a X I‘VV-A .vulw—f (.r‘L .3
._ J..W:Vv I’M... 9. z

.34.... .
‘lclea-zi 5
I ‘y

.11. .

é...

4 . x
.nkflsfi
. z. x
.. , xii?! w.
. int: stMmflMuhwuuh
D . x.” x
I:

. a

.r L??? ve.‘ ‘
gigglimn ..x?L... ‘

{$44401 . Lanna “Pagan...
lunfnlli In...“

vi?“ \2
, .1 n at (4:.

:5 {loin

.1 . ..r.i.auto.§u.lf
Olglal - $X.o‘3.v‘ — a .i“
3...: x c. , .. It.» . 0
4 Amy. ”Karen 71% .
Wu»...

mm»! .5}. _

.. 926i
{3. .iflgfiflh.

«.

.. ..:)....:..t.
sfififlux I.

0501.:
1293.. 2:3

.1. 3... i!
M1 h... $138.3
. ..

 

 

111111111111111111111111111111111

DOC

 

 

 

 

This is to certify that the

dissertation entitled

A STRUCTURAL ANALYSIS OF OVERBUILD COMPETITION
AND CONVERGENCE IN THE MULTICHANNEL TELEVISION
INDUSTRY

presented by

HOEKYUN AHN

has been accepted towards fulfillment
of the requirements for

Ph.D. Mass Media

degree in

 

 

Barry R. Litmnmfim

Major professor

Date 8-— ‘ (fiam

MSU i: an Affirmative Action/Equal Opportunity Institution 0-12771

f:
L.

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

_ DATE DUE

DATE DUE

DATE DUE

 

1 280021101

 

NOV 1 8 2003

 

flMTiiwi

 

 

11111.3 1 A 20115

 

 

 

 

 

 

 

 

 

mm mm.“

 

A STRUCTURAL ANALYSIS OF OVERBUILD COMPETITION AND
CONVERGENCE IN THE MULTICHANNEL TELEVISION INDUSTRY

By

Hoekyun Ahn

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Communication Arts and Sciences — Mass Media

2000

ABSTRACT

A STRUCTURAL ANALYSIS OF OVERBUILD COMPETITION AND
CONVERGENCE IN THE MULTICHANNEL TELEVISION INDUSTRY

By

Hoekyun Ahn

Subsequent to the deregulation of the cable television industry in 1996, a
number of questions have begun to surface concerning the extent of competition,
the pricing behavior of cable systems, the feasibility of competition in cable
television in the long term, the existence of economies of scope, implications of
open entry for convergence, and ultimately long term prospects of market
structure in telecommunications. Employing a structural analysis of competition
and convergence, the present research provides the first systematic study of
overbuild competition in the multichannel television industry to answer these
questions.

The overbuilding business has always been risky. Changes in technology and
public policy notwithstanding, significant obstacles to competition remain today.
Most importantly, sharp price reductions in overbuild markets significantly
influence entry decisions in the multichannel television industry. Market predation
compresses profit margins for the incumbent as well as the entrant. Large MSOs,
however, may still enjoy substantial cost advantages and nonnegative profits
because of lower input costs, associated with both programming and operating.

Thus, an entrant suffers “double squeeze” where predatory pricing pushes post-

entry price down from above, while higher input costs further squeeze profit
margins for the entrant from below. Moreover, there are economic forces
associated with convergence that may further impede entry into the multichannel
television industry. If an entrant is compelled to provide a bundle of services, i.e.,
not only video but also data and voice, to compete with the incumbents product
space, convergence would mean extra capital requirement for the entrant.
Together, the combination of these factors may reduce competition, rather than
strengthen it in the multichannel television industry in the near future.
Considering all these factors, overbuild competition may not materialize as
expected. Large-scale competition between the two monopolies will take place in
the high-speed lntemet access service first, rather than in the multichannel
television market. Telcos are more likely to focus on data and voice services.
Thanks to the extreme difficulties in the overbuild business and cable industry’s
early lead in the deployment of broadband services, cable operators will be ready
more quickly than telcos to provide a “bundle of integrated services”. The “joint
dominance” in the video and high-speed data markets will be one of the critical
factors that confers cable MSOs strategic advantages in the competitive battle in
telecommunication. To the extent that consumers would like to get their
telecommunications services from a single provider. cable operators will have
significant strategic advantages over their telephone competitors in the

increasingly converging telecommunications industry.

Copyright by
Hoekyun Ahn
2000

To Hyunsook

ACKNOWLEDGMENTS

The author is especially appreciative of the encouragement, support and
generosity provided by Dr. Barry Litman, committee chair and director of this
dissertation. His guidance during the entire Ph.D. program made the successful
completion of this degree possible.

Special thanks is also expressed for the dedication and advice given by
Dr. Thomas Baldwin. He provided the author with enthusiastic effort and support.
It was a great fortune for the author to have him as academic advisor for a
master’s degree and as a member of the guidance committee for this
dissertation.

The author would like to thank Drs. Charles Steinfield and Bonnie Reece
for their advice and academic expertise. Thanks also to Dr. Charles Salmon,
Director of the Mass Media Ph.D. program, and Dr. Bradley Greenberg,
University Distinguished Professor and genuine scholar, for their support.

This dissertation would not have been possible without the cooperation of
many cable system managers and staff across the country. Their willingness to
be interviewed, and to provide information is greatly appreciated.

Friends provide the emotional strength and encouragement to persevere
through doctoral studies: several of mine are integrally a part of my doctoral
experience and this document: Drs. Hyun Oh Yoo, Joonho Do, and Seonghoon
Jeon for sincere friendship and concern; Jacqui for being my officemate and

thoughtful advisor; Minsuk for being a dedicated colleague; Denise and Vanessa

for secretarial help and much more. Thanks should also go to my colleagues in
the Department of Telecommunication for their support and for being there.

To my life-time guidance committee of two—my parents—many thanks.
No amount of formal education or advanced degrees can match their wisdom,
understanding, support, and love. My wife, Hyunsook, has always supported and
encouraged me with faith and love. Thanks for her endless patience. My
daughters, Jiyong and Jiwon, supported me with love that made the completion

of this dissertation possible. For that and so much more, thanks.

vii

Chapter

TABLE OF CONTENTS

Title
INTRODUCTION

Research Questions
Organization of the Present Research

THE DEVELOPMENT OF OVERBUILD
COMPETITION IN THE MULTICHANNEL
TELEVISION INDUSTRY

From Regulation to Competition
New Entry into the Video Market

OVERBUILD COMPETITION AND MARKET
PREDATION

Search for Competition in Cable Television:
Theory and Evidence

Theories of Market Predation

Description of Methodology and Model

Description of the Data

Descriptive Analysis

Econometric Results

CONSOLIDATION AND CONCENTRATION OF
THE MULTICHANNEL TELEVISION INDUSTRY

Horizontal Concentration in the Multichannel
Television Industry

Consolidation of the Regional Markets

Vertical Integration in the Multichannel Industry

Concentration, Consolidation, and Competition

CONVERGENCE AND COMPETITION

Structural Effects of Convergence on Competition

Current Development of Service Convergence

Comparison of Broadband Services: Technology
and Economics

The Strategic Use of Convergence

viii

Page

10

1O
14

38

38

56
76
89
92
93

106

106
117
125
127
146
147
157
174

183

6 SUSTAINABILITY OF OVERBUILD COMPETITION

Financial Performance of Overbuild Competitors
in Simulation Studies

ls Overbuild Competition Sustainable?

Are New Markets Big Enough to Sunnount
Double Squeeze?

The Success of 038 and Its Impact on Overbuild
Entry

Survivor Analysis

Analysis of Tobin’s q

7 CONCLUSION
Prospects of Competition in the Multichannel
Television Industry
Issues and Policy Recommendations

Bibliography

ix

206

211
214

219

225
229

236
236
245

258

LIST OF TABLES

Table 1 Average Monthly Rates of Competitive and Noncompetitive Groups

Table 2 Variable Definitions

Table 3 Descriptive Statistics of the Sample
Table 4 Estimates of Price Equations

Table 5 Estimates of Quality Equations

Table 6 Assessment of Competing Technologies

Table 7 Concentration in the National Market for the Purchase of Video
Programming: 1996-1999

Table 8 Number and Subscriber Size of Major Cable System Clusters
Table 9 Cable Modem Customer Rankings

Table 10 Classes of DSL Technologies

Table 11 Number of DSL-Ready Lines

Table 12 Technology Deployment & Costs to Residential Consumers
Table 13 Changes of Market Structure From 1990 To 1999

Table 14 Market Valuation of Cable Systems

Table 15 Estimates of Tobin’s q for the period of 1993-1999

78
93
97
1 02

110

113

118
163

1 70
1 73
1 77
228
230
231

LIST OF FIGURES

Figure 1 Breakdown of Online Universe 194

xi

Chapter 1

INTRODUCTION

Research Questions

The last mile to the residential consumer, historically served by telephone
and cable television incumbents, has generally been the least competitive and
most bandwidth-constrained part of the communications network. Compared to a
long distance line, a facility serving the last mile is likely to be expensive on a per
customer basis to deploy, to have linear rather than exponential growth in use,
and to require high maintenance per user. In addition, existing telephone and
cable facilities were largely paid for in past decades, when construction costs
were lower. They were also paid for by captive ratepayers, under regulatory
protection from competition and/or inherent economic conditions that conferred a
de facto monopoly and ensured recovery of costs, however slowly. The
incumbent local exchange carrier (LEC) in each area is also financially strong
and well staffed. These factors, among others, have combined to make entry
against telephone and cable incumbents very difficult. This has denied
residential consumers all the benefits of competition, which lowered prices,
increased choices and usage, and sped up innovation in telephones, long
distance, and mobile services over the past decades (FCC, 1999a).

Rising public dissatisfaction with cable service rate hikes and the allegedly
poor quality of customer service led to the passage of the Cable Television

Consumer Protection and Competition Act in 1992. But prior studies show that

rate regulation becomes substantially less viable as an instrument to protect
consumers from the monopoly market power as the complexity of the regulated
product increases. The Cable Act of 1992 was eventually superseded by the
Telecommunications Act of 1996, which mandated a sunset of the complex and
controversial benchmark price regulation and paved the way for potential entry of
telephone companies into the cable television business. Over the last decade,
Congress and the FCC tried to constrain cable rates, failed, and then
deregulated.

A number of approaches to curing the “government failure” have been
proposed, but the most popular and widely accepted approach has been the
repeal of the cross-ownership bans that have prohibited LECs since the early
1970s (see Hazlett, 1995b; NTIA, 1988, 1991). This policy would supposedly
introduce the new breeze of competition by LECs in the multichannel television
distribution market. Certainly, allowing telephone companies to provide cable
service would raise the prospect of another effective provider of multichannel
video programming to television viewers, thus offering direct competition to
incumbent cable system operators. Furthermore, the opportunity to provide
programming should increase telephone companies' incentives to construct
integrated broadband network facilities to carry that programming to the home. In
turn, these facilities could make available a vast array of voice, data, and video
services to subscribers. Indeed, there are clear public interest benefits in
permitting local exchange carriers (LECs) to provide or package video

programming themselves over their own facilities, so long as they also offer

video services under FCC safeguards designed to prevent discrimination against
competitors and cross-subsidization that could harm ratepayers.

In spite of substantial evidence of welfare enhancement from competition
presented in the prior studies, one must distinguish desirability from feasibility. It
is one thing that competition is desirable. It is another whether or not overbuild
competition in cable television is sustainable over a dynamic long run. The long-
tenn sustainability of overbuild competition would depend not only on the
demand and cost conditions for cable services, but also on the possibility of
product differentiation and the strategic interaction between competitors in the
marketplace (Smiley, 1990). Even though legal barriers to entry in the cable
industry have considerably diminished in recent years, other structural entry
barriers still remain formidable. Worse yet, even after entry does occur, first-
mover advantages and strategic behavior of the incumbent may suffocate the
prospects of overbuild competition in cable television, ultimately impeding future
entry.

In the early 1990s when LECs were not allowed to own and operate cable
services in their own telephone service territory, “video dialtone” was the
technological rage. But it went nowhere. Under the “video dial tone” concept, an
LEC could construct, operate, and maintain a transport facility within its local
service area. The channel capacity of the transport would be leased to
unaffiliated programmers on a common carrier basis. Lease rates for channel
capacity would be regulated to ensure that they were reasonable and

nondiscriminatory (NTIA, 1991). This concept of video dialtone was critically

flawed in that it failed to envisage the fundamental mechanism of business
operation in the capitalist society. Unless an appropriate level of economic
incentives and profitability is ensured, however attractive an idea might be, it
would not materialize in the real world.

The incidence of overbuild competition in the US. cable television
markets has been quite limited. Historically, a small number of communities have
issued multiple franchises for cable television service, leading to overbuild
competition between operators in the same market. In 1983, only 14 of
approximately 5,600 cable operators faced competition over a substantial portion
of their service territory. By 1989, the number of competing systems totaled 93
out of 9,050 systems nationwide. (see Television and Cable Factbook 1983,
1989). With the enactment of the 1996 Telecommunications Act, however, the
landscape of the cable industry has changed substantially in anticipation of LECs
entry into the video programming delivery market. To some degree, the hope (or
hype) has been realized in some markets across the country, but it is far from
what policy makers, economists, and the public have expected.

Thus, it is imperative to keep in mind that an idealist view of competition
would not produce meaningful consequences in reality. Subsequent to the
deregulation of the cable television industry in 1996, a number of questions have
begun to surface concerning the extent of competition, the pricing behavior of
cable systems, the feasibility of competition in cable television in the long term,
the existence of economies of scope, implications of open entry for convergence,

and ultimately long term prospects of market structure in telecommunications.

In order to understand the nature of competition currently in place in the
telecommunication industries and the evolutionary path to telecommunication
convergence, it seems necessary to review key economic models of industrial
organization that explain dynamic interaction between incumbent firms and
new/potential entrants. Key questions involving competition and convergence in
telecommunication include: (1) what kind of competition will there be? (2) how
will entry be affected by an incumbent’s market position and entry deterrence
strategies? and (3) to what extent will entry into the multichannel television
industry be viable considering competitive responses by incumbent firms and
strategic advantages available to them?

It might be too early to judge the prospects of cable competition. Yet, it is
possible to project the future market structure and performance of the cable
industry if the observation of overbuild competition currently in place is to provide
any helpful guidance to the understanding of factors that might influence entry
decisions. Employing a structural view of competition and convergence, the
present research provides one of the early systematic studies to answer these

questions.

Organization of the Present Research

The present research is organized as follows. Chapter 2 presents the
development of overbuild competition. Of all of the competitors that have
emerged to challenge incumbent cable operators, perhaps none is making such

a serious run as Ameritech Corp., the Midwest regional phone company. But as

far as specific incursions into cable's customer base, the impact is far from clear.
With the exception of Ameritech and SNET, no LECs have made significant
large-scale facilities-based entry into the video industry. Even the relatively
visible entry by the two LECs is currently under serious review. More prominent
are non-LEC affiliated firms such as RCN and Knology. These relatively
unknown firms aggressively pursue the business opportunities in selected
markets around country with not only video but also voice and broadband
services.

In Chapter 3, an empirical analysis of overbuild entry is presented. The
primary goal of Chapter 3 is to investigate competitive effects in cable price and
service quality dimensions. Chapter 3 begins with the debates of the efficacy of
two distinct policy options that have been adopted and are in place in order to
ensure high quality and reasonable price of the multichannel television industry.
The discussion of regulation and competition is followed by a literature review of
previous studies on competitive effects of overbuild competition in the
multichannel television market.

After presenting the famous debate of predatory pricing, first triggered by
McGee (1958), the applicability of market predation to the cable industry is
discussed. A number of factors associated with the technological and business
operation of cable television systems were found to render predatory pricing
rational, plausible, and profitable when overbuild competition occurs in a local
market.

The effects on the competitive communities and their incumbent operators

can be difficult to assess. To be sure, the presence of competition has caused
operators to sharpen and improve their offerings, to launch new services and
perhaps to speed up plant rebuilds. But the primary focus of the present
research is how the market responds to competitive overbuild and how the
reaction may influence the entry rate in the multichannel television industry.

A regression analysis of 247 local cable systems provides evidence of
welfare-enhancing effects that overbuild competition can generate in two
dimensions of cable service: price and quality. The inclusion of competitive
markets and municipal cable systems enabled the research to figure out how
incumbent cable systems would respond to competitive entry and how post-entry
price levels might influence entry decisions, by comparing the price charged by
duopoly systems to that of city-owned cable systems. The analysis also
attempted to investigate the economic effects of multiproduct offerings as well as
“potential competition.“

Chapter 4 addresses the implications of horizontal concentration and
consolidation in the cable industry for the vigor of overbuild competition. Three
ownership issues characterize the cable industry: horizontal concentration and
relationships among cable operators; vertical relationships between system
operators and program suppliers; and clustered cable systems, whereby cable
operators consolidate ownership within geographical areas. More specifically,
this chapter examines the status of ownership concentration in markets for the

delivery of video programming and discusses how it can erect barriers to

competition, affecting the competitive prospects and viability of overbuilds in the
multichannel television industry.

In essence, horizontal concentration and the accompanying monopsony
power of large MSOs will have a detrimental effect on viability of overbuild
competition. However, the financial viability of overbuild entry also depend on
other factors such as changes in demand, technological breakthroughs, and
potential for market expansion.

Chapter 5 deals with the economic consequences of convergence in the
telecommunication industry. Chapter 5 begins with the economic implications
and effects of convergence on competition. The economic implications of
convergence is analyzed in an attempt to provide explanations as to how it gives
rise to a significant shift in economic perspectives of telecommunication services
and how the new development in the telecommunication technology may
influence the evolution of market structure in the future. The opportunities for
changing one’s market position are much more limited in a mature industry. The
early history of an industry is of interest because it determines the structure of
the mature industry and, as a consequence, its long-run performance. In
addition, the evolutionary process determines the returns to a variety of kinds of
investment, including those in research and development, which have
implications for the dynamic aspects of the performance of the economy.

Chapter 5 also presents the current state of integration of voice, data, and
video services, followed by an analysis of the significance of broadband services

in shaping the future of the telecommunication industry. The emergence of the

high-speed lntemet access service market is viewed as the bridge of
telecommunication convergence, connecting video and voice services that have
historically been separated. The chapter also presents the discussion of how
firms can take strategic advantage of their multiproduct operations in dealing with
increasing threats of competition in the market. In essence, convergence will
confer cable MSOs three main strategic advantages: bundling capability, first-
mover advantages, and dynamic resource allocation. Chapter 5 concludes with
the discussion of why and how cable MSOs will dominate the converging
telecommunications market from each of these strategic perspectives.

Based on the major findings of the preceding three chapters, Chapter 6
discusses the sustainability of overbuild competition. In addition to market
predation and first mover advantages, the feasibility and sustainability of
overbuild competition is further discussed in conjunction with the impact of DBS
senrice on the long term prospect of overbuild entry. The assessment will be
checked against the findings of two separate analyses, that is, likelihood of
overbuild survival and Tobin’s q-ratios.

