and that 35% of revenues would constitute gross profit. This implies that the average customer would generate $210 annually, suggesting that a delay of 1.5 years would essentially offset the infrastructure cost advantages reported for VoIP.
Liberal Rules for Telecoms page 9 Thomas W. Hazlett
3. Broadband Regulation: the U.S. Experience
The U.S. market also provides “convergence” evidence with respect to broadband. Both telephone and
cable TV networks pass nearly all U.S. homes, and so residential competition has been facilitated by
network upgrades to either – digital subscriber line (DSL) service layered onto traditional phone systems,
cable modem services supplied by cable TV operators. Conveniently for policy analysis, distinct
regulatory regimes have governed these broadband offerings, and these regimes have changed over time.
This allows us to compare and contrast, evaluating how the two approaches to convergence fare; the one
stressing policies of mandatory network sharing, and the other focusing on investment incentives for the
creation and advancement of separate networks.
3.1 Three Historical Regimes for DSL
3.1.A. Unregulated Cable v. Regulated DSL (until 20031Q)
Cable TV operators began offering cable modem services in 1995  and did so without any obligation to share their network infrastructure with other firms. Attempts to require cable operators to provide “open 20 A 1999 FCC report concluded that access regulation access” to independent ISPs were unsuccessful.21would risk deterring investment in the rapidly evolving market. The FCC later concluded that cable modem service should be classified as an interstate information service, and therefore be exempt from
common-carrier or open access obligations at both the federal and state/local level. The U.S. Supreme 22Court upheld the FCC‟s determination in June 2005.
In contrast to unregulated CM services, telephone company DSL services have historically been subject to
various regulatory obligations. When DSL was first offered in the 1990s, incumbent local exchange
carriers (ILECs) faced three major types of open access rules. First, telephone companies were required to 23provide the broadband transmission component of DSL services on a common-carrier basis. Second,
telephone companies were mandated to provide the copper loops used to provide DSL service on an 24unbundled basis. Third, the FCC‟s so-called line sharing rules required telephone companies to lease just 25the high-frequency portion of the loop (“HFPL”) used to provide DSL services. Regulators then set the price for the HFPL far below the price for an unbundled loop as a whole, substantially reducing Digital 26Competitive Local Exchange Carrier (dCLEC) costs.
3.1.B. Cable Unregulated/DSL Partially Deregulated (20031Q-20052Q)
20 The AOL/Time Warner merger, consummated in early 2000, imposed unique third party access
obligations. The provisions required the merged firm to offer AOL Broadband only after permitting two
independent ISPs to utilize Time Warner Cable infrastructure. The rules did not regulate wholesale prices,
nor did they regulate Time Warner‟s Road Runner broadband ISP. 21 Federal Communications Commission Cable Services Bureau, Broadband Today, A Staff Report to
William E. Kennard, Chairman, FCC (Oct. 1999);
http://www.fcc.gov/Bureaus/Cable/Reports/broadbandtoday.pdf. 22 National Cable & Telecommunications Assn. v. Brand X Internet Services, 545 U. S. 967 (2005). 23 Federal Communications Commission, In the Matter of Appropriate Framework for Broadband Access
to the Internet over Wireline Facilities, Report and Order and Notice of Proposed Rulemaking, CC Docket
No. 02-33, pp. 19-20 (Sep. 23, 2005). 24 47 C.F.R. ? 51.319(a)(1). 25 Federal Communications Commission, In the Matters of Deployment of Wireline Services Offering
Advanced Telecommunications Capability and Implementation of Local Competition Provisions of the
Telecommunications Act of 1996, Third Report and Order, CC Docket No. 98-147 (Dec. 9, 1999); http://www.fcc.gov/Bureaus/Common_Carrier/Orders/1999/fcc99355.txt. 26 47 C.F.R. ? 51.319(a)(1)(i). dCLECs are competitive local exchange carriers specializing in data
Liberal Rules for Telecoms page 10 Thomas W. Hazlett
27 As a result, dCLECs would have to pay for In February 2003, the FCC eliminated DSL line sharing rules.the entire local loop in order to supply DSL service to customers over the incumbent carrier‟s lines, or
strike a commercial agreement with the carrier to share a loop. The rationale for the reform was that, with
lessened network sharing obligations, telephone carriers would invest more heavily in bringing broadband
services to residential customers.
3.1.C. Unregulated Cable/Unregulated DSL (20053Q-present)
In August 2005, the FCC removed the remaining open access regulations for Internet connections bundled
with transport. With the Commission determining that DSL fell under Title I of the Communications Act,
broadband Internet access became treated as an “information service” exempt from common-carrier 28regulation. This put DSL services on regulatory parity with cable modem service.
3.2. Testing the Policies
Given this policy variation and the existence of rivalry between cable modem (CM) and DSL networks, we
can examine how broadband subscribership responds to changes in regulatory obligations. Network
sharing rules are tested on the evidence yielded by subscriber choices. Across three broadband regime switches, the theory that convergence requires continued network sharing regulation implies:
(1) pre-1Q2003: CM unregulated, DSL regulated with “line sharing”
Prediction: DSL subscribership will exceed CM subscribership.
(2) 1Q2003-4Q2006: DSL “line sharing” eliminated 1Q2003
Prediction: DSL subscriber growth will decline from trend.
(3) 3Q2005-4Q2006: DSL classified “information service” 3Q2005
Prediction: DSL subscriber growth will further decline from trend.
3.2.A. Growth of DSL vs. Cable Modem Prior to 1Q 2003
While DSL and cable modem technologies were developed at roughly the same time, unregulated cable
companies expanded the availability and penetration of their services much more quickly than regulated
telephone companies. By year-end 1999 cable dominated the emerging residential broadband market:
residential and small business DSL lines totaled just 0.29 million, while cable modem subscribers
numbered 1.40 million. Cable continued its dominance through year-end 2002, when it served 11.34
million, double the number of DSL lines (5.53 million). See Fig. 4.
27 FCC, FCC Adopts News Rules for Network Unbundling Obligations Of Incumbent Local Phone
Carriers (Feb. 20, 2003); http://hraunfoss.fcc.gov/edocs_public/attachmatch/DOC-231344A1.pdf. 28 Federal Communications Commission, FCC Eliminates Mandated Sharing Requirement on Incumbents’
Wireline Broadband Internet Access Services, Decision Places Telephone and Cable On Equal Footing,
Press Release (Aug. 5, 2005); http://hraunfoss.fcc.gov/edocs_public/attachmatch/DOC-260433A1.pdf.