Finally, Chapter 7 summarizes the main findings of the present research.
Key policy issues and recommendations are presented in conclusion. They
include, among others, discussions of regulation of price and service, program
access issues, ownership limits, and cross subsidization in the multichannel

television industry.

Chapter 2

THE DEVELOPMENT OF OVERBUILD COMPETITION IN MULTICHANNEL
TELEVISION INDUSTRY

Regulations of Entry and the Telecommunication Act of 1996

Prior to the Telecommunications Act of 1996, local exchange carriers
(LECs) were prohibited from providing video program services to subscribers
located within their designated service areas. The Telecommunications Act of
1996 radically changed the market structures of the traditional electronic media
and telecommunications industries. It deregulated the barriers to entry between
both the cable and telephone industries and allowed them to enter each other's
service areas.

Since the passage of the Telecommunications Act of 1996, the major
players in the telephone and cable industries, regional Bell operating companies
(RBOCs) and multiple system operators (MSOs), have been expected to invade
each other's markets. The result would change the historically distinct markets

into a single integrated multimedia industry.

Modified Final Judgement and the Cable Act of 1984

Historically, telephone companies have been barred from the television
business due to their proclivity toward anticompetitive conduct. This was one of
the main reasons behind the 1984 AT&T breakup. Regulations designed to
protect against telephone company anticompetitive conduct often fail because

local telephone companies (telcos) are so complex and characterized by such

10

joint and common costs that no set of behavioral regulations could realistically
prevent competitive abuses. Because discrimination against competitors and
cross-subsidization are very difficult to detect, prevent, and rectify through the
behavioral regulations, telcos’ entry into video programming has been blocked
through such structural regulations as line-of-business restrictions and/or cross-
ownership ban.

Federal restrictions on telephone company entry into video programming
stem from three sources: FCC regulations, the 1984 Cable Act, and the AT&T
Consent Decree. In 1970, the FCC adopted rules barring all telcos from
providing video programming to subscribers in their local sen/ice area, either
directly or indirectly through an affiliate. When these rules were adopted, cable
was in its infancy. At the time, the FCC was concerned that telcos would, absent
such restrictions, be able to engage in improper cross-subsidization, block the
development of new broadband cable services, and use their control of
telephone poles and conduit space to prevent or hinder competition from
independent cable companies (NTIA, 1988).

When Congress enacted the Cable Act in 1984, it codified the FCC’s
cable-telco crossownership ban. The Cable Act states “it shall be unlawful for
any common carrier...to provide video programming directly to subscribers in its
telephone service area.“ These Cable Act provisions generally codified an FCC
rule implemented in 1970.

In addition, for a number of years, the BOCs were precluded from

diversifying into cable by the terms of the AT&T consent decree. Among other

11

things, that decree barred the BOCs from providing “information services," which
were universally assumed to include cable service. This restriction was broader
than that contained in the FCC’s rules or the 1984 Cable Act, as it barred the
BOCs from entering the cable business anywhere in the country, not just within
their own service area. However, the court presiding the consent decree lifted
this restriction in 1991 so the BOCs now were allowed to own and operate cable

systems located outside of their local service areas.

LECs’ Challenge to Cross-ownership Bans

These legal‘barriers appeared to be disintegrating since the early 1990s.
Some local authorities see little reason to protect unregulated monopolies, and
were quick to issue competing franchises. Even more important, however, are
the federal courts which have declared franchise monopolies to be
unconstitutional restrictions on the free speech/free press rights of these cable
competitors which have been excluded. The clear trend is for increasing market
rivalry within the industry.

In 1994, in Chesapeake & Potomac Telephone Co. of Virginia v. United
States (1994), the Fourth Circuit court ruled that the telco-cable cross-ownership
ban imposed by Section 533(b) of the 1984 Cable Act violates the First
Amendment rights of the telephone companies. The appeals court upheld a
federal district court’s decision that the telco-cable crossownership ban violates
Bell Atlantic’s First Amendment rights to free speech, permitting the regional Bell

Operating Company (RBOC) to provide video programming in its own service

12

area. The Ninth Circuit Court of Appeals confirmed the decision in the Bell
Atlantic case in favor of another RBOC, US. West (U. S. West Inc. v. United
States, 1994). Since then, three more federal district courts in other states also
have given the same rights to the RBOCs including Bell South (Be/[south Corp.
v. United States, 1994), NYNEX (NYNEX Corp. v. United States, 1994), and
Ameritech (Ameritech Corp. v. United States, 1994).

A year before these decisions, a federal district court in The Chesapeake
and Potomac Telephone Co. v. U. S. (1993) held that Section 613(b) of the
Communication Act, as amended, 47 U.S.C.A. §533(b), violated the First
Amendment because it was not tailored to serve significant government interest
in promoting competition in the cable television market. This telco-cable cross-
ownership ban prohibits a telephone company from directly providing video
programming to subscribers “in its service area.“ Significantly, the statute has not
been interpreted to prohibit a local telephone company from providing video
transport services. Thus, a telephone company is permitted to run a cable into a
subscriber’s home and then to lease channels of communications on that cable
to unaffiliated entities, such as cable operators. According to this “video dialtone”
concept, the telephone company only runs afoul of the statute by “exercising
control or discretion“ over the programming transported over its facilities. The
court found that §533(b) directly abridges the telephone companies’ right to
express ideas by means of a particular, and significant, mode of

communication—video programming.

13

The Telecommunications Act of 1996

The Telecommunications Act of 1996 changed the landscape of the
telecommunications industry by introducing open entry policy. Most significant
and relevant provisions in the Act include:

o Repealing cable-telco crossownership prohibition so that telcos could offer cable services
in their service territory

0 Prohibiting telco buyouts of cable systems in their service area except in “very limited
circumstances“

0 Allowing telcos to offer video services through a separate subsidiary

Under these conditions, telcos may choose to be regulated as cable
system, common carrier or newly created “open video system.” The Act also
preempted state and local regulations barring cable operators and others from
providing local telephone services. LECs are required to negotiate with new
telephone entrants for interconnection, number portability, dialing parity, access
to rights-of-way, and reciprocal compensation. Thus, the telecommunications

market is opened to competition from each monopoly from now on.

New Ent_ry into the Video Market

Since the passage of the 1996 Telecommunications Act, some industry
insiders claim that direct competition between cable systems in the same
geographical area is the new reality. The current estimate of overbuilds rivalry
finds more than 100 US cable markets. This competition has recently been
intensified by a number of factors: deregulation of cable prices and services,
vanishing virgin cable territory, lucrative cable investment opportunities, the

superiority of newer video technology, the ending of exclusive monopoly

14

franchises by the courts, and new legislation. This section describes how LECs
and others responded to such environmental changes surrounding the

multichannel television industry.

Ameritech New Media (ANM)

Nearly four years after entering the cable business as an aggressive
overbuilder, Ameritech has acquired local franchises in four major Midwestern
markets-Chicago, Cleveland, Columbus and Detroit, and extended service to
more surrounding communities. Alone among the Baby Bells, Ameritech is
broadly fighting cable on cable's battlefield, constructing expensive 750-MHz,
hybrid fiber-coaxial (HFC) systems in large, densely populated urban and
suburban areas.

ANM has been building major clusters, at least one in each of the 5
midwest states, where it provides telephone service. Its concerted franchising
efforts have brought Ameritech's cable arm a total of 107 cable franchises
spanning l.5 million homes in Michigan, Illinois and Ohio. It is expected to be
among the top 30 MSOs in the country in the year 2000 (Applebaum, 1999). As
of November 1999, Ameritech New Media is one of only a handful of cable
companies in the country challenging existing cable operators. Approximately
250,000 customers are watching Ameritech's cable TV service americast in
Chicago, Detroit, Cleveland and Columbus, OH, and their satellite cities and

suburban towns.

15

Detroit is ANM's largest cluster, with 600,000 households and 1.5 million
residents under the franchise, followed by the Columbus, Ohio, area, where it
has 530,000 households and 1.15 million residents under contract. According to
ANM, 34 out of 40 communities under franchise in the Detroit suburbs are up
and running (Multichannel News, Jan. 27, 1999).

Ameritech also services Area 5 in metro Chicago, where 500,000
residents and 185,000 households represent the company's second-largest
cluster. Its continued move into the Chicagoland area is setting up a fierce battle
against AT&T, which also has been building a huge cluster in the Chicago
market through a series of system swaps and acquisitions in the past years
(Estrella, 1998e).

The system's general manager is responsible for technical operations
only. Marketing, ad-sales and customer-service personnel take their marching
orders from Ameritech's headquarters, as do employees with similar roles in the
company's other metro markets. That way, ANM gets operating scale and
efficiencies. ANM's 120—channel capacity means no rush to push into digital-
cable services. Interactive services are under consideration, with a number of
vendor and service-provider discussions going on.

After nearly four years of relentless video expansion, Ameritech now faces
a crossroad. SBC Communications Inc., new parent company of Ameritech has
strongly opposed constructing or operating costly wireline and wireless cable
systems, even on its home turf. SBC's track record suggests that it may not have

the stomach for Ameritech's aggressive cable expansion over time, particularly

16

as the bills pile up and profits still seem a distant prospect (Breznick, 1999b).
Thus, industry observers have shown suspicion that Ameritech's long-term
standing in the wireline cable business is somewhat in doubt, given the fact that
it has been acquired by SBC, which scrapped Pacific Telesis Group's nascent
cable build-out after buying PacTel (Gibbons, 1998b).

Indeed, in October 1999, as long-rumored, hopeful expectation in the
cable industry, SBC determined to put a hold on all of Ameritech New Media’s
expansion plans. Ameritech suspended deployment of new cable operations and
suspended negotiation of new franchise agreements. SBC announced that the
operations would be suspended until a “deep review” of all business operations
was completed. As a result, Ameritech has notified some communities that it is
stopping negotiations on new cable franchises, at least until the end of the year
(Jones, 1999). The decision by SBC would have a direct impact on cable
competition in the Midwest. The future of Ameritech’s cable operation has been
in doubt since SBC announced plans to buy the LEC in May 1998. Ameritech
continued its aggressive cable overbuilds throughout the lengthy consideration of

the merger, but that has abruptly come to a halt.

SNET Personal Vision

SNET Personal Vision holds a statewide cable franchise in Connecticut,
and currently offers service to 20 localities. The video arm of Southern New
England Telephone began its video operations in September 1996, when the

Connecticut Department of Public Utility Control granted SNET the first statewide

17

cable TV franchise in the US. In March 1997, Southern New England
Telecommunications Corp. (SNET) launched its newest line of business, SNET
Personal Vision, and began signing its first cable TV customers. With this
milestone, SNET became the first company to compete in the cable TV market in
Connecticut, adding yet another dimension to SNET‘s menu of information,
communication and entertainment services.

On August 25, 1999, however, SNET applied for and received permission
from the Connecticut Department of Public Utility Control (DPUC) to suspend
construction of its statewide Hybrid F iber-Coaxial (HFC) network while it
considers alternative technologies for video deployment, such as digital
subscriber line (DSL) technology. SNET maintains that HFC is not an efficient
technology for deployment of video services. The DPUC required SNET to
continue video service where it has already begun to do so, and declined to
relieve SNET of any of its statewide construction obligations. SNET is required
to file a proposal on new technologies and for statewide construction by October
1, 2000 (FCC, 2000).

FCC reported on concerns that SBC, then in the process of purchasing
Ameritech, might sell or abandon Ameritech’s cable overbuild efforts, given that
SBC previously sold or abandoned PacBell multichannel television services. In
addition, SBC's Chairman Edward Whitacre, on May 19, 1998, declined to
commit to maintaining SBC’s video efforts. SNET and SBC said they will
continue to operate SNETs current video service in 20 towns and will make it

available to six more communities where the plant has already been built

18

(Breznick, 1999b). But as of this writing, SBC is focusing on evaluating

alternative technologies for delivering cable service.

GTE Media Ventures

GTE Americast is providing cable television service to customers in
California, Florida, and Hawaii. GTE is franchised to serve pockets of
unincorporated Ventura County and the cities of Oxnard, Thousand Oaks,
Newbury Park and Camarillo.

GTE uses both wireline and wireless technology. GTE provides cable TV
service using high-speed, high-capacity fiber optic in: Cerritos, California;
Ventura County, California; Tampa Bay area, Florida. Those hardwire networks
have about 90,000 subscribers (Farrell, 1998b). GTE's cable presence in St.
Petersburg is part of a Pinellas County-wide buildout that began in 1996. More
than one-half of St. Petersburg has been built out since the franchise agreement
took effect in the fall of 1996.

In addition, the LEC deployed wireless network technology in Oahu,
Hawaii (MMDS system). GTE entered wireless cable through acquisition,
purchasing Oahu Wireless Cable, in May 1997 and commercially launched a
new wireless-cable service in April 1998. GTE competes with entrenched
incumbent Oceanic Cable, the Hawaii division of Time Warner Entertainment, in
the most cabled state in the United States. In Hawaii, cable penetration on Oahu
is about 74 percent. Cable has been successful on Oahu, at least in part,

because the island's mountainous terrain does not lend itself to off-air broadcast

19

reception. But that same terrain poses challenges for wireless cable (Hogan,

1998b).

BellSouth Interactive Media

BellSouth Corp. launched its first digital wireless cable service in
November 1997 in New Orleans. Cox Communications Inc. is the New Orleans
cable incumbent, with some 270,000 subscribers in the region (Gibbons, 1997).
BellSouth has about 20,000 to 30,000 subscribers in New Orleans and its
service is priced about the same as that of the incumbent cable operator (Farrell,
1998b)

BellSouth Corp. began its initial rollout of digital-wireless cable service in
the LEC’s home market of Atlanta in June 1998 (Farrell, 1998b). BellSouth
service was not universally available to everyone in the Atlanta market because
not everyone will have a good line of sight. But BellSouth is rolling out service in
phases in Atlanta, initially targeting more than 700,000 households, or about
one-half of the local market. To reach additional homes, the company will
strengthen its system with signal repeaters (Hogan, 1998c).

BellSouth Corp. launched its digital-wireless-television service in its third
market, in October 1998, competing with Time Warner Cable in Orlando, Fla.
BellSouth plans on passing between 500,000 and 600,000 homes in the Orlando
area, and 75 percent to 80 percent of them are within the line of sight of the
service (Farrell, 1998a). The Orlando launch is one of four markets in south

Florida where the Atlanta-based telco plans to provide wireless video service—

20

Jacksonville, Daytona Beach and Miami are expected to be added as of this
writing (Haugsted, 1998).

SBC and Pacific Bell Video

Although SBC was the first telco to show interest in launching video by
acquiring two cable systems in 1992, they have since been soured on the
business. In recent years, SBC has acquired telcos and promptly jettisoned their
video plans. In California, for instance, SBC shut down or sold off Pacific Telesis'
fledgling wireline and wireless cable systems in the San Jose, Los Angeles and
San Diego areas after buying the West Coast regional Bell in 1997. Similarly, in
Connecticut, SBC recently halted construction of SNET's ambitious alternative
cable plant across the state after taking over the telco. Besides the California
and Connecticut video operations, SBC sold off cable systems that it owned in
the Washington, DC, area and suspended its own fledgling video efforts in
Texas in recent years (Breznick, 1999b).

The Pacific Bell Digital TV launched a service with limited distribution in
Los Angeles in 1996. Pacific Bell service launched commercially May 28, 1997,
in parts of Los Angeles and Orange counties. Vifith MMDS platform, PacBell
chose the Los Angeles areas partly because the city is a very attractive market
for wireless cable thanks to good line-of-sight. PacBell estimated that 85 percent
of the LA markets were within the line of sight of the company’s two digital
towers (Hogan, 19989).

21

Pacific Bell Digital TV competed with MediaOne in several markets within
the MSO's Greater Los Angeles cluster and Charter Communications Inc. in
Southern California. Pacific Bell Digital TV took a targeted marketing approach,
going after higher-end cable customers. PacBell was focusing its efforts on
Orange County and in MediaOne markets that had not yet been completely
rebuilt.

However, after an 18-month long operation, SBC sold a majority
ownership of Pacific Bell Video Services to Prime Cable, an Austin, Texas-based
MSO with about 530,000 subscribers (Hogan, 1998h). As part of the assets in
the transaction, PrimeOne acquired the rights to market the Tele-TV brand and
renamed PrimeOne Tele-TV. Pacific Bell remains a minority partner in Tele-TV
(Haugsted, 1999a). The name was created several years ago as a joint venture
by three local telephone companies—including Pacific Bell, Nynex Corp. and Bell
Atlantic Corp—which were looking to expand into video services. In 1997,
however, PacBell became the only telco to actually launch a TeIe-TV-branded
service, using digital MMDS boxes. Bell Atlantic Corp. and Nynex Corp. shelved
plans for digital MMDS in 1997 (Gibbons, 1997). After SBC bought PacBell, it
was widely believed that SBC would seek to sell off its wireless cable assets
(Hogan, 1998h). At the same time, SBC withdrew from the Americast consortium
in 1997 after its merger with Pacific Telesis Group. It also decided to back away
from video-network deployments, shutting down PacTel’s cable construction in
California. SBC converted its own cable system in Richardson, Texas, into a

data test (Multichannel News, November 24, 1997).

22

Bell Atlantic

Bell Atlantic has struggled to find a video strategy. The telco initially
thought the answer was to buy its way in and made headlines in 1992 when it
announced it was acquiring Tele-Communications Inc. But the deal was not
consummated. In the mid-19905, Bell Atlantic aggressively invested in wireless
cable systems but technological glitches with digital transmissions gave the
company serious competitive problems and the MMDS investments turned out to
be unattractive quickly (Neel, 1999a).

After passage of the 1996 Telecommunications Act, Bell Atlantic created
FutureVision as its video unit. Since then, FutureVision spent $70 million to build
the video delivery network in New Jersey. The LEC eventually shut down its
wireline operations at the end of 1998, after three years of delivering service to
some 2,600 customers in Toms River, NJ. When the service was launched, it
was hailed as a high-tech alternative to existing video providers. Regulators
hoped it would spur the kind of competition that would benefit consumers. But it
soon became apparent that the network was too expensive to deploy beyond the
small confines of its beta test area in Dover Township.

Stepping up its competition to cable in the Northeast, Bell Atlantic \hdeo,
now new-bom video arm, began selling the DIRECTV service in Boston and
Pittsburgh in May 1999 and plans to expand to New York City. Bell Atlantic rolled
out service in New York in 1999 in conjunction with DIRECTV‘s planned debut of

local broadcast signals. Assuming that Congress completes passage of

23

favorable "local-into-Iocal" legislation sometime soon, DIRECTV intends to start
beaming local stations to the New York and Los Angeles markets and then
extend service to up to 20 other major markets.

More of Bell Atlantic's existing DBS subscribers are renting hardware than
buying it. That makes Bell Atlantic's economic model more like the leasing one
followed by Primestar Inc., which DIRECTV recently purchased, than DIRECTV‘s
standard puchase model. Bell Atlantic is now heavily promoting DIRECTV's
hardware and programming packages in most major markets outside New York.
Since introducing DBS service in the Washington, DC, and New Jersey areas in
1998, the regional Bell has expanded to Baltimore, Philadelphia, Boston,
Pittsburgh, northern Delaware and most of central Maryland. Plans call for
extending service to Richmond and Norfolk, Va., Providence, R.l., and other

secondary markets in the year 2000 (Breznick, 1999c).

US West

US West has operated a cable system in Omaha, Nebraska since 1994,
competing the incumbent cable operator Cox with an overbuild of the existing
telephone network. The LEC’s Scottsdale system represents US West's first
foray into switched digital networks since the telco withdrew applications to build
SDV networks in five cities in mid-1995. US West has been delivering video
services over a hybrid fiber/coax network in Omaha since September 1995. As
far as US West is concerned, its decision to build a fiber-to-the-curb, switched

digital video network in a planned community in Scottsdale, Ariz., represents an

24

opportunity for the telco to test the architecture's economic and technical
viability. (Levine, 1997b).

US West has gotten its broadband digital-TV and data service under way
in Phoenix, amid signs that the telephone industry is finally getting serious about
moving to a full-service-network platform. After years of trial and error in US
West's and other telcos“ territories that largely led to dead ends, the Phoenix
launch represents the first large-scale commercial deployment of a US. telco
network suited to delivering an integrated package of voice, data and TV
sen/ices over a single line (Dawson, 1999a).

In Phoenix, US West uses very high speed digital subscriber line ("VDSL")
for distribution of video, high-speed lntemet access, and telephone service over
existing copper telephone lines, and is still the only company in the country using
this distribution technology. In addition, US West announced that it has begun
trials of an interactive service that integrates customers’ telephone, lntemet
access, and existing television service. This service, delivered over dial-up or
DSL connections to television sets, allows customers to switch between Web
surfing and television watching, to do both simultaneously, to view caller ID
information on their television screens, to answer the telephone via a
speakerphone on a set-top box, and to check e-mail. Full-scale deployment of
this service was expected in the fall of 1999, but has not yet occurred (FCC,
2000)

25

Broadband Service Operators (BSOs)

RCN Corp.

RCN was spun off from former MSO C-TEC Corp. in 1997 to create a
competitive local exchange carrier (CLEC). RCN offers bundled phone, video
and lntemet services in several Northeast cities including New York, Boston, and
Washington DC. and Northern California. RCN has service operations in the
Northeast corridor, which spans from Boston to Washington, DC. (Haugsted,
1999c).

RCN already serves about 80,000 customers total in Manhattan, where
the company has activated more than 500 buildings for its bundled services
(Hogan, 1998d). RCN’s aggressive marketing campaigns have already attracted
"significant interest" in the company among residents in New York's outer
boroughs. RCN Corp. expanded its New York City customer base beyond the
island of Manhattan and into Queens with fiber optic network in 1998. The
company started offering its dial-up phone and lntemet services there in April
1998. In Queens, RCN offers a full complement of telecommunications services
including phone and lntemet services over its own fiber optic network (Hogan,
1998d).

RCN Corp. also started delivering cable-television, high-speed lntemet
and telephone service in Gaithersburg, Md., a Washington, DC, suburb. In the
District of Columbia, its cable-television product has achieved 30 percent
penetration, according to the company’s press release. It also said that 80

percent of RCN's facilities-based customers in Washington subscribe to more

26

than one service. All of RCN's Washington customers to date are multiple-
dwelling unit residents (Multichannel News, July 26, 1999).

In June 1998, the FCC approved RCN Corp.'s application to provide
open-video-system (OVS) service in 109 Philadelphia-area communities, as well
as in 33 communities in the San Francisco area. RCN's plans are ambitious: In
its FCC applications, RCN revealed that it will provide 330 analog video-
programming channels, keeping 110 for itself and 220 for unaffiliated
programmers. Ordinarily, OVS operators are not obliged to obtain local
franchises, although they must comply with the same must-carry, retransmission-
consent and PEG-access (public, educational and government) requirements as
incumbent cable operators. (Heam, 1998a).

As recently as 1999, RCN entered in the Chicago market by purchasing
the 21st Century Telecommunications Group, an overbuilder already operating in
one of the densest areas in the country. With a vision of a facilities-based
provider supplying a bundle of communications products over cable, including
video, high-speed lntemet and telephony, 21st Century Telecom Group acquired
the franchise to provide bundled service in Area 1 of Chicago. The 360 applied

for franchises to serve three more of the five areas (Estrella & Haugsted, 1999d).

Knology Holdings Co.
While RCN Corp. has been grabbing headlines with its bundled-
telecommunications forays into big Northeastern markets, another company has

been quietly following a similar approach in smaller Southeastern cities.

27

Knology—according to its Web site (www.knology.com), the name comes from
"combining knowledge with technology"—-is branching out from that initial base.
The West Point, Ga.-based firm has targeted Southeastern cities with population
bases of more than 100,000 for new digital overbuilds. So far, Knology has
purchased overbuilt systems in Columbus, Ga., Panama City, Fla., and
Huntsville and Montgomery, Ala. It has obtained 28 other franchises to overbuild
markets in Alabama, Florida, Georgia and South Carolina. Its current systems
and franchise areas pass 246,575 homes. As of July 1999, Knology has 91,838
households for its bundled offerings, 80,068 of which are specifically for video
services (Grebb, 1999).

The list of other companies in which Knology's original backer, ITC
Holding Co. Inc., owns equity indicates a track record in data and telephony.
They include MindSpring Enterprises Inc., an lntemet-service provider with about
340,000 subscribers; DeltaCom Inc., a long-distance reseller now in 14
Southeast markets; lnterstateNalley Telephone Co., a local telephone company
in parts of Georgia and Alabama; and PowerTel Inc., a wireless phone operator
in 25 Southeast markets (Gibbons, 1998a). Knology's sister companies provide it
with a variety of services. Knology uses DeltaCom’s fiber network to provide
long-distance service and an lntemet link. Knology and DeltaCom jointly run a
network-operations center in West Point. Also, MindSpring provides lntemet
content and customer support to Knology. Like RCN, which has forged links with
electric utilities in several markets, Knology reached a strategic partnership with

South Carolina utility Scana Communications Inc., which owns 7 percent of

28

Knology. Other Knology equity-holders include AT&T Corp—its AT&T Venture

Funds owns 14 percent of Knology (Gibbons, 1998a).

LyncStar Integrated Communications

LyncStar Integrated Communications LLC, offers cable to multiple-
dwelling units via open-video systems in 15 states. The OVS operator competes
with Time Warner in Austin, Texas. LyncStar also serves Houston, where its 20
properties served constitute its largest cluster. It has already secured a
traditional cable franchise in the west Texas community of Midland and another
in Little Rock, Ark. With some 15,000 customers over OVS systems already up
and running in 15 states, LyncStar recently received its third OVS license in
Colorado when the FCC approved its application to serve Fort Collins, a college
town north of Denver. The company also has OVS licenses to serve portions of

Denver and Aurora, Colorado (Estrella, 1998a).

Entry Strategies

The telcos can provide multichannel television service in one of four ways.
One option is to transmit programming over the upgraded, broadband networks
currently under development. Telephone calls, interactive sen/ices, and cable
offerings would all be sent over the same network. A second option is to build an
entirely separate network only to provide cable services. A third alternative is
wireless cable, whereby signals are broadcast over the airwaves to a receiver

installed at each subscriber‘s house. The technology used for wireless cable is

29

known as a multipoint, multichannel distribution system (MMDS). The fourth

option for telcos is to use the direct broadcast satellite (DBS) method.

System Topology

Bell Atlantic and Pacific Telesis have used the first method, while
Ameritech employs the second approach. Bell South, Bell Atlantic, Pacific
Telesis, and NYNEX have also tried the third option, because the first two
involve technological and regulatory hurdles (Gooley, 1996).

Ameritech Corp., GTE Corp. and Southern New England
Telecommunications Corp. have used cable-network technology overlaid on
existing copper plant, and BellSouth Corp. is widely offering TV over wireless
cable, as well as preparing to experiment with a fiber-to-the-home delivery
system in Atlanta (Dawson, 1999a). Both SNET and Ameritech are building their
own fiber optic cable systems in their respective territories. Separate from their
telephone network, Ameritech and SNET deliver cable TV programming with
their hybrid fiber-and-coaxiaI-cable system. From a reliability standpoint, a
wireline system is superior because of the possibility of weather problems with
the wireless technology.

Ameritech has been constructing its HF C network for cable sen/ice in the
LEC’s biggest markets and surrounding areas in an attempt to achieve
economies of scale. Ameritech New Media offers a package of video

programming that it delivers via a new, two-way, hybrid fiber/coaxiaI-cable

30

system, separate from Ameritech's telephone network. SN ET, too, was granted
state-wide cable franchises.

On the other hand, BellSouth and GTE’s overall video strategy includes
both wireless and wireline technology, depending on the logistics of the particular
market (Hogan, 1998c). BellSouth has taken a cautious approach to wireless
video, launching the service in a few markets where the technology's Iine-of-sight
constraints do not present a major problem. In the areas where the telco cannot
reach customers with its digital-wireless signal, traditional coaxial cable is being
used.

In using wireless cable, the telcos are making a tradeoff. MMDS lets them
provide only a limited number of channels and service offerings, not the full
range of interactive services they envision. This approach has two distinct
advantages, however. One of the benefits of entering a new market with MMDS
technology is that the company can blanket a city with the technical ability to
serve the market much quicker than would be possible with wireline technology.
MMDS lets them provide video services more quickly than if they had to wait for
completion of their broadband networks. This gives them a new revenue
source—and a higher profile—much sooner. Clearly, the speed is very attractive
with MMDS and the operator only need invest in subscriber homes, not in every
home passed (Hogan, 1998c). Second, MMDS is less costly than building a new
network.

Despite these advantages, though, most telcos are still building

broadband networks. Once these broadband networks are completed, the telcos

31

will simply use wireless cable as a complementary business to the full range of
interactive services that broadband networks will allow. BellSouth has embraced
wireless technology for this reason. Deploying a wireless-video network is
cheaper and, with new digital technology, it is possible to offer as many channels
as a conventional hardwire system. GTE, too, decided to go with MMDS
because it provided a low-cost alternative to building a hardwire network. MMDS
also offers some advantages over direct-broadcast satellite. It has the capability
of delivering local signals (Farrell, 1998b).

Even though MMDS costs less to launch than deploying a hardwire cable
network, it is not cheap enough to offer a significant discount to consumers.
Wireless cable offerings are priced between 10 percent and 15 percent lower
than the incumbent cable provider's service. At that marginal discount, wireless
cable subscribers tend to be budget-minded individuals who will drop the service
the minute that they get a better deal from DBS.

The BellSouth and GTE wireless systems is digital, meaning that it has
better picture quality and channel capacity than analog MMDS. GTE has also
added enhanced services to its MMDS system, like near-video-on-demand and .
time-shifted television programming. And the company is investigating lntemet
data services over wireless.

Nevertheless, MMDS has suffered over the years with a critical problem
vis-a-vis wireline cable service: LOS characteristics. They prohibit transmission
of signals through obstacles like buildings or trees. MMDS is a way into the

market, but “it’s a desperate way in“ (Farrell, 1998b). In fact, telephone

32

companies have been burned by wireless video in the past: Bell Atlantic's plans
for a nationwide wireless video service based on CAI Wireless Systems Inc's
network were scrapped in 1996, after the telco could not overcome LOS
problems. If the technology is so bad, why are GTE and BellSouth embracing it?
Both companies jumped into wireless believing that their significant resources
and marketing know-how were enough to make it a success (Farrell, 1998b).

BellSouth and the others are not hinging all of their bets on wireless. In
fact, BellSouth is using wireless video only in markets where it has at least 70
percent LOS. For those homes that cannot be reached via MMDS, the telco is
using a traditional hardwire strategy.

Some Bell companies are exploring yet another approach to offering cable
television, reasoning that no single technology may be entirely appropriate for all
local markets. This alternative technology is asymmetric digital subscriber line
(ADSL), which allows transmission of video signals over existing copper
telephone lines through the use of equipment installed at both ends of the line.
As a result, the telcos have become increasingly interested in ADSL since early
1996. Like MMDS, ADSL is not a long-term solution for providing a full array of
interactive services to customers, but it does give telcos faster entry into the
cable television market, while taking advantage of their existing copper lines.

At a time when telcos are scrambling to upgrade their plants to create one
homogenous system, BellSouth is instead opting for a "mosaic" approach,
knitting together telephone, wireless, fiber optic and conventional copper cable

lines to bring high-demand data and lntemet products into its territory. The

33

mosaic is part of a BellSouth strategy to shift emphasis from technology to
delivering service, even as the telco goes through a hefty fiber optic upgrade in
its Atlanta and South Florida markets. GTE also uses a mix of solutions as they
make sense with each particular market.

By strategically selecting combinations of telephone, wireless, fiber optic
and cable lines, BellSouth is opening up its menu of high-speed data, voice and
entertainment senlices for more customers than if it had waited for the whole-
system upgrades. The system involves making an economic study of the various
service options to decide the best means of transportation and the best
combination of services.

That mix of technology may be a challenge to manage, but it gives
BellSouth a natural progression for upgrades and new service offerings.
BellSouth also provides the high-speed data service, which uses traditional
copper telephone wire or fiber-optics. For areas where fiber-optic lines are wired
to the curb, there is even a plan to experiment with fiber-optic lines wired into the
homes (Brown, 1999b).

AT&T once adopted the fourth method, DB3, for their video services. In
early 1996, AT&T acquired a small interest in Hughes Electronics' DIRECTV,
and it plans to market the popular service to its customers. MCI also has bid
successfully in an FCC auction for its own DBS license. Currently, Bell Atlantic,
SBC, Cincinnati Bell, and GTE have an agreement with 038 providers.

GTE Corp. signed a multiyear agreement with DIRECTV Inc. to market

and distribute the direct-broadcast satellite company's digital-video service.

DIRECTV also signed with Bell Atlantic Corp. and SBC Communications Inc. to
offer its service in the regional Bell operating companies’ service territories.
Unlike DIRECTV's earlier pact with AT&T, no money changed hands in these
deals.

As part of its one-stop-shopping telecommunications plan, LECs allow
consumers to purchase or lease the Digital Satellite System hardware and order
DIRECTV programming with one telephone call. GTE’s deal with DIRECTV does
not mean the company is pulling back on its other video ventures, which include
wireline and wireless-cable service in California, Florida and Hawaii. But it does
allow GTE to offer video as part of its bundled telecommunication offer in more
markets sooner. Currently, GTE is marketing DIRECTV in areas where it does
not offer a competing video service. But over the next several years, GTE plans
to offer both DIRECTV and GTE's own video services in some of the same
markets.

GTE had also announced recently that it "will write down investments in
hybrid fiber-coax (HFC) technology associated with its video trial markets
because of advances in new technologies that will eliminate the need for
duplicate networks." GTE plans a move to VDSL (very high-speed digital

subscriber line), a wire that can integrate voice, data and video on twisted

copper.

35

Programming

Americast is the national consortium created by Ameritech Corp.,
BellSouth Corp., GTE Corp., SBC Communications Inc. and Disney
TeleVentures — a wholly-owned subsidiary of the Walt Disney Company — to
develop, produce, and market video programming for its telco members. The
telcos realized that developing entertainment content for their future cable
ventures would be risky, so they joined together to spread the financial risk. In
addition, they recognized they know little about the business, so they aligned
themselves with one of the world’s premier entertainment providers in Disney.
The Americast consortium hopes to combine Disney's creative and marketing
excellence with the communications and service expertise of the communication
partners to provide quality cable programming and services.

Americast offers customers over 90 channels of cable television
entertainment, including local broadcast, expanded basic, premium and Pay-Per-
View. With Americasts interactive program guide, StarSight, users can navigate
through up to four days of television scheduling. There is no more watching and
waiting for a scrolling list of programs. As viewers flip through the channels, at
each stop, a message on the screen will indicate which show is currently in
progress and the length of the show, even if a commercial is showing. StarSight
also offers a parental control feature to help parents make educated decisions
on appropriate programming for kids (SNET, 1997).

Americast has been scaling back, scrapping its plans to create original

programming and handing over other marketing and programming tasks to

Disney. SBC has dropped out, although Southern New England
Telecommunications Corp. signed on, and it is selling Americast-branded cable
service in Connecticut. Americast claims 235,000 video subscribers in 80
communities.

Americast offers consumers several packages of programming, including
local broadcast stations, today's cable TV favorites such as ESPN, CNN and
Nickelodeon, and newer channels such as The Sci-Fl Channel, The History
Channel, ESPN2, and Teen Disney. Americast's “Premiercast’ includes premium
channels. Its “Express Cinema,” the in-home movie service, is comparable to
PPV programming. It is fed on a near-video-on-demand basis. Customers get
access to up to 30 hit movies every month, starting as often as every 30 minutes.
Viewers use their universal remote control, R.E.D., to select movies, as well as
special events such as concerts and boxing matches, via the service's Interactive
Program Guide.

“Local Connections“ provides up-to-date community information 24 hours
a day. It allows Americast customers to access dozens of community information
channels to learn local movie theatre show times, concerts and plays, plus library
and community events. And customers with expanded basic cable service can
find information including local restaurants that deliver, phone numbers for

schools, clubs and community services, and a five-day weather forecast.

37

Chapter 3

OVERBUILD COMPETITION AND MARKET PREDATION

Search for Comgtition in Cable Televisi_on: Theory and Evidence

Even though substantial inroads into the cable industry are evident in
recent years, they exhibit little impression that overbuild competition will flourish
in the near future in terms of entry rates and scales. Since entry decisions and
post-entry equilibrium will heavily depend not only on demand levels, market
growth, and revenue potential, but also incumbents’ strategies and competitive
reactions—particularly price reduction and product differentiation—it is
necessary to take these factors into consideration in the assessment of overbuild
competition. Among these factors, market growth potential will be discussed in
Chapter 5 in conjunction with new service offerings such as lntemet access
service. The primary goal of the present chapter is to investigate competitive
effects in the cable price and service quality dimensions. This chapter presents
theoretical discussion of price response with particular emphasis in the possibility
and rationale of market predation by an incumbent firm, followed by model
specification, major procedures of data collection, and summary findings of the
empirical investigation of a sample of 247 observations.

The chapter begins with the debate surrounding the efficacy of regulation
and competition. A review of previous studies is also provided to start discussion
of competitive effects of overbuilds. The primary findings of this chapter are

presented in an analysis that was designed to provide empirical evidence of the

38

effects of competition in cable television.

Efficacy of Regulation and Competition

A number of regulatory regimes have existed over the brief history of
cable television in the United States. Since its inception, the US. cable industry
has been subjected to virtually every conceivable regulatory regime from
unregulation through tight rate regulation to deregulation and reregulation, and
now deregulation again under open entry policy. These swings may indicate that
government authorities and cable consumers have continuously been
bewildered by the surprisingly myopic regulatory plans for cable services as well
as the opportunistic behavior of the cable television industry.

These shifts in regulation and policy prompted academic researchers to
begin investigating the very fundamental assumptions that have been taken for
granted heretofore in cable television. They investigated, among others, such
issues as economies of scale (Noam, 1985; Owen & Gottlieb, 1986; Owen &
Greenhalgh, 1986), exclusive franchising policy (Hazlett, 1986, 1987, 1991), the
effects of rate regulation (Hazlett, 1996, 1997; Mayo & Otsuka, 1991; Otsuka,
1997), and the effects of competition (Beil, Dazzio, Ekelund, & Jackson, 1993;
Emmons & Prager, 1997; FCC, 1994b, 1999b, Jayaratne, 1996).

In the early years, the provision of cable television service in the US. was
regarded as a natural monopoly due to the underlying technology. Until recently,
it has been assumed that a single firm can serve the market at lower average

cost than two or more firms. Johnson and Blau (1974), for example, argued that

39

municipal cable franchises that allow only one system to serve the entire market
at any given time are efficient because “if a second firm is built alongside the
first, an unnecessary duplication of resources is accomplished“ (p.325). Due to
the economies of scale in cable television, they viewed that “such competition
would be wasteful” (p.325). The natural outcome of this view meant an exclusive
franchise in the local cable market. In the absence of regulation, however, the
single franchisee would charge the profit-maximizing monopoly price. This
behavior will result in a misallocation of resources and a reduction in social
welfare relative to the competitive ideal. The traditional solution to the natural
monopoly problem is to impose some form of rate regulation, granting the
franchising government the authority to approve or deny proposed rate
increases.

Ideally, rate regulation can be designed to promote cost minimization
while effectively restraining the market power of monopolists that might lead to
monopoly prices. By squeezing price-marginal costs margins, rate regulation can
increase supply toward the optimal level of output, preventing social losses due
to monopoly pricing behavior. In theory, the efficacy of regulation has been
challenged from many perspectives, notably by Stigler (1971) and Peltzman
(1976). In the Stiglerian view of “capture theory“, “regulation is acquired by the
industry and is designed and operated primarily for its benefit“ (Stigler, 1971,
p.3). In the Peltzmanian perspective of regulation, the regulator would maximize
political support from both consumers and producers in establishing an optimal

regulated price (Peltzman, 1976). In practice, municipal rate regulation is not

40

always effective. Particularly, its effects on cost, price, and quality were
ambiguous. Thus, the efficacy of rate regulation is an empirical question.

Early academic studies of cable television (Comanor & Mitchell, 1971;
Noll, Peck, & McGowan, 1973; Park, 1972) suggested that the relaxation of FCC
rules was necessary to help stimulate cable growth in urban areas. These
studies and a landmark decision by the US. Supreme Court in the HBO case
precipitated the rather dramatic deregulation of cable television in the early
1980s. Since the Cable Communications Policy Act of 1984 ended local
regulation of basic cable rates in all but a few markets by 1987, basic cable rates
rose rapidly. Nevertheless, it was also apparent that the number of channels
included in the typical basic service package increased significantly at the same
time.

The price hikes of cable television service in the 1980s have led to some
empirical research on the relationship between cable price and rate regulation.
For example, Mayo and Otsuka (1991) investigated the effects of various
regulatory forms on cable price, using a simultaneous equations model that was
designed to reflect the relationship between demand and price. Their analysis of
1,300 cable systems revealed that while regulation did not lead to economically
efficient (marginal cost) prices for basic service, it did act to keep prices
considerably below monopoly levels. In addition, there were some significant
differences in the effectiveness of the various types of regulation practiced in the
pre—deregulation period in the 1980s. While ROR regulation or state-level

deregulation had no effects on price, common tariff regulation led to significantly

41

lower basic cable prices than local franchise regulation. Interestingly, Mayo and
Otsuka also found that where regulation had the effect of constraining basic
prices, cable firms responded by charging higher prices for premium services.

While the Mayo and Otsuka study attempted to find the price-restraining
effects of rate regulation, Rubinovitz (1993) investigated the effect of
deregulation on price increases in cable television. An increased exercise of
market power, due to the change from a regulated to an unregulated
environment, enabled cable operators to increase cable price by 18 percent
since deregulation, holding quality improvements and other cost changes
constant (Rubinovitz, 1993). Thus, these two studies reveal that rate regulation
has been effective, to a certain degree, in controlling the price of cable television
services. In contrast to these findings, however, Prager (1990) found no
evidence to support the view that regulation effectively constrains the rates
charged for basic cable television service.

With respect to output, Hazlett (1997) analyzed key market evidence to
observe consumer reaction to the rate regulation imposed under the 1992 Cable
Act. Contrary to the stated goals of rate regulation, Hazlett found that basic cable
output, as measured in penetration, did not positively deviate from trend during
the period of rate reductions. A number of subscribership measures exhibited
statistically significant declines in growth, in fact, during the two—year window in
which rate reregulation under the 1992 Act lowered cable subscription prices.

In the matter of quality, Otsuka (1997) discovered that franchise

regulation has produced higher service quality than does unconstrained

42

monopoly. However, this does not necessarily mean the successful
accomplishment of the stated goals by the regulators. These performance levels
are far from the public’s expectations, not to mention economists’, because the
findings simply suggest the service quality under rate regulation is better than
that of a “monopoly”. Indeed, Prager (1990) found that rate regulation in cable
television significantly reduced both service quality and firm responsiveness to
community needs and interests.

An alternative solution to rate regulation is competitive bidding for a
franchise, proposed first by Demsetz (1968). He argues that competitive bidding
for the right to become the sole producer in a naturally monopolistic industry
would yield the same benefits as rate regulation at a lower cost to society. Other
economists have challenged Demsetzian approach to controlling monopoly
behavior, however. Most notably, this approach may lead to cost inefficiences
and allow cable operators to engage in ex post opportunistic behavior
(Williamson, 1976). This may involve raising prices above the agreed-upon level
or shirking with respect to other product characteristics, such as the quality of
service. On the other, reputational concerns can limit the extent of such
opportunistic behavior (Prager, 1990; Zupan, 1989b).

Despite the deregulation of cable television, the typical local government
issues an exclusive franchise as an efficient and necessary regulation of a
“natural monopoly“. Beginning in the mid-19803, however, this practice was
seriously challenged in both the court of law (Preferred Communications vs. City
of Los Angeles, 1986) and academia (Hazlett, 1986, 1987; Lee, 1983). In the

43

Prefen'ed case, the US. Supreme Court upheld an appeals court decision that
exclusive franchising by municipal governments implicates the freedom of the
press and it is in violation of the First Amendment.

Researchers noted problems embedded in the policy of exclusive
franchising and the inefficacy of rate regulation by government in an attempt to
curb monopoly market power. Critics of monopoly franchising have argued that
local government policies, rather than the economics of cable television, have
created monopoly market structure. Posner (1972), for example, blamed the
practice of exclusive franchising as the factor substantially responsible for
discouraging competition in the cable industry. He states that “the immediate
cause of [monopoly] is not any inherent characteristic of cable television, but the
fact that a cable company must obtain a municipal franchise...[and]
municipalities do not grant more than one cable franchise in any area within their
jurisdiction“ (Posner, 1972, p.111). Johnson and Blau (1974) have asserted that
three major economic factors (i.e., costs, pricing, and investment) led to an
inevitable natural monopoly in cable television, which, in turn, dictates an
exclusive franchise solution. Hazlett (1987) sharply attacked the practice of
exclusive cable franchising in a critique of Johnson and Blau. Hazlett argued,
however, that monopoly market structure in cable was a product of three factors,
none of which supports the monopoly franchise solution. First, multiple entry was
illegal in the great majority of local jurisdictions. The expense of court challenges
to such licensing exclusions acts as a significant entry barrier. Second, this legal

entry barrier is due, not to natural monopoly constraints, but to local

governments’ objective of maximizing their rents by actively promoting a
monopoly arrangement. Third, the assumption of overinvestment due to
duplicative costs by competitive entry is a myth because market’s efficient
allocation of investment capital will result in the firrn’s intemalizatlon of all
duplicative costs. They will ultimately discipline entrepreneurs to avoid
overinvestment. In short, open entry and competition would make consumers
and society better off in terms of economic welfare, but municipalities rarely opt
for such a policy.

Hazlett further refined his “free market, open entry approach” to cable
television in subsequent research. Applying public choice theory, Hazlett (1990,
1991) hypothesizes that municipal franchising authorities generally will have an
incentive to grant monopoly franchises to incumbent cable operators in
exchange for side payments to municipal officials and political pressure groups.
He argued that municipalities intentionally impede competition and show anti-
consumer intent by trying to protect the cable monopoly. Resources expended to
obtain these rents add nothing to consumer or producer surplus because they
are wasted in unproductive or purely rent-seeking activities. These resource
misallocations, according to Zupan (1989a, 1989b), can result in a significant
portion of cable costs in the form of nonprice concessions (e.g., l-nets, excess
channel capacity, community programming). Such concessions, of course, would
raise the costs associated with building and running a cable system. Zupan
(1989b) found that nonprice concessions accounted for 26 percent of building

costs and 11 percent of operating expenses. Moreover, while the benefits from

45

nonprice concessions are both relatively concentrated and focused in the hands
of individuals who tend to be active and influential in local community politics, the
benefits from expenditures on “plain vanilla“ cable services are relatively diffuse
and accrue to the average cable consumers (Zupan, 1989a). If these costs are
ultimately immersed in the consumer price, the efficiency-enhancing potential of
franchise bidding schemes may be curtailed.

As criticism of rate regulation and exclusive franchise has become the
rage in recent years, entry has become a favorite direction for reform. While the
1992 Cable Act prohibits local franchising authorities from granting an exclusive
franchise without “legitimate“ reasons, the 1996 Telecommunications Act

introduces open entry policy to replace rate regulation.

Empirical Foundation of Cable Competition

There have been a few surveys and econometric studies comparing
competitive markets with monopolies. Every inquiry found that competition in
cable lowers price significantly. Bolstering those results is the fact that
unmeasured, nonprice aspects of competitive supply are also more consumer-
friendly than service characteristics found in monopoly markets. Several studies
have specifically examined the impact of competition on cable service prices and
service offerings. Menine (1990) found that the price of basic cable service was
18 percent lower and the number of basic channels 21 percent higher in a 1990
sample of 26 competitive markets relative to a matched sample of monopoly

markets. The cost per channel was 33 percent higher in the monopoly market.

Applying a simplified OLS method to the same sample, Levin and Meisel (1991)
concluded that overbuild competition reduced the price of basic cable service by
22 percent to 30 percent.

Bell, Dazzle, Ekelund, and Jackson (1993) estimated the effect of
competition on social welfare, applying Hazlett’s public choice theory and Mayo
and Otsuka’s (1991) model of simultaneous equations. Drawing on a sample of
178 systems (48 competing, 130 monopoly), Bell et al. (1993) found that
competition caused significant decreases in price for both basic and pay service,
which, in turn, have a positive effects on the demand for basic cable service
measured by total penetration in a market. More specifically, Bell et al. (1993)
showed that competition reduced the price of basic cable service by $3.21 and
the price of a typical premium channel by $1.15 in 1989. Aggregating this
monthly data to yearly figures over the entire non-competing cable systems in
the United States in 1989, they approximated the annual deadweight loss to be
$3.6 billion. They concluded, therefore, that the reregulation of the cable price
accomplished in the 1992 Act would fail to produce significant consumer benefits
and that competition may be a viable substitute for rate regulation of cable
services.

There are a few empirical studies that have specifically examined the
impact of nonprivate ownership on prices or service offerings in the cable
television industry. Using a pre- and post-deregulation design, Emmons and

Prager (1997) analyzed the underlying characteristics and behavior of competing

Versus monopoly operators on the one hand, and private versus nonprivate

47

ownership on the other. They concluded private cable systems facing direct
competition from another private system offered basic service at prices that were
20.1 percent lower in 1989 than those charged by private monopoly systems.
Emmons and Prager (1997, pp.739-740) concluded “the price differentials were
even greater for municipally owned systems (23.9 percent in 1983, 25.6 percent
in 1989), other nonprivately owned systems (50.3 percent in 1983, 47.8 percent
in 1989), and systems operating in markets where a privately owned firm
competed directly with a nonprivately owned firm (70.8 percent in 1983, 73.7
percent in 1989)“. This study may be criticized in two aspects. First, the effects of
competition and ownership modes derived by regression coefficients were
overestimated in their interpretation. Emmons and Prager erred in interpreting
the coefficients from their regression analysis. The coefficient on MUNI was -
.256 and they interpreted it as 25.6 percent. Because their estimation was based
on a log-level (or semi-log) model, however, more precise estimates would be
22.6 percent [exp(-.256) -1 =.226] for municipal ownership and 18.2 percent for
competing systems [exp(-.201) — 1 =.182] in 1989. Similarly, the price differential
is 38 percent instead of 50.3 percent for other nonprivately owned systems, and
52 percent Instead of 70.8 percent for systems competing with a nonprivately
owned firm. This adjustment would not be as crucial for small percentage
changes, but it seems necessary with such substantial effects in Emmons and
Prager. More critically, Emmons and Prager’s estimation did not control for
“quality“. They justified it by stating that “quality differentials among the various

subsamples were statistically insignificant, except for the other nonprivately

48

owned group in 1989“. But in fact, there were statistically very significant
differentials among the subsamples in both 1983 and 1989, presented in Table 1
(p.736) and Table 2 (p.737). For example, the average number of basic channels
delivered on privately owned monopoly systems was 17.73, whereas 24.63 on
competing private systems, 12.75 on municipal systems, and 9.10 on other
nonprivate systems in the 1989 data. These differentials in the number of
channels, used here as a proxy for “quality“, were significantly different from zero
all at the .01 level. Apparently, the omission of this key variable in their
estimation led to the unusually large differences in price due to the failure of
“quality“ control.

The FCC found in a 1993 study that basic rates charged by private cable
systems engaged in head-to-head competition with another private cable system
were 22.1 percent lower than rates charged by monopoly systems on a per-
channel basis (FCC, 1993). That differential was estimated in a price equation
that adjusted for other factors such as the number of system subscribers and the
number of satellite channels carried. (When the competitive dummy subsample
was extended to include municipal systems and private systems competing with
municipals, the price difference increased to about 27.9 percent.) The same
survey data were reworked in the FCC's Febnrary 1994 rate regulation
rulemaking. In an attempt to devise rate caps under the 1992 Cable Act, the
FCC (1994b) conducted research on the effects of competition on cable price.
The agency used a sample of 420 franchise areas, of which 237 were in the

noncompetitive subsample, 148 in the low-penetration sample, 39 in the

49

overbuild sample, and 11 in the municipal sample. The FCC found that in 1992,
franchise areas with a complete overbuild (100 percent overlap in a franchise
area) had basic cable prices that were 16 percent lower than those in monopoly
franchise areas with no overlap. A further 1 percent adjustment was made to
reflect heterogeneity in the degree of overlaps within the “overbuild“ subsample.
Municipal cable systems were found to have rates that were 21 percent lower
than comparable competitive franchises and 37 percent lower than comparable
noncompetitive franchises (FCC, 1994b). Based on these findings, a further 7
percent rate rollback was implemented in 1994, following the 10 percent price
reduction in the previous year.

Using the same data, but with a different pricing equation, Jayaratne
(1996) found that rates in the overbuild areas were 12 percent lower than
monopoly rates. Jayaratne criticized the use of “overlap" measures in the FCC
study. Instead of using a dummy variable for competition, the FCC's estimation
included a continuous variable of “overlap“, as measured by the fraction of
households passed by both the incumbent and new entrant in a given franchise
area. Thus, the problem due to this measure of “overlap“ in the FCC study is that
the FCC’s equation forces price to be a continuous function of the degree of
overlap since it would have a continuous value in the interval [0,1]. As a result,
the pricing behavior of overbuilds with very little overlap (e.g., 10 percent of the
entire franchise area) is assumed to have a price that is close to a monopolist
price. In reality, however, pricing behavior would be quite different from that of

the monopoly case if further expansion is anticipated or if price is a

50

discontinuous function of overlap (Jayaratne, 1996). To illustrate the point,
consider the following case. In a particular section in the franchise area, price
may have only two values; the competitive price on overbuild streets and the
monopoly price in the rest. Further, if the local franchising authority requires a
uniform price in the entire franchise area, price would take only a single value
regardless of the degree of overlap. Thus, the functional form used in the FCC
study would produce a biased estimate of overlap, hence an incorrect estimate
of competition on price.

Two comments seem to be warranted on these studies by the FCC (1993,
1994b, 1999b) and Jarayatne (1996), which used the same data as the FCC.
First, the unit of analysis in these studies would very likely produce incorrect
results in their estimation. Both the FCC and Jarayatne used each and every
franchise as the unit of analysis, based on “Community Unit“. A Community Unit
Identification Number (CUID) is a unique identification code assigned by the FCC
to a single operator within a community unless the operator also serves a
separate area within the same community. In that case, there would be two
separate CUIDs, one for each area served. Thus, a company is considered to be
an operator for each CUID it serves. This means that if a company serves 50
CUIDs included in the sample, that company was referred to as 50 systems. But
in general, a single cable system serves multiple franchise areas, that is
communities, within a market. Therefore, these studies cannot properly control
for system specific characteristics such as the number of subscribers without

losing the possible effect of scale economies. If the number of subscribers in

51

each community is assigned to each unit of analysis when the community in
question is part of a larger market served by a single system, the estimation
would suffer from the omission of scale effects. On the other hand, if the number
of subscribers in the equation is the total subscribers to the system, assigned to
each community simultaneously, it would result in a critical error in
measurement. Either way, the estimation is biased.

Second, the estimation of overbuild effects in the FCC and Jayaratne
studies may or may not be precise. Facing difficulties of collecting this
information, the dependent variable used in the FCC survey was the average
franchise area revenues from regulated senrices per subscriber per month.
Jayaratne took advantage of the FCC dataset, but modified the dependent
variable to include only monthly subscription charges for all regulated tiers. Thus,
both models had the heterogeneity problems with respect to “overbuilding“. This
is why the FCC opted for a continuous variable OVERLAP rather than a dummy
variable OVERBUILD, while Jayaratne attempted to improve the estimation by
including OVERBUILD*OVERIAP as well as a pure dummy OVERBUILD.
Unfortunately, however, neither of these two would be able to solve the
heterogeneity problems.

When only a small portion of the community is wired by the new entrant,
the incumbent will most likely employ price discrimination. The literature is
replete with selective price-cutting. Hazlett and Spitzer (1997) summarize that
“when an entrant begins constructing a competing system, the incumbent

typically cuts prices just in those areas (literally, the blocks) where the competitor

52

has placed its lines. Prices have been sharply lower in areas where rivalry has
been in evidence“ (p.29). Indeed, the telephone survey conducted for the present
study revealed that selective price-cutting was not uncommon. For example,
Time Wamer in the Tampa, Florida areas charged $19.95 in the areas overbuilt
by GTE, compared to $29.95 in the nonoverbuild blocks for the same service. If
selective price—cutting is available and actually employed, the use of an
interaction term (e.g., competition dummy by overlap) would have a high
possibility of producing incorrect estimates. In the above Tampa Bay case, the
correct measure of competitive effect is achieved if and only if the price charged
in the select overbuild neighborhoods (i.e., $19.95) is used as the dependent
variable and the competition dummy only as the criterion variable. In other
words, the degree of ovenap is irrelevant and unnecessary in measuring the
effects of competition on price. Regardless of the degree of overlap within a
franchise area, the effect of competition is the same in that market. Suppose a
cable system with 50 percent overbuild charges $20 in overbuilt areas and $30 in
the rest. This two-tier pricing will generate a weighted average revenue of $25
per subscriber per month. And the competitive effect would be a differential
between $25 and monopoly price. Suppose the average price of monopoly
systems is $30. Then, the methodology used in the FCC and Jayaratne studies
would result in $5 differential between monopolies and duopolies, whereas the
true effect of overbuild competition is $10. In short, the two studies suffer from
the underestimation of the overbuild effects on price. More frustrating is the fact

that no one knows whether or not the effect of competition based on this method

53

is even incorrect.

The most recent research also confirms previous studies. The FCC’s
Price Survey, conducted in 1998, reports that the gap in average monthly rates
between competitive and noncompetitive operators has widened. The average
monthly rate (for the BST, CPST, and equipment) charged by cable operators
facing effective competition was $27.15 and $28.71 as of July 1, 1997 and 1998,
respectively. For those not facing effective competition, the average monthly rate
was $28.56 and $30.53, respectively, during the same time period. This
represents a differential of 5.2 percent and 6.3 percent, respectively, in average
monthly rates between the competitive and the noncompetitive. Further, the
average monthly rates charged by systems facing head-to-head competition (i.e.,
systems within the competitive group that meet the overbuild test or the LEC
test) was 14 percent less than the average monthly rate charged by

noncompetitive systems.

Table 1 Average Monthly Rates of Competitive and Noncompetitive Groups

 

 

 

 

 

 

 

Date Competitive Noncompetitive $ Difference % Difference
Group Group
8/31I93 $20.51 $22.23 $1 .72 8.4%
7I14l94 $21.04 $21.61 $1.03 2.7%
7/1/95 $22.88 $24.43 $1 .55 6.8%“
711/96 $25.42 $26.21 $0.79 3.1%“
7l1l97 $27.15 $28.56 $1.41 5.2%*
7l1l98 $28.71 $30.53 $1.82 6.3%“

 

 

 

 

 

Source: FCC 1996 Price Survey, FCC 1997 Price Survey, FCC 1998 Price Survey.
Average rate is for BST, the most highly subscribed CPST, a remote, and a converter. *
_s'gn'rficant at 95% level of confidence.

 

 

 

The average monthly rates for the competitive and noncompetitive groups
for the period of 1993 through 1997 are shown in Table 1. The differential in
monthly rates between competitive and noncompetitive operators was 8.4
percent in 1993. Due to the FCC’s rate roll-back, the differential decreased
significantly to 2.7 percent in 1994. The gap has increased from 3.1 percent on
July 1, 1996, to 5.2 percent on July 1, 1997, and to 6.3 percent on July 1, 1998.

Under the FCC' definition, effective competition was found where more
than one operator serves a community (the “overbuild“ test), where there is low
subscribership (the “low penetration“ test), where the system is owned by a
municipality (the “municipal“ test), or where the competing service provider is
owned by, or affiliated with, a local telephone company (the “LEC“ test). The
FCC reports that the average monthly rate charged by competitive operators
varies significantly according to which test was used to determine effective
competition but for all subcategories was lower than the average monthly rate
charged by the noncompetitive group. The differentials between the
noncompetitive group and each subcategory of the competitive group also vary
widely. For 1998, the regression equation results show that operators belonging
to the overbuild subcategory had rates that were 11.5 percent lower. Similarly,
operators belonging to the LEC test subcategory of the competitive group had
rates that were 15.5 percent lower than the noncompetitive group. The municipal
systems subcategory was found to charge rates that were 26.4 percent lower
than the noncompetitive group. On the other hand, operators in the low

penetration subcategory charged rates that were statistically indistinguishable

55

from the rates charged by operators in the noncompetitive group (FCC, 1999b).
The present study goes beyond previous research in several important
ways. It first attempts to discover consumer welfare in two dimensions—price
and quality—that overbuild competition may generate. From what is found, this
study will further attempt to make inferences as to whether overbuild competition
will be feasible and pervasive in the cable television industry. In so doing, it
seems necessary to find how incumbent cable operators respond to entry. To
make grounded inferences, a benchmark group of cable operators is included in
the analysis. Including municipal cable operators in the data sample and
differentiating competing systems according to a number of system
characteristics is expected to produce a richer story about the consumer welfare
effect and feasibility of competitive entry in the cable television industry.
Because price reductions are constantly reported as the most visible and
dramatic effect of overbuild competition, the primary focus here is placed on
price competition. Price competition also is probably the most critical, dynamic
element in the competitive struggle between incumbents and entrants in any
industry. Our search for the sustainability of overbuild competition begins with

the old, yet unresolved, debate on predatory pricing.

Thegries of Market Predation
The Chicago School and the “Chain Store Paradox“

It has long been recognized in industrial organization literature that the

response entrants expect from incumbent firms would be a major factor in

56

determining the attractiveness of entry, and much of the traditional literature on
entry deterrence effectively hinges on the threat of predation. Various types of
entry deterrence models tried to explain the dynamic interaction between the
dominant firms and fringe firms or the incumbent and new entrants. An old
question in industrial organization is whether an incumbent monopolist can
maintain its position by threatening to wage a price war against any new firm that
enters the market. The Standard Oil case at the turn of the century has been
considered a prototypical model of predatory price cutting. Standard Oil was
allegedly engaged in such anticompetitive conduct to drive its competitors out of
the market. This idea was heavily attacked by Chicago School economists such
as McGee (1958, 1980), Bork (1978), and Easterbrook (1981) on the grounds
that a price war would hurt the incumbent more than colluding with the entrant.

In his pathbreaking article dealing with the famous case of allegedly
predatory pricing by the old Standard Oil Company of New Jersey, McGee
(1958) attacked the Supreme Court decision in the landmark antitrust case of the
Standard Oil Company (Standard Oil Company of New Jersey v. U. S. 221 US 1,
47, 76) on two main grounds. First, there was little evidence to support the
contention that Standard achieved its monopoly position through predatory price
cutting. Second, predatory price cutting by Standard would have been irrational
because it demanded an excessive sacrifice of profits.

McGee (1958) argued, on the basis of the trial record in the 1911
Standard Oil case, that the old Standard Oil did not use predatory price cutting to

acquire or keep monopoly power. Instead, according to McGee, Standard used

57

acquisitions or mergers because they are always a preferred alternative to
predatory pricing. In large part because predatory price cutting imposes greater
costs upon predator than prey, he also doubted how rational and effective—and
therefore how frequent—predatory pricing would be in effectuating mergers, or in
general. McGee’s conclusion that monopoly achieved by the acquisition of the
rival is cheaper than monopoly achieved by the elimination of the rival in
economic war was modified by later contributors, including Telser (1966) and
Bork (1 978).

According to Bork (1978), a firm contemplating predatory pricing warfare
will perceive a series of obstacles that make the prospect of such a campaign
exceedingly unattractive. The losses during the war will be proportionally higher
for the predator than for the victim. The costs of predatory price cutting would be
prohibitively high for the predator because he faces the necessity of expanding
his output at ever higher costs, while the victim not only will not expand output,
but has the option of reducing it and so decreasing his costs.

Furthermore, three other considerations make such a strategy even more
unattractive and improbable (Bork, 1978, pp.144»155). First, the modern law of
horizontal mergers will make it all but impossible for the predator to purchase the
victim, so the campaign will have to last until the victim’s organization and assets
are dissolved. Second, ease of entry will be symmetrical with ease of exit. The
easier it is to drive a firm from the market, the easier it will be for that firm or
another to reenter once the predator begins to collect his monopoly profits.

Conversely, the more difficult entry is, the more difficult and expensive it will be

58

to drive a rival out. This symmetry of entry and exit means that predatory pricing
is a poor investment even if the predator has the reserves to bear the
disproportional losses required. Finally, while costs due to predatory pricing are
immediate, the anticipated monopoly revenues, being deferred, must be
discounted by the current rate of interest.

In summary, the Chicago school views predation by price cutting as
irrational, improbable, and inconsistent with the fundamental theory in
economics, that is, price theory. The theoretical argument presented by McGee
(1958, 1980), Telser (1965, 1966), Bork (1978), and Easterbrook (1981)
suggests that predatory price cutting is unlikely to exist. Easterbrook, for
example, refers to predation as both a unicorn and a dragon—animals much
talked about but not found in the real world. They contend, therefore, it is unwise
to construct rules about a phenomenon that probably does not exist, or which,
should it exist in very rare cases, the courts would have grave difficulty
distinguishing from competitive price behavior.

This view seemed to be confirmed by a game theoretic analysis by Selten
(1978), who theorized possible outcomes of competition in a game between a
multimarket incumbent and new entrants. Selten’s model involves a finitely
repeated game. He considers an incumbent monopolist in N geographically
separated markets (is, chain store). Firms serving several geographically
distinct markets are the prototype in the Chain Store game. The incumbent faces

N potential entrants, one in each market. The potential entrants must make their

59

entry decisions sequentially, one potential entrant making its decision each
period.

Suppose the one-shot game is repeated 20 times in the context of a chain
store which is trying to deter entry into 20 markets where it has outlets. The
outcome of this finitely repeated game would be much different from that of the
one-shot game because the chain store would fight the first entrant to deter the
next 19. Sherer and Ross (1990), for example, point to

“the demonstration effect that sharp price cutting in one market can have

on the behavior of actual or would-be rivals in other markets. If rivals

come to fear from a multimarket seller's actions in Market A that entry or
expansion in Markets B and C will be met by sharp price cuts or other
rapacious responses, they may be deterred from taking aggressive
actions there. Then the conglomerate’s expected benefit from predation in

Market A will be supplemented by the discounted present value of the

competition-inhibiting effects its example has in Markets B and C” (p.338).

In essence, a response that is predatory at some time in a particular
market may signal to would-be entrants the likelihood of equally rapacious
responses in other markets at later dates. Intuition suggests that the chain store
may predate in the first few markets to establish a “reputation“ for toughness and
thereby deter the remaining entrants. However, the Intuitive appeal of this
reasoning is not supported in a straightforward game-theoretic model in Selten
(1978). Selten first acknowledges, and it seems intuitively appealing, that in the
early rounds of this game the incumbent would adopt the costly predatory action

in order to persuade later entrants that they should best stay out, and that only

near the end of the game would it be willing to share a market. Yet Selten also

60

points out that this strategy cannot be consistent with the natural solution
concept (that of perfect equilibrium) to employ in such situations.

The repeated game in the Chain Store context is much more complicated
than the one-shot game. Selten relies on the backward induction approach to
solve the problem. A player’s decision of what action to choose will depend on
what actions both players took in each of the previous periods. In picking his first
action, a player looks ahead to its implications for all the future periods, so it is
easiest to start by understanding the end of a multi-period game, where the
future is shortest. Consider the situation in which 19 markets have already been
invaded. In the last market, the subgame is identical to the one-shot entry
deterrence game; the entrant will enter and the chain store will collude,
regardless of the past history of the game. In the next-to-Iast market, the chain
store can gain nothing from building a reputation for ferocity because it is
common knowledge that he will collude with the last entrant anyway. This
backward induction leads to the well-known “Chain Store Paradox“.

A standard technique for solving repeated games of this kind, the
“backward induction“ starts by asking what behavior would be rational in the last,
that is, N“, stage. Since there would be no more entries, what happens at stage
N can have no precedential effect, and so the only rational behavior at that stage
would be for the entrant to enter and the incumbent to acquiesce. If so, then at
the N-1th stage, what is done cannot influence behavior at stage N, so both
players would choose the rational strategies for that stage only—enter and

acquiesce. The same reasoning is iterated at stages N-2, N-3, and so on back to

61

stage 1. Thus emerges the paradox: if the participants foresee rationally what will
ensue at each stage all the way to the end of the game, acquiesce, not fight, is a
Nash equilibrium solution. The incumbent lacks incentive to discourage
subsequent entry, and as a result it ends up sharing its profits with entrants in
every market.

Apparently, Selten’s model is inadequate to justify Scherer’s prediction
that reputation effects will play a role. This inadequacy arises because the model
does not capture a salient feature of realistic situations. In practical situations,
the entrants cannot be certain about the payoffs to the monopolist. Rosenthal
(1981) investigated the original chain-store game and pointed out that the
paradoxical result in Selten’s analysis is due to the complete and perfect
information formulation in the model. He argued that in a more realistic
formulation of the game, the intuitive outcome will be predicted by the game-
theoretic analysis.

The chain store paradox seemingly supports the Chicago School’s
position that predation is irrational. This support is very limited, however. In other
market settings, predation emerges as a rational, profit-maximizing strategy
(Ordover, 1998). The challenge is to make predation for reputation a rational
strategy. Several possibilities arise. The goal of each is to make “backward
induction“ impossible. For example, the game may have no end from which to
reason backward. In such a model, the incumbent can in fact credibly threaten to
predate against eariy entrants and build reputation for toughness. More

Importantly, the players in the game may not have complete and perfect

62

information as assumed in the chain store game. I now turn to how this

modification can lead to quite distinct outcomes in the game.

The Harvard School and “Reputation Models“

McGee’s argument has been criticized on several grounds. First, the point
of the reputation models is that much of the benefit from an episode of predation
is the impact that it has on future entry or on entry in other industries. Indeed, if
the monopolizing firm shows a willingness to merge with any rival, it may face a
stream of entrants who enter just for the possibility of being bought out
(Rasmusen, 1989). Thus, the simple single-market calculus is inappropriate in a
multi-market or multi-entrant context (Ordover & Saloner, 1998). Second, there
are legal constraints on mergers. The elimination of the rival is most
advantageous when it results in a large increase in market concentration.
However, it is in these circumstances that a proposed merger is likely to violate
antitrust laws. In choosing between two unlawful methods of increasing its
market power, therefore, a firm may prefer predation to merger because the
fonner is much more difficult to detect (Posner, 1979).

The crucial point in McGee’s analysis of predatory pricing is that the
practice involves predator and victim in unnecessary loss of profits. Such loss or
sacrifice of profits is independent, however, of whether the deliberate price
cutting by the predator takes the price below cost (e.g., below its long-run
marginal cost or average cost). All that is necessary is that the price is taken to a

level lower than would otherwise prevail. In the duopollstic market situation which

63

is postulated, the initial price could be at any level, from the competitive price at
one extreme to the monopoly price at the other (Yamey, 1972). The aggressor
may be able to achieve its objective of eliminating or disciplining the rival and of
discouraging potential entrants by means of price cutting falling short of
predatory pricing. The aggressor has an obvious incentive of minimizing the
extent of its price cutting to achieve a particular result, and has a choice of
tactics. A smaller cut may in some circumstances be as effective as a larger cut,
especially when the rival has reason to suppose that the aggressor will go further
if necessary. On the other hand, a sharp initial cut may sometimes convey the
intended message more emphatically and achieve the intended result more
quickly.

Critics of McGee also point out that if the predator faces rivals in other
markets as well, it may be concerned about the effect of its pricing in one market
on its rivals in another market. Yamey (1972, p.131), for example, argued that
“the aggressor will, moreover, be looking beyond the immediate problem of
dealing with its current rival. Alternative strategies for dealing with that rival may
have different effects on the flow of future rivals.“

These contributions by Yamey (1972) and others gave rise to what may
be called “reputation models“. Reputation models simply assume that it is
feasible to drive the rival out and focus on a particular aspect of the profitability
of predation, namely the effect of reputation on future entry.

In fact, the “chain store paradox“ can be escaped if the sequence of

entries is neither clearly bounded in number nor ordered in time. More critically,

such paradox is resolved if the incumbent’s response to early entries alters
potential entrants’ assessment of the profitability of entry. If potential entrants
estimate the probability that an incumbent will fight additional entry on the basis
of the behavior they observe in connection with early entry attempts, then it may
pay the incumbent to react to early entries by fighting, despite the loss of profits
in the markets entered. The more often the incumbent responds by fighting, the
higher is the updated probability of fighting, as perceived by uncertain potential
entrants (Sherer 8 Ross, 1990). In this regard, Kreps and Vifilson (1982) show
that a relatively small probability that the incumbent will fight at the first stage is
sufficient to make fighting rational for the incumbent. In effect, the incumbent
fights to engender a reputation for toughness that deters additional entrants. If
the game has a finite number of well-ordered stages, the incumbents incentive
to react by fighting diminishes as the set of unchallenged markets declines, and
eventually, the incumbent will give way to some entrant, after which deterrence
will fail repeatedly.

Kreps and Wilson (1982) reexamined Selten’s model, adding to it a
“small“ amount of imperfect information about players’ payoffs. They found that
imperfect or incomplete information is sufficient to give rise to the “reputation
effect“ that one intuitively expects in the finitely repeated chain-store game or in
the finitely repeated prisoner's dilemma. “If rivals perceive the slightest chance
that an incumbent firm might enjoy ‘rapacious responses,“ then the incumbent’s
optimal strategy is to employ such behavior against its rivals in all, except

possibly the last few, in a long string of encounters. For the incumbent, the

65

immediate cost of predation is a worthwhile investment to sustain or enhance its
reputation, thereby deterring subsequent challenges“ (Kreps & Wilson, 1982,
p.254).

The two models presented by Kreps and Wilson are variants of the
Selten’s chain-store game. In flieir first model, a multimarket monopolist faces a
succession of potential entrants. But unlike Selten’s model, the entrants are
unsure about the monopolist’s payoffs in Kreps and Wilson (1982). This model
shows that there is a unique “sensible“ equilibrium where, no matter how small
the chance that the monopolist actually benefits from predation, the entrants
nearly always avoid challenging the monopolist for fear of the predatory
responses. The second model enriches this formulation by allowing, in the case
of a single entrant with multiple entry opportunities, that also the incumbent is
uncertain about the entrant's payoffs. The equilibrium in this model is analogous
to a price war: Since the entrant also has a reputation to protect, both firms may
engage in battle. Thus, each employs its aggressive tactic in a classic game of
“chicken,“ persisting in its attempt to force the other to acquiesce before it would
itself give up the fight, even if it is virtually certain at the outset that each side will
incur short-run losses.

Milgrom and Roberts (1982a) also take issue with those who doubt the
rationality and relevance of predatory strategies. Predation will emerge even if it
involves losses that cannot be directly recouped in the given market. The viability

of this predatory strategy does not depend on being able to induce rival’s exit.

Rather, all that is needed is that the predator be able to drive the rival’s return
from entry below that available elsewhere.

Milgrom and Roberts argue in their game theoretic analysis that if a firm is
threatened by several potential entrants, then predation may be rational against
early entrants, even if it is costly when viewed in isolation, because it yields a
reputation which deters other entrants. Predation emerges as a rational, profit-
maximizing strategy, not because it is directly profitable to eliminate the particular
rival in question, but rather because it may deter future potential entrants. The
mechanism by which this deterrent effect comes about is that by practicing
predation the firm establishes a reputation as a predator. This reputation then
leads potential entrants to anticipate that the incumbent firm will behave similarly
if they should enter, and hence, entry appears less attractive to them.

Milgrom and Roberts (1982a) contend that as soon as the complete
information assumption on the game is relaxed, the logic of Selten’s backward
induction breaks down. The possibility of actions taken in the past being a useful
guide to future behavior in similar situations now opens up, and with this,
reputation comes into play. Practicing predation now gives one a reputation as a
predator which is valuable in deterring entry. Thus, if entry occurs at an early
stage, it will meet a predatory response, because any other response
encourages further entry. Recognizing this, potential entrants at these early
stages will enter only if the market is so lucrative that they are willing to face
certain predation. Only as the horizon draws near and the number of markets

which may still be entered declines will the firm be willing to share a market.

67

One of the major properties in these models is that the value of a
reputation—and the costs one will be willing to incur to obtain it—has a positive
relationship with the frequency with which the reputation may be used, as
measured by the length of the horizon and the inter-period discount factor.

Asymmetric information plays a key role in Milgrom and Roberts (1982a),
since it provides the rationale for entrants to base their expectations of the finn’s
future behavior on its past actions. There are numerous reasons why this
element of uncertainty should exist. On the one hand, the entrants could be not
completely sure about the game being played. For example, it might be that the
incumbent firm could actually be involved in some bigger game so that the finn’s
equilibrium strategy in this bigger game calls for it to prey in earlier stages.

New entrants can by no means be certain that existing firms will give
ground. When potential entrants consider entering a market, they devote a good
deal of thought to finding an entry strategy that strikes the best compromise
between securing scale economies and minimizing the risk of price warfare. One
common strategy is to enter at a small scale with a plant that can be expanded
gradually when and if demand growth permits. This is done more readily in
industries with flexible, divisible production process than in continuous process
industries like petroleum refining and cement, where major equipment items may
come only in large chunks (Sherer & Ross, 1990).

The presence of informational asymmetries can lead a firm operating in
several markets to adopt a predatory strategy against entrants, even though

such behavior is irrational when viewed in the context of a single market in

68

isolation and even though there are only a finite number of potential entrants.
This same point is made in the one-sided uncertainty model in Kreps and Wilson
(1982).

Milgrom and Roberts (1982a) also recognized that predation will only
rarely need to be practiced. The credible threat of predation will deter all but the
toughest entrants, and so the occasions when the firm will be called upon to
cany out its threat will be infrequent.

In sum, these “reputation“ models demonstrate that predation is a rational
and optimal strategy in a repeated game with incomplete information. The
models enhanced our understanding of entry deterrence found in earlier studies
by Bain (1956), Wenders (1971), Spence (1977) and Dixit (1979, 1980), which
asked “what can the monopolist do prior to the entrant’s decision point to make
predation optimal in the short run?“ These earlier studies suggested, inter alia,
expanded capacity, limit pricing, sales networks and so forth. The relevance of
this question is that the threat of predation is only credible if predation is ex post
the optimal response, so the monopolist must make it so in order to forestall
entry. The new game-theoretical analyses by Kreps and Wilson (1982), Milgrom
and Roberts (1982a), and Kreps, Milgrom, Roberts, and Wilson (1982) suggest
that in repeated play situations, the actions taken by the monopolist need not
make predation actually ex post optimal—what they must do is to make
predation possible, and perhaps, increase the probability assessed by the

potential entrants that it is ex post optimal. If deterrence is the objective, as

69

Kreps and Vifilson (1982) argue, it is the appearance, not the reality of ex post
optimal predation, that may be important.

The Milgrom and Roberts model has the merit of showing that predation in
one market to protect another market is theoretically possible. However, it does
this at the cost of a number of restrictive assumptions, which it shares in part
with the Selten model. For example, there is the assumption that entry occurs
sequentially, in one market at a time by a different entrant. Consequently, if
predation occurs today, its aim is to prevent entry in a later period, not to prevent
simultaneous entry in other markets. Furthermore, their models require that
predation, once it has begun in any market, must continue in that market, so
formally they have a series of one shot games. Since giving up is not permitted,
predation, once started, is automatically credible for any market.

Easley, Masson, and Reynolds (1985) have managed to relax this
assumption. They allow for the possibility that entry may occur simultaneously in
several of the monopolist’s markets, possibly to stop or prevent predation.
Entrants may now enter new markets to force a rise in profits in previously
entered markets. They also allow for a monopolist to prey only for the gain of
extra time, that Is, to slow down entry in other markets, with full knowledge that
the predation will not preclude eventual entry into all of their markets. The
incumbent multimarket monopolist recognizes that it will subsequently cease its
predation and face complete entry into its markets. Despite this fact, it can attain
a greater present value by preying and retarding the rate of growth of

competition. The fact that eventual entry is not stopped does not mean that

70

predatory behavior has not occurred or that it has been unsuccessful. Profits can
be made from slowing the inevitable. Easely et al. (1985) demonstrate that
“successful predation may stifle competition without either eliminating
competitors or even stopping (some) competitors from entering!” (p.453).

Easley et al. (1985) defined predation as “the selection of strategy in any
entered market which does not maximize present value in that market when it is
considered in isolation, but which is selected for the purpose of slowing or
stopping future entry of equally efficient firms“ (p.448). This definition provides
for a very convenient tool for real world analyses. The price here needs not equal
or close to marginal cost or long run average cost, which oftentimes is not
available for academic research. In fact, the qualification of predatory pricing has
been a subject of long debates in industrial economics (Areeda & Turner, 1975;
Bork, 1978; McGee, 1980).

It is at times convenient to think of this strategy as a low price. Unlike “limit
pricing“ which will be discussed later, predatory pricing is only used in response
to entry. These acts serve to slow or stop entry and need not require elimination
of an entrant. They are knowingly selected in a fashion that does not maximize
the monopolist’s present value in each market separately. Thus, in Easley et al.
(1985), the predatory markets may yield non-negative cash flows. This means a
“long purse“ (T elser, 1966) or “deep pocket“ is not required for predation to have
anticompetitive effects in some markets in perpetuity.

In a model with slight modification to the original chain-store game, Phlips

(1995) provides a striking case where predation must occur. Suppose each

71

player is immediately informed about the moves of all other players, and thus
knows the history of the moves taken each point in the game by it and the other
players. Further suppose the history of the incumbent's payoffs is not known to
the potential entrants. In this game, predation must be practiced by the
incumbent. The basic intuition is that if the chain store ever failed to prey during
the game, all potential entrants immediately know with certainty that the
incumbent is not a fanatic predator. And then, the chain store paradox becomes

valid again.

Is Predation Rational in Cable Television?

How can these theories of market predation fit with the cable industry? To
check the plausibility of predation theories with respect to cable television, let us
recall some key elements in the Chicago school perspective. In this view, a
predator must lower price across the entire market rather than simply discounting
to those customers an entrant might serve. This places a large burden upon the
predator because predation implies pricing below marginal cost; hence per unit
losses are being sustained across an expanded level of output. In contrast, the
entrant, attempting only to take some market position, need not commit to
sustaining losses across the entire demand curve. This creates an asymmetry in
the losses sustained by the incumbent and the entrant. It is this disparity in the
distribution of losses during a predatory war which, in the Chicago school’s
viewpoint, makes predation such an incredible threat and, therefore, an unlikely

strategy.

72

But some industry-specific characteristics might affect the likelihood of
predatory pricing as a plausible tool. Industries that exhibit sunk costs, joint
capital, and network extemalities can be particularly susceptible to market
predation (Gabel & Rosenbaum, 1995). Take, for example, the
telecommunication industries—cable and telephone to be more specific—where
an incumbent firm operates “chain stores“ in multiple markets that share joint
capital. Suppose a rival firm enters only some of these markets. Pricing in the
shared markets below marginal cost imposes losses on the rival. The incumbent
firm, however, can offset those low prices with earnings from monopoly markets.
As long as revenues from all markets cover combined marginal costs and the
joint capital cost, the incumbent remains profitable. In short, the joint nature of
the capital imposes a cost disadvantage on the entrant relative to the incumbent.

When joint costs are sunk, the entrant is in a more tenuous position. To
overcome cost disadvantages due to the joint nature of the capital, the entrant
would consider entry into all markets. The sunk nature of the investment,
however, discourages such entry because the investment will be lost if entry is
unsuccessful. Therefore, the entrant is destined to be at a cost disadvantage to
the incumbent firm unless it undertakes the risky sunk investment of entry into all
markets simultaneously (Gabel & Rosenbaum, 1995). In essence, sunk and joint
capital costs make predatory pricing a very powerful, plausible and profitable
strategy for the dominant firm to oust its rival entrant from the market and deter A

future entry.

73

Gable and Rosenbaum (1995) illustrate that these problems of sunk and
joint costs, along with network extemality, enabled AT&T to successfully engage
in predation in the Madison, Wisconsin market at the turn of the 20th century.
The reward to AT&T was not so much in its opportunity to raise price after its
rival telephone company exited the market, as in the protection such predation
provided for AT&T’s remaining monopoly markets (Gabel & Rosenbaum, 1995).

Given many similarities between the telephone and cable industries, it is
no surprise to find the strategic (ab)use of fixed costs problems and first-mover
advantages in the cable industry (Barrett, 1996; Hazlett, 1990, 1995a). In his

attack on monopoly franchising, for example, Hazlett (1990) states;

“The existence of sunk cable plant, however, precisely identifies sub-
markets of entrants and thus mitigates the problem of infra-marginal
customers for the incumbent. Predatory price cuts need not be sustained
across all customers, and the losses of incumbent and entrant will diverge
in reverse: the entrant will sustain losses across its entire output range,
the incumbent across a limited slice of its output. Indeed, focusing on just
these segments where entry can physically occur, the incumbent is able to
employ a panoply of scale advantage, for example, in marketing and
brand name identifiability, further biasing competitive losses towards the
entrant“ (pp. 1 05-106).

Furthermore, capital fixity in cable television may offer increased
opportunities for the incumbent to engage in predation because it will likely
broaden the range of predatory pricing. As extension into sub—markets for cable
television service causes the entrant to make nontrivial and nonsalvageable
investments, these fixed costs can make the incumbent’s price, which is above

its average variable cost (AVC), lie below the entrant's average total cost (ATC).

74

In this case, an incumbent who sets prices equal to or above ATC may be able
to discourage an equally efficient entrant. This will facilitate predation by lowering
the cost of predation to the incumbent. In sum, “the easy geographical
identification and separation of customers, coupled with the necessity of bond
posting in specific capital, makes predation a relatively cheap defensive strategy“
(Hazlett, 1990, p.107) in the cable industry.

As discussed earlier, predation by incumbents in overbuild markets is an
empirical question. Despite the Chicago school’s attack on the likelihood of
predation in general (Bork, 1978; McGee, 1958), each industry will have a
different level of possibility for predatory pricing. In the cable television industry,
as noted by Hazlett (1990), predation may be highly likely because “the
existence of sunk cable plant precisely identifies submarkets of entrants and
thus mitigates the problem of infra-marginal customers for the incumbent“.
Contrary to Bork’s argument that the disparity in the distribution of losses due to
predatory pricing makes predation such a weak threat and hence an unlikely
strategy, predatory price cuts need not be sustained across all customers in
cable television markets. In fact, the entrant will have to sustain losses across its
entire output range while the incumbent can employ selective price-cutting
strategy across a limited slice of its output. Further, focusing on those segments
where entry occurred physically, the incumbent is able to employ an array of
scale advantages including marketing and brand name recognition, further

biasing competitive losses toward the entrant (Hazlett, 1990). Hazlett concludes

75

that potential and actual competition would be more prevalent were it not for
franchise restrictions and predatory behavior by incumbents.

This study is designed to investigate the effects of competition in general,
and the existence of predatory pricing in particular in the open entry era since the
passage of the 1996 Telecommunications Act. It is hoped that the investigation
of competitive interaction with respect to the price dimension will provide us with
a glimpse of the feasibility of overbuild competition, and ultimately market

structure, in the cable industry in the years to come.

Description of Methodology and Model

 

In the following analysis, dependent variables are monthly rates for basic
9885331933“ the \number of channels in the basic service package—Since
(cable systems maintain a different rate structure, it is necessary to standardize,
or define, this package for a comparable measure. To reflect competitive effects

on price more precisely, it seems warranted to measure monthly rates for the
standard basic package or tier that must include most popular cable
programming networks such as CNN, ESPN, TNT, and USA as well as
superstations (i.e., WGN and TBS). Occasionally, cable systems maintain a
mini-tier with superstations for an additional monthly charge, separate from the
standard basic service. In such instances, this monthly charge was included in
the standard basic service rate. In addition, a number of cable operators moved

the Disney channel from premium tier to the basic package in response to

competitive entry. Thus, the dependent variable in the price equations is monthly

76

rates for the standard basic package and the Disney channel (BDRA TE).
Another dependent variable (BDHRA TE) in the price equations is monthly rates
for the standard package which includes the Disney channel and a primary
premium channel, HBO. In very rare cases (in fact, only 2) where HBO is not
carried, the monthly rate for Showtime was used. These dependent variables
enter the equations in logarithmic form.

On the right-hand side of each equation are a number of exogenous
variables that have been found to influence the choice of price and quality
through their effects on the demand for cable television service, the marginal
cost of production, and the price-cost margin. More specifically, the model

estimates the following relation:

Price = f(lNCOME, sues, DENSITY, CHANNEL, PEG, TOPBMSO,
DUOPOLY, MUNICIPAL, POTENTIAL, DIGITAL, INTERNET, VOICE,
ONESTOP)

Quality = f(lNCOME, SUBS, ADI, DENSITY, TOPBMSO, DUOPOL Y,
MUNICIPAL, POTENTIAL, DIGITAL, INTERNET, VOICE, ONESTOP)

Complete definitions for the empirical variables included in these
equations are provided in Table 2.

77

\/

Table 2 Variable Definitions

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Variable Description Expected Sourcea
Sign
Regressand
BDRA TE Rate charged for basic service including the [1]
Disney channel
BDHRA TE Rate charged for basic service including the [1]
Disney channel & HBO
CHANNEL Number of satellite-delivered, cable-only [2]
channels, excluding broadcast signals, C-SPAN
l&ll, home shoppini and PEG access channels
CHANNELC Number of satellite-delivered, cable-only [2]
channels, including C-SPAN l&ll, home
shopping, and PEG access channels
CHANNELT Number of satellite-delivered, cable-only . [2]
channels, including broadcast signals, C-SPAN
l&ll, home shopping, and PEG access channels
Regressor
INCOME Median household income (S) +I+ [3]
SUBS Number of subscribers to basic cable service -/+ J4]
ADI ” ADI TV market ranking of franchise area -/+ [4]
DENSITY Homes passed divided by number of plant -/? [4]
miles
PEG Number of Public, Educational, Governmental +In.a. [1]
access channels
. TOP8MSO Dummy variable = 1 if firm is one of the top 8 -/+ [1, 4]
MSOs as of June 1999, 0 otherwise
OVERBUILD Dummy variable = 1 if multiple operators and -/+ [1, 2, 5]
overbuild areas, 0 otherwise
MUNICIPAL Dummy variable = 1 if system is owned & -I? [4, 6]
operated by a local municipality, 0 otherwise
POTENTIAL" Dummy variable = 1 if system faces “potential -/+ [5, 7]
competition,” 0 othenNise
DIGITAL Dummy variable = 1 if system offers digital 7/7 [1]
cable service, 0 otherwise
INTERNET Dummy variable = 1 if system offers high-speed ?l+ [1]
lntemet access service, 0 othenNise
VOICE Dummy variable = 1 if system offers telephone 7!? [1]
service, 0 othenIvise
ONESTOP Dummy variable = 1 if system offers all of -/+ [1]
conventional cable, lntemet access, and
telephone services, 0 otherwise
' Sources:

[1] Telephone survey

[2] Online survey (wwww. tvguied. com, tv. yahoo. com, msn. Clicktv. com)
[3] US Department of Commerce (1994). CounLnd City Data Book
[4] Warren Publishing (1999). Television and Cable Factbo_oi_<__

[5] FCC (1999). Community Unit Reference UT“

 

78

[6] American Public Power Association (1999). Annual Directory 8. Statistical Report
E7] Various trade journals including Multichannel News, Cable World, and Cablevision
ADI TV market rankings were converted to ADI in the following manner (see Mayo and
Otsuka, 1 991 ):
= (103 — ADI) if the cable system is located in the top 100 ADI TV markets
= 2 if the cable system is located below the top 100 ADI TV markets
= 1 if the cable system is located outside ADI TV markets

° Potential competition” was operationally defined as one of the following three
conditions:
(1) A new firm has been granted a cable franchise, but not activated as of survey
time
(2) An identifiable firm has been in the franchise negotiation process
(3) The franchise area has been reported in major trade journals that an identified
firm was considering entry in the market, 0 otherwise

79

Estimating the above models amounts to a simple, reduced form
comparison of monopoly and duopoly rates on the one hand, duopoly and
municipal on the other. No attempt was made to recover structural hypotheses in
the present study. Since the primary interest of the present study is to determine
how price and quality choices vary with market structure and ownership type
(and not particularly concerned with estimating such structural parameters as the
own-price elasticity of demand), the analysis of data will focus on estimating
reduced-fonn equations for the determination of price and quality.

Because direct data on the marginal cost of basic service were
unavailable, the model includes several variables that are believed to determine
marginal cost in the price equations. These include the number of basic
subscribers (SUBS), a measure of population density (DENSITY), the number of
channels offered by the cable firm (CHANNEL).

To the extent programming cost is related to service price, the number of
cable-only channels delivered via satellite would have a direct effect on
subscription fees for basic cable. In previous studies, total number of satellite-
delivered cable channels has been used to control the quality of cable service. In
this study, the basic service package is more specified according to each
channel’s contribution to programming cost. In general, a basic service package
is the bundle of multiple programming products that have different attributes in
their relationships to cost functions for a local cable system. According to distinct
cost functions for each channel, a basic cable package may be divided into

separate groups. Most cable channels are carried on the local system for a

80

license fee. Most popular cable channels (in terms of audience rating) such as
USA, ESPN, TNT, and CNN, are in this group. But some cable channels do not
incur any programming cost due to license fees. For example, home shopping
channels (e.g., Home Shopping Network, QVC, and Valuevision) do not charge
the cable operator license fees for their programming. Instead, these
programmers make payments to the system operator, as a form of commission,
for each order originated in the system’s franchise areas. In other words, the
monetary payment is in reverse: home shopping networks pay a commission to
system operators in return for the carriage of their programming. No license fees
are necessary for public, education, government (PEG) access channels,
although these channels, too, are included in the basic package. PEG access
channels, however, may have a positive effect on programming cost for the
cable system operator because these channels will require programming
personnel and facilities such as studios and equipments (Zupan, 1989a, 1989b).
Yet another distinct relationship is found with broadcast signals and C-SPAN l
and C-SPAN II. The point is that it would produce more precise estimates of
programming cost to exclude all these channels from basic package. Thus,
variable CHANNEL here is the number of satellite-delivered, cable-only
channels, but not broadcast signals, home shopping, C-SPAN l, C-SPAN II, and
PEG access channels.

According to (Zupan, 1989a, 1989b), nonprice concessions such as
franchise fees and community programming significantly raise cable costs.

Among various types of nonprice concessions, Zupan found that community

81

programming on access channels accounted for about 4 percent of the total
annual operating costs. Thus, PEG access channels may have a positive
relationship with price of cable senrices, while their contribution to revenue
generation is invisible. To test this effect on price, it is necessary to separate
PEG access channels from the basic cable package. Variable PEG is the
number of public, education, government access channels.

The primary measure of cable system size employed here is the number
of subscribers (SUBS) to basic cable service. It is the output level of the system
that may have an important bearing on the average cost, hence price ultimately,
for cable services. However, it should be noted that the size of local cable
systems measured in terms of subscribers may not have a significant effect on
consumer price for two reasons. First, the extent of scale economies in cable
television service has been challenged by both academic research (Noam, 1985;
Owen & Greenhalgh, 1986) and numerous entries by small cable systems
around the country. Second, the number of subscribers would be a poor
measure of programming cost per subscriber at the local level because license
fees are paid on a subscriber basis to the cable programming networks. In
contrast to previous studies by Ford and Jackson (1997), therefore, the average
programming cost is the same on all level of system size measured in the
number of subscribers at the local system level. Only true fixed costs incurred for
headend facilities may be reduced as the system size increases. But its effect,

too, may be mitigated by diseconomies of scale. In a study by Ford and Jackson

82

(1997), the number of subscribers to the local cable system was found to have
no effect on programming costs and prices.

Indeed, programming cost per subscriber in cable television service may
be related to the total number of subscribers at the national MSO level. Chipty
(1995) found tremendous bargaining power of big M803 and its significant effect
on programming costs. Dertouzos and WIldman (1998), Ford and Jackson
(1997), and General Accounting Office (1999a) all revealed that big MSOs
enjoyed a very stiff “volume discount” for programming fees. For example, one
cable channel cost $0.02 per subscriber for MSOs with more than 3 million
subscribers, $0.22 per subscriber for MSOs with between 500,000 and 3 million
subscribers, and $0.29 per subscriber for MSOs with fewer than 500,000
subscribers (Ford & Jackson, 1997). Despite the significant discounts for large
MSOs, previous studies (FCC, 1994a; Jayaratne, 1996; Thorpe, 1985) found
that they actually Charged higher rates than small M803 and independent
systems. Jayaratne (1996) attributed this to higher quality programming and
associated higher programming costs by large MSOs. To capture this MSO size
effect, variable TOP8MSO is included in the equations to be estimated. As of
1999, the biggest 8 MSOs in the country collectively serve more than 60 percent
(42 million homes) of the total wireline cable subscribers. Adelphia
Communications, the eighth largest MSO, had about 2.3 million subscribers
(Cablevision, May 17, 1999, p.43).

Income levels (INCOME) in the franchise areas have to do with cable

service price in two ways. First, it may have a positive relationship with cost for

83

cable service. Some prior studies, including Emmons and Prager (1997) and
Ford and Jackson (1997), used the median household income as a proxy for
local wage rates as they may be positively related to marginal cost. However,
they failed to control the effect of income level in the community on cable price
because they used income at the county level. There are tremendous variations
in income among towns even within a county. For example, it is not well devised
if one tries to capture an effect of income on price in the Perrine, FL area
($30,481) by using median household income for Dads county ($16,925). In this
study, the income level is controlled by using the median household income at
the local community level to avoid this problem. Where a cable system serves
multiple communities, the median household income for the system’s primary
operation is used. Another problem is to use the income level as a proxy for
wages. Because the operation of a cable system is not labor-intensive, and thus
wages do not occupy much in expenses, local wage rates would not produce a
meaningful effect on marginal cost. Thus, it seems more appropriate to conceive
income levels as a proxy for the intensity of demand in a particular community.
Prior studies have found that cable subscribers tend to have a higher income
level (Baldwin, McVoy, 8 Steinfield, 1996).

Marginal cost for cable services decline as the number of homes passed
by a cable system increases in a given area. The cost of adding a subscriber to a
cable system generally declines as more homes are wired per mile by coaxial
cable. DENSITY is defined as the number of homes passed, divided by total

cable plant size measured in miles.

Despite theoretical debates of predation and case studies supplied by
Hazlett (1990), Barrett (1996), and trade journal reports, the issue of predatory
pricing has remained unsolved largely because these observations were refuted
as an anecdote. But more importantly, in order to show predation, a researcher
must have access to various cost information including long-run average cost
and marginal cost for both the incumbent and the entrant. Even with sufficient
information in a few cases, it would be very difficult for researchers to conduct
the systematic analysis of the prevalence of predatory pricing as a competitive
strategy.

Faced with these obstacles, the present study attempts to take a different
approach. First, a regression analysis of the 247 cable systems allows the
researcher to discover price difference between monopolists and duopolists. As
consistently evidenced in all previous studies, overbuild markets are expected to
charge significantly lower prices for basic and premium channels. To measure
the effect of overbuild competition, the equations include a dummy variable
OVERBUILD defined as a situation in which cable television service is offered by
two or more wireline cable systems in direct head-to-head competition within the
same service area. In the absence of cost information, however, it would be
almost impossible for the researcher to determine whether or not price
differences, if any, would indicate the existence of predation. To determine this,
another group of cable systems was included in the sample. A group of 33
municipal cable systems (MUNICIPAL) will be used as a comparison variable,
and the comparison of its coefficient to that on OVERBUILD should help the

85

researcher determine the intensity of price wars in the overbuild areas. So, the
group of municipal cable systems will provide a benchmark test to the extent that
these systems are providing their residents with cable services, not for economic
“profits”, but for other reasons. It is assumed, as in the Emmons and Prager
(1997), that privately owned cable system operators choose the price and quality
of basic cable service in order to maximize profits, taking as given the demand
for their product, the technology, the prices of inputs, and market structure.
Nonprivate, city-owned cable systems are permitted to Choose price and quality
levels that differ from those offered by privately owned firms, to allow for the
possibility that they pursue alternative objectives, such as maximizing consumer
welfare. Variable MUNICIPAL is assigned a value of 1, if the system is owned
and operated by a city.

To utilize this comparison effectively, it is necessary to make a few
assumptions on the behavior and motivation of municipal cable systems. Indeed,
it is conceivable to assume that municipal cable systems will behave differently
in many ways. First, many cities operate their own utilities operations (e.g.,
electricity and water). Most city-owned cable systems share some principal
elements of cable networks such as poles, ducts, and conduits with other city-
owned operations. The sharing of key network facilities may lead to cost
reductions for municipal cable systems. Second, prices for cable services will
also likely deviate from those of private monopolies and duopolies because the
primary objective of city-owned cable systems is not “maximizing profits”.

Virtually all managers for municipal cable systems revealed in telephone

86

interviews with the author that their motivation of municipal cable systems was
not so much “making money” as “enhancing residents’ satisfaction” with cable
services. As such, their concern is “break even” rather than “maximizing profits”.
The performance of municipal cable systems will play two critical roles in this
study. First, it should help the researcher observe price deviations between
private and non-private systems. Second, it will shed light on strategic use of
price by the incumbents. As a benchmark, prices charged by municipal cable
systems will be compared to those of incumbents facing a new entrant as well as
monopolists.

Another interesting variable included in the equations is the threat of
“potential competition”. Contestability theory suggests that where entry is
possible, even “potential entry“, not active, can still form an important competitive
constraint on incumbent behavior (Baumol, Panzar, & erlig, 1982). In the
presence of such a potential threat, a monopolist will exercise restraint in
contestable markets. Due to the number of strong assumptions such as “no fixed
costs” and “zero exit costs“ which the contestability theory in its original form is
premised on, it would be hard to apply the contestability principle to the
telecommunication industries that typically require both significant fixed and exit
costs. But the idea may still be tested if it is carefully crafted.

In this study, the threat of “potential competition“ is embedded as a
dummy variable. “Potential competition“ (POTENTIAL) is present if a local cable
system faces one of the following three conditions: (1) a new firm has been

granted a cable franchise in a market in question, but the system was not active

87

as of survey period (January 2000); (2) a specified and identifiable firm has been
in the process of franchise negotiations with a local franchising authority; (3) a
market has been reported in major trade journals that an identified firm was
considering entry into a specific market. Thus, if a market met any of these
conditions, the incumbent was considered to face “potential competition“.
Models of industrial organizations focus almost entirely on economies of
scale as the traditional source of efficiencies in most industries. Yet there are
other related factors that may be important sources of cost savings as well for
the multi-product firm. A growing body of economic literature pays more attention
to cost savings that arise from simultaneous production of different outputs over
shared facilities. These economies of scope may materialize when there are
special cost savings that accrue to a multi-product firm that are unavailable to
firms producing only one of the products but not both. In short, the single firm
can supply a bundle of products demanded by the market at a lower total cost
(and price) than some combination of two or more single-product producers.
Such special synergistic effects could arise from: achievement of new markets,
stronger combined product lines, buying raw materials in large volumes,
eliminating duplicate general and administrative functions/departments, and risk
reduction in capital markets. To capture this subadditivity factor and reflect
recent development in the cable industry, the price equations include three
products dummy variables: INTERNET, VOICE, and ONESTOP. If a cable
system provides cable modem service as well in its franchise areas, INTERNET

is assigned a value of 1, otherwise 0. Large cable MSOs have begun to provide

88

local telephone service within their service areas. A few private or municipal
cable systems also operate their own telephone services in small towns. VOICE
is a dummy variable to investigate cost savings or lack thereof from the joint
operation of cable and telephone networks. Yet another group of MSOs, most
notably MediaOne, Comcast, and Cox, is moving toward the provision of a
complete package of telecommunication senrices from a single source. These
companies claim that consumers will be offered the convenience of “one-stop
shopping”. To capture the effects of (dis)economies for these multi-product firms,
a dummy variable ONESTOP is included in the equations. If a cable system
provides cable modem service as well as telephone service in its franchise

areas, ONESTOP is assigned a value of 1, otherwise 0.

Description of the Data

The following analysis is based ori_a__sample o1247 cable SY$t9fD§§EfE§§
the country. The sample comprises three subsamplesfirst, this study began
with a group of 91 cable systems that are under overbuild competition. This
competing subsample includes all systems that could be identified as facing
direct competition from another cable operator in some portion of their franchise
territory. The researcher collected information on market structure in each local
market included in the sample, using various sources. The starting point for the
identification of overbuild competition in the local market in question was the

FCC's large compilation called the Community Unit Reference List, which

contains more than 30,000 US communities and cable systems in each

89

community. This archive contains very basic information on cable systems at the
franchise-granting community level. Available in this list are community and
county of service area, type of cable system operation, name of system operator,
system identification, and beginning year of service. Among 30,000 communities,
about 300 communities were found to have granted more than one cable
franchise. Because a single cable system usually provides service in surrounding
communities as well and operator specific information is available at the system
level, not at the community level, the total number of competing systems
available for research reduces to 119 cable markets. The problems of missing or
incomplete infon'nation further reduced the number of competing markets to 91
for the present study.

The municipal subsample consists of all nonprivate, city-owned cable
systems. According to the 1999 Annual Directory & Statistical Report published
by American Public Power Association, there were 76 cable systems that were
owned and operated by municipal government as of 1999. Among these 76
systems, all municipal cable systems serving more than 1,000 subscribers were
included in the sample. These amounted to 33 systems. But one municipal
system was unable to provide information on some system characteristics and
was dropped in the analysis. The private monopoly subsample includes 150
randomly chosen cable systems that face no direct competition and are privately
owned. Among these, 26 systems were dropped because of incomplete data,

leaving 124 monopoly cable systems. Thus, the total number of observations

90

included in the regression analysis was 247 systems encompassing 124
monopoly, 91 duopoly, and 32 municipal systems.
Data on cable system characteristics were collected from the 1999 edition

of the Television and Cable Factbook. Demographic data on franchise areas

————-——

\_‘_‘ '

were collected from US government publications, including the County and City

 

Data Book and Survey of Cunent Business.
The following empirical analysis is based on cable system data collected

from five different sources. First, telephone surveys were used to gather

 

information on prices for standard cable service and one premium channel
(HBO). The telephone surveys also checked if the cable system provided high-
speed lntemet access and telephone services other than the plain old cable
service in its franchise areaswfeiond, the Community Unit Reference List file
compiled by the FCC was used to find out the existence of overbuild competition.
Third, information on channel lineups on each cable system was collected
through online television listings services such as TV Guide Online
(www.tvguide.com), Gist TV by Yahoo (tv.yahoo.com), and Microsoft’s intemet
service MSN (msn.clicktv.com). These online services are preferable because
not only do they provide most updated information on cable programming in
each segment by zip code within a cable franchise area, but they also allow the
researcher to confirm the FCC data on the presence of overbuild competition.

Fourth, the 1999 edition of Television and Cable Factbook provided other system

characteristics such as the number of subscribers and plant size. Fifth, the

91

County and City Datebook published by the Department of Commerce was

utilized for household income levels.

Descriptive Analysis

Table 3 presents descriptive statistics for each variable. The first four
variables show that the sample in the present study represents the cable industry
very well. The average number of subscribers is 46,397 out of 74,994 cable-
available homes. This amounts to the average penetration of 68.3 percent. Paul
Kagan Associates, Inc. shows that basic cable penetration was 68.9 percent as
of 1998 (NCTA, 1999). The average number of homes per mile, a market density
measure, is 92.5, which is almost the same with the national average of 93
homes per mile, as reported by Paul Kagan Associates, Inc. These system
characteristics are very close to those used in simulation studies by Smiley
(1986) and by Hazlett (1995b). This similarity should allow for the direct
comparison of the following econometric analysis with findings in the prior
research that was based on hypothetical scenarios.

As of January 2000, the average monthly rate for basic service including
the Disney channel was $31.39, whereas the typical cable subscriber was paying
$11.57 for HBO. On average, cable consumers received about 42 cable-only
channels. Additionally, they were offered a little more than 2 PEG access
channels.

More than half (55 percent) of the systems in the sample were affiliated

with the nation’s biggest 8 M803 and provided digital cable services. lntemet

92

access service over cable (cable modem) was available on half of the systems
included in the sample, while local telephone service was offered in about 11
percent of the 247 systems. About 10 percent of these systems provided a
complete package of video, data, and voice services for the convenience of

“one-stop” shopping.

Table 3 Descriptive Statistics of the Sample

   

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Variable S.D. Minimum Maximum
SUBS 46397.16 78413.25 401 585000
HOMES 74994.35 139413.10 625 1201093
PRATE 68.25 15.06 10.68 99.65
DENSITY 92.52 70.33 21.90 562.13
INCOME 29031 .75 9808.33 12573 63976
BDRATE 31.39 6.94 15 53.89
HBO 1 1 .57 1 .81 2 17
CHANNEL 41.62 9.73 13 59
PEG 2.33 2.04 0 1 1
TOP8MSO .5543 .4980 0 1
OVERBUILD .3696 .4837 0 1
MUNICIPAL .1284 .3352 0 1
POTENTIAL .0973 .2969 0 1
DIGITAL .5469 .4988 0 1
INTERNET .4941 .5009 0 1
VOICE .1137 .3181 0 1
ONESTOP .0980 .2980 0 1

 

 

 

 

 

 

Econometric Results

Competitive Effects on Price

The estimates of the price equations are presented in Table 4. Both price
equations and quality equations—which will be presented later—were estimated
using the log-level model. The results for the two equations are very similar in
terms of each model’s overall explanatory power and each variable’s effect. The

model shows that INCOME is a significant factor for both basic sen/ice and

93

premium channels, although its effect is economically very small. Note INCOME
is measured in thousands of dollars and its coefficient suggests that a $1,000
increase in the median household income in a community is predicted to
increase basic service price by only .2 percent. Similarly, the number of
subscribers (SUBS) is also significant, but its effect on price is negligible. In fact,
its coefficient says that the addition of a thousand subscribers would increase
price of basic service by .4 percent, which does not provide any significant
economic meaning. Nonetheless, it is worth noting that the relation between
SUBS and BDRA TE and BDHRA TE is positive and statistically significant.
Contrary to the general notion of scale economies in cable television, the system
size measured in the number of subscribers does not produce any price benefits
for the consumer, holding other factors constant. It is not surprising in cable
television, considering the disparity between cost and price (Ahn, 1997). This
finding in fact is consistent with previous studies by Noam (1985) and Owen and
Greenhalgh (1986). The coefficient on TOP8MSO may reveal another example
of the disparity between cost and price in cable television. As discussed above,
the biggest MSOs have a strong bargaining power over the cable programming
networks and they receive substantial amount of “volume discounts“ vis-a-vis
mid- or small-sized MSOs (Chipty, 1995; Dertouzos 8. Wildman, 1998; Ford &
Jackson, 1997). But these MSOs charge, ceten's paribus, 5.6 percent more than
smaller cable systems [exp(.0553) =1 .0568]. This result is consistent with the
findings of Thorpe (1985), FCC (1994) and Jarayatne (1996).

As expected, the number of channels is a statistically very significant
factor for cable service price. The addition of one more channel is predicted to
increase basic service price by .75 percent, which is a nontrivial effect size in an
economic sense.

On the other hand, the coefficients on PEG suggest that the number of
PEG access channels is negatively correlated with price. Although their effect on
BDRA TE is statistically insignificant while the effect on BDHRA TE is significant at
the .10 level, its effect size is about .8 percent per each additional PEG access
channel. It is clearly not negligible and the direction of its effect is against the
expectation. It is difficult to interpret this result. One plausible explanation is a
spurious correlation. In other words, the number of PEG access channels might
have a strong correlation with another attribute of the cable system or of the
community that this study did not intend to capture. One of those attributes may
have to do with the influence of the local franchising authorities. The number of
PEG access channels are basically determined in the franchise negotiation
process and franchise authorities try to reserve as many access channels as
possible for the public, educational, and governmental use. In franchise
negotiations in the past, cable operators tended to promise many nonprice
concessions including community programming channels in return for an
exclusive franchise. Obviously, more powerful franchise authorities demand
more access channels regardless of actual usage. This bargaining power may
also be exerted to maintain the low price in a community where the local

franchising authorities have regulatory control over price. Even though the rate

95

regulation of cable service was sunset in March 1999, its impact may still remain
due to path dependence. In short, there might be a spurious correlation between
the number of PEG access channels and price such that another third attribute
(e.g., the influence of franchise authorities) might simultaneously determine both
access channels (upward) and the price level (downward) in a community.

The coefficients on competition and nonprivate ownership variables are all
large, negative, and statistically very significant in both equations. First, consider
OVERBUILD where two wireline cable operators compete in the same market.
Overbuild competition leads to a price reduction by 20.7 percent [exp(-.2324) -
1=-.2073]. This effect has been consistently found in prior studies. Second,
municipal cable systems offered basic cable service at prices that were 25.7
percent [exp(-.2969) - 1=-.2568] lower than those charged by monopoly
operators. This is comparable to Emmons and Prager’s 22.6 percent price
differential in 1989 between municipal and monopoly cable systems (1997).

The coefficients on POTENTIAL are not only significantly different from
zero at the .05 level, but the effect size is quite large. They suggest that
competitive threats from a “potential entrant“ lead to a price of basic cable
senIice that is 7.7 percent lower than that of monopolists. This differential may
be an indication of “limit pricing“ by incumbents in an effort to prevent actual, full-
blown entry. The incumbent may try to signal that post-entry equilibrium would
not be profitable for “potential entrants” and that they would face price warfare

should entry occur.

Table 4 Estimates of Price Equations

Dependent Variable:

Dependent Variable:

 

 

 

 

 

 

 

 

 

 

 

 

 

 

In(BDRA TE) In(BDHRA TE)

Independent Coefficient t-statistic Coefficient t-statistic

Variable (Std. Errog (Std. Error)

INCOME (in .0022221 1.792* .0021535 2.181”

thousands) (.0012399) (.0009875)

SUBS (in .0004058 2.437“ .0002997 2.259"

thousands) (.0001665) (.0001326)

CHANNEL .0074285 4807*” .0057592 4679*“
(.0015454) (.0012308)

PEG -.0087930 -1.424 7 -.0081602 -1.659*
(.0061751) (.0049179)

DENSITY .0001626 .995 .0001974 1.517
(0001634) (.0001301)

TOP8MSO .0553969 2.047“ .0364250 1 .690
(.0270650) (.0215550)

OVERBUILD -.2324208 8992*” -.1873848 -9.103***
(.0258471) (.0205850)

POTENTIAL -.0804296 -2.062** -.0633207 -2.038**
(039003) (.0310661)

MUNICIPAL -.2969203 8149*” -.2703530 9317*”
(.03643®) (.0290186)

DIGITAL -.0319171 -1.211 -.0176349 -.840
(.0263503) (.0209858)

INTERNET .0456307 1.758' .0403348 1 .952"
(.02595@) (0206672)

ONESTOP -.0860918 -2.085** -.0827502 -2.516**
(.0412902) (.0328841)

Constant 3.1620770 55.617*** 3.533274 78.032***
(.0568545) (.0452797)

 

 

 

 

 

F (12, 234) = 17.07
Prob. > F = 0.0000
R 2 = .4667 (n = 247)
Adj. R 2 = .4394

 

F (12, 234) = 19.30
Prob. > F = 0.0000
R 2 = .4975 (n = 247)
Adj. R 2 = .4717

 

Note: *, **, *” indicate values that are significantly different from zero at the .10, .05, and
.01 levels, respectively.

97

 

It might be somewhat surprising to see this result given the fact that the
incumbent may launch price warfare at any time alter actual competition takes
place in its service areas. But it should be noted that this 7.7 percent differential
is not a price reduction from a pre-entry monopoly price level. Instead, it is a sign
of price-checking power exerted by the threats of potential competition. That is,
when “potential competition“ exists in its service areas, an incumbent would
restrain from using its monopoly power, while others still exploit their immunity
from competition to increase price year after year. Where actual competition is
present, this trend has been reported frequently in the cable industry (FCC,
1997; Hogan, 1998b). In Sacramento, California, for example, Sacramento
Cable Television (the incumbent operator) launched extensive remarketing
campaigns in the neighborhoods that were expected to be overbuilt before a new
entrant started its service, offering free installation and 3 months of free service
(Hazlett, 1995a). The price-constraining power of “potential competition” would
be very likely, particularly when entry is imminent. Oftentimes, dramatic price
cuts right after competition also bring about consumer backlash since consumers
and franchising authorities tend to question the credibility of pre-entry prices.
Recall the operational definition of “potential competition“ in this study. One of
the three conditions to meet the definition of “potential competition” was that “a
new firm has been granted a cable franchise in a market in question, but the
system was not active as of survey period”. Clearly, this does not mean “actual
competition“, but does suggest competition is “highly likely and credible and

imminent”. To the extent that this definition correctly measures “potential

98

competition“, the econometric results suggest that “potential competition“ plays a
significant role in keeping price in check, if not reducing it.

One way to cope with the squeezed profit levels due to competitive forces
is to expand the line of services using the same cable infrastructure. Another
interesting group of variables in the regression analysis attempted to measure
whether or not cable prices are reflected with (dis)economies of scope. While
statistically insignificant, price of standard (analog) basic service charged by
cable systems offering digital cable packages (DIGITAL) is about 3 percent lower
than that charged by systems that provide only traditional analog service
packages. This may reflect operators’ pricing strategy that attempts to
“recapture“ channels from analog to offer digital tiers. On the other hand, cable
systems that also provide lntemet access service (INTERNET) over cable are
predicted to Charge 4.6 percent more than systems that do not offer such a
service. And this difference is significantly different from zero at the .10 level.
This seems to imply that there exists cross-subsidization from the cable
operation to the high-speed lntemet service within a given service area. While no
concrete evidence of cross-subsidization is available, there is no reason to
believe that diseconomies of scope play a role in increasing cable price. A more
plausible explanation is, therefore, cable operators utilize their monopoly profits
from the plain old cable television operation to market such new services as the
lntemet access service through the internal transfer of profits (Baldwin et al.,
1996). Interestingly, however, if a cable system provides lntemet access as well

as telephone services (ONESTOP), the system is predicted to maintain a price

99

for cable service that is much lower than othenIvise. This is statistically significant
at the .05 level, and may suggest the existence of economies of scope in the
production of multiple services over the cable network. This may have a
significant implication for synergistic effects from the efficient use of cable
networks. A dummy variable VOICE did not enter in the equations because of

the multicollinearity problems with INTERNET.

Competitive Effects on Service Quality

Table 5 presents the estimates of quality competition. In estimating the
effects of overbuild competition, two dependent variables are used as an
indicator of service quality. The first, In(CHANNELC), is the number of cable-only
channels in the standard basic service tier including satellite-delivered channels,
PEG access channels, and home shopping channels. The second,
In(CHANNELT), is the total number of channels in the standard basic service tier
including broadcast channels as well as all the cable-only channels. Satellite-
delivered cable-only channels are costly to carry since cable operators have to
pay a license fee on a per-subscriber basis. In case of broadcast stations, cable
operators have to make investments in channel capacity to make room for local
and distant signals. PEG access programming requires not only channel space
but also personnel and studios. Although much of the broadcast signals and
access channels are mandated by either the federal must-carry rules or local
franchise requirements, cable operators can choose to import distant signals and
to increase access channels. In fact, competition seems to give rise to the

importance of community programming as cable operators strive to differentiate

100

their services from those offered by overbuilders and DBS providers. It is
presumed here, as in prior studies (Emmons & Prager, 1997; Jayaratne, 1996),
that increased channel offerings represent greater service quality to cable
subscribers.

Economic theory predicts that a monopoly system operator will add cable
channels up to the point where the incremental revenue from increased demand
for cable service stimulated by enhanced service quality equals the cost of the
last added channel (MR=MC). If two operators compete in the same market,
each would perceive that increased quality would attract not only more
households to subscribe to cable, but would also draw current subscribers away
from its competitor. Hence, overbuild competition is likely to lead to more
channels than would a monopoly (Jayaratne, 1996).

Of course, adding channels cannot be done overnight. The decision to
increase the number of channels will be impacted by a system’s channel
capacity and oftentimes will require system rebuild. Thus, it is a long-term effect
rather than short term.

The median household income (INCOME) is a significant predictor of
CHANNELT, but not statistically significant in explaining CHANNELC. Although it
may suggest that cable service quality would be better in high income
communities, its effect is not so large. On the other hand, the number of off-the-
air (OTA) signals available in a community, measured by ADI ranking, is
significantly related to the cable operators’ decision of how many channels would

be carried on their system.

101

Table 5 Estimates of Quality Equations

Dependent Variable:

Dependent Variable

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

In(CHANNELC) In(CHANNELT)

Independent Coefficient t-statistic Coefficient t-statistic

Variable (Std. Error) (Std. Error)

INCOME (in .0021490 1.222 .0036156 2.426“

thousands) (.0017581) (.0014906)

SUBS (in .0002950 1 .440 .0002445 1.408

thousands) (.0002049) (.0001737)

ADI .0010591 2.526“ .0013084 3680*“
(.0004192) (.0003554)

DENSITY .0004261 2.073” I .0004794 2.750“
(.0002056) (.0001743)

TOP8MSO .0863209 2.581 “* .0618432 2.181”

(0334436) (.0283542)

OVERBUILD .1072472 3.429“ .0854082 3.221 “*
(.0312778) (.0265180)

POTENTIAL .0857553 1.771* .0958863 2.336“
(.0484188) (.0410505)

MUNICIPAL -.0031513 -0.069 -.0018542 -0.048
(.0454903) (.0385677)

DIGITAL .0414891 1 .252 .0385075 1 .370
(.0331491) (.0281046)

INTERNET .1332212 4293*“ .1054631 4008*“
(.0310325) (0263100)

ONESTOP .0688678 1 .355 .0729323 1 .692*
(.0508280) (.0430931)

Constant 3.387049 67.591” 3.605559 84.867“
(.0501106) (.0424849)
F (11, 235) = 17.97 F (11, 235) =23.14
Prob. > F = 0.0000 Prob. > F = 0.0000

R 2 = .4569 (n = 247) R 2 = .5199 (n = 247)

Adj. R 2 = .4314 Adj. R 2 = .4974

 

 

 

Note: *, **, *** indicate values that are significantly different from zero at the .10, .05, and
.01 levels, respectively.

102

 

The effect of OTA signals on price has been found in the past when cable
programming was not as rich as today, but more recent studies have not
reported such effect. A positive correlation between TV market size and cable
service quality found here imply that cable operators prefer to compete with
broadcast stations on the quality or viewing choice dimension rather than on the
price dimension. This finding is consistent with Jayaratne (1996) who found that
cable operators met broadcast competition by offering more cable-only channels,
but not by lowering cable rates. This is interesting because cable operators were
found to employ price-cutting strategy when they face overbuild competition. But
in dealing with competition from broadcasting, cable operators pick up viewing
choice strategy rather than price cutting. It is no surprise, however, if one
considers the weak substitutability of broadcast signals these days. Closely
related to this finding is home density in a community. Unlike the null effect of
home density in the price equations discussed above, the number of homes per
mile (DENSITY) is a significant and positive predictor in both quality equations. It
suggests that cable consumers receive more channels, instead of discounted
price, where density may help operators achieve cost savings. It also implies that
cable systems in urban areas are more likely to have high channel capacity or to
upgrade their systems.

A similar analysis may apply to MSO affiliation. Looking at the M80 level,
the bigger MSOs (TOP8MSO) offer significantly better service relative to the
smaller MSOs. In the price equations, TOP8MSO was found to have a positive

effect on price, contrary to a priori expectations. This confirms Jayaratne's

103

finding that while MSO affiliation is related to higher rates, the larger MSOs also
offer better service quality. Like density, possible cost savings for TOP8MSO
may be transferred to the consumer as a form of higher quality rather than lower
price. In the quality estimation, cable operators affiliated with the largest 8 MSOs
are predicted to offer 6 to 9 percent more channels than smaller MSOs, holding
other factors constant.

The coefficients on OVERBUILD suggest that an operator facing overbuild
competition offers, ceten's pan'bus, 9 to 11 percent more channels than a
monopolist, depending on the model estimated. This would translate to 4 to 5
additional channels on a duopolist cable system since the typical monopolist
offers 41 cable-only channels or 52 basic channels in total. These estimates are
statistically significant at the .01 level. This finding is inconsistent with Emmons
and Prager (1997), but consistent with Jayaratne (1996). While Emmons and
Prager found no evidence that competition affected the quality of basic cable
service, Jayaratne reported that a cable operator in an overbuild area offered 26
to 34 percent more cable-only channels than a comparably situated monopolist.
The present study finds that the effect is somewhere inbetween these two
previous studies. Interestingly, potential competition (POTENTIAL) was also
found to influence service quality upward. The effect size is similar with actual
competition and significantly different from zero. On the other hand, municipal
cable systems maintain the same service quality, measured by the number of

channels, as private cable systems.

104

Finally, cable systems that sell high-speed lntemet access services
(INTERNET) over cable offer significantly more channels, ceteris paribus,
relative to systems that do not offer such services. ONESTOP is a significant
predictor for the total number of channels, which includes broadcast signals, but

an insignificant predictor for the number of cable-only channels.

105

Chapter 4

CONSOLIDATION AND CONCENTRATION OF THE MULTICHANNEL
TELEVISION INDUSTRY

The multichannel television industry is a fairly concentrated industry and is
also characterized by significant ownership ties among cable companies and
related firms. This chapter investigates the effects of concentration in the
national market and consolidation of regional markets on overbuild competition
in the multichannel television (MCTV) industry. There are three ownership issues
that characterize the cable industry: horizontal concentration and relationships
among cable operators; vertical relationships between system operators and
program suppliers; and clustered cable systems, whereby cable operators
consolidate ownership within geographical areas. More specifically, this chapter
examines the status of ownership concentration in markets for the delivery of
video programming. It also discusses how this can erect barriers to competition,
affecting the competitive prospects and viability of overbuilds in the multichannel

television industry.

Horizontal Concentration in the Multichannel Television Indust_ry
According to the 1999 Annual Report by FCC (1999 Report), a total of

80.9 million households subscribed to multichannel video programming services
as of June 1999; 81 percent penetration among US. television households (see
Table 6 ). The 1999 Report also reveals that cable television is still the dominant

technology for the delivery of video programming to consumers in the MCTV

106

marketplace, although its market share continues to decline since the mid 19909.
As of June 1999, about 82 percent of all MCTV subscribers received their video
programming from a local franchised cable operator, a 10 percent decrease from
1994. The cable market share declined even more than indicated in the study
because the subscriber totals of phone companies, which have overbuilt
incumbent cable operators, were included in the 82 percent.

While the number of cable subscribers continued to grow, reaching 66.7
million, the total number of noncable MCTV households grew to 14.2 million
homes. Much of the increase in the number of noncable MCTV subscribers is
attributable to the growth of DBS. DBS appears to attract former cable
subscribers and consumers not previously subscribing to an MCTV. As of June
1999, 10.1 million households subscribed to DBS, representing 12.5 percent of
all MCTV subscribers. There also have been a number of additional cable
overbuilds in the last 5 years. Table 6 also shows that OVS, a new concept first
introduced in the 1996 Telecommunications Act as one means of providing
multichannel television service for LECs, is practically dead. Over the last 5
years, the number of subscribers to and market shares of other distribution
technologies such as MMDS and SMATV remained as fringe services or
continued to decline. On the other hand, DBS continues its expansionary trend
of gaining new subscribers since its initial launch in 1994.

The very success of DBS poses some regulatory challenges as well. The
30 percent national ownership limits are no longer enforceable as DBS becomes

a robust substitute to cable. Since these rules are applied to the number of

107

homes passed by a single entity that provides multichannel television services,
the virtual ubiquity of DBS signals renders these 30 percent limits obsolete. lf
DBS is properly viewed as a close substitute to cable and included in the
measures of concentration, as in the recent FCC Annual Report, such national
ownership limits must be increased considerably to reflect the coverage of DBS.
If DBS does not enter in the relevant product market, the multichannel television
industry should be defined as “moderately concentrated,“ which may warrant
more stringent ownership restrictions.

Considering these ownership restrictions are not strongly enforced at the
present time, however, one policy option would be to repeal them altogether.
They also would be inconsistent to the principle of regulatory symmetry and
“level playing field” from the broader perspective of converging
telecommunications. It is too myopic to limit the number of homes passed for
video programming when the industry is quickly moving toward the provision of
complete packages of video, voice, and data. This may require a much greater
subscriber base than the traditional telephone or cable services.

Not only do these ownership limits implicate the First Amendment issues,
but such limits solve little problems as intended. An MSO even without reaching
30 percent of US. households can still ask significant concessions from the
program suppliers. Thus, the ownership limits do not provide small MSOs or new
entrants with the protection from stiff discounts, which would have obvious

implications for overbuild competition.

108

Should the 30 percent limits be increased or repealed, it might exacerbate
the problems associated with horizontal concentration, which will be discussed
later. To resolve these issues, it might be necessary to introduce an alternative
policy designed to safeguard “fair and nondiscriminatory“ terms and conditions
for multichannel television programs. The policy must strike the very conduct of
possible anticompetitive practices between distributors and suppliers, rather than
maintain structural restrictions that do practically nothing for sound

competitiveness in the multichannel industry.

109

o:

0 £282 680.. d mam.» 00". ”co-Sow

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

$666 $86 $86 $86 .98- >85. 6 6866
$86- 886 $8.8 $86 $2 $2 8.65 6866
86.8 86.8 686 8.6 28.58...» «6
$86. $86 $86 $666 $.66 $86 .96» >52 6 686..
$66.66 $86.. $66.6. $66.8 $8.86 $8.86 689.6 6866
866666. 68866 86686 86.86... 68.866 6868 285366 666
$666 $86 $666 $6.6 $86 $36 .966 >65. 6 686..
$666.- $6.6- $6...1 $ .66. $86 $. .8 88:6 6866
.9666... 86666.6 66466.6 866666 8668.6 8666.6 235366 am:
$66.. $66.. $8.. $8.. $8.. $8.. .98 >62 6 6866
$66.66 $36.- $.66 $8.6. $6.6. $36.. 885 686...
6866.... 686.6 8666... 8666... 8668 6868 22.535 >556
$66.. $8.. $6.... $8.. $46.. $86 .93 >5: 6 6866
$86.- $86. $9.6. $8.8 $8. .4 $8. .6 689.6 686..
68. .66 68.86.. 86.8... 86.8... 8668 86.86 235866 66:...
$8.8 $8.8 $6.66 $2.66 $66.8 $66.8 .96» >55. 6 6866
$66.. $8.. $66.. $666 $666 $666 88.6 6866
86.868 86.8...8 8666.66 8.86.8 68.8.66 68.86.66 22.5866 668
$66.6 $6666 $6666 $.63. $64.... $66.66 6266881666966
$.66 $86 $66.. $666 $6.6 $86 8.65 6866
.3688 86.866 6666866 6866666 8668.8 8668.8 86:83.. >5...
$8.. $8.. $86 $6... $666 $66.. 6688 6866
68.8666 68.8666 86.6866 86.8666 686866 86866 8.2.83: 3

tmm: >mo_occo :-

 

6c_mo_oc..oc._- 62:06:60 .0 EOE66066< o 956...

 

Consolidations within the cable industry continue as cable operators
acquire and trade systems. The seven largest operators now serve almost 90
percent of all US. cable subscribers. The FCC’s examination of upstream
national MCTV concentration currently reveals a relatively low level of
concentration. The top five firms in the upstream MCTV nationwide programming
market are AT&T (with a share of 20.5%), Time Warner (15.95%), DIRECTV
(9.23%), Comcast (8.26%), and MediaOne (5.84). The share of subscribers of all
MSOs in these top five MCTV operators (except DIRECTV) has changed little
over the years. The four largest cable MSOs (AT&T, Time Warner, MediaOne,
and Comcast) have served approximately 54 percent of all MCTV subscribers
since 1995. The only prominent change is DIRECTV, which accounted for 9.23
percent nationwide.

These shares relate to the broader MCTV market rather than specifically
to the cable market. To assess the potential for market power resulting from
concentration, the reported MCTV shares an be appropriately translated into
HHl figures. The FCC finds that in terms of HHI, national concentration among
the top MCTV operators has declined since 1997. In assessing national
concentration of the multichannel television industry, the FCC started in 1997 to
consider the presence of all MCTV operators and MCTV subscribers, not just
cable MS09 and cable subscribers (FCC, 1998). Given the close substitutability,
it was an appropriate adjustment to reflect the continuing growth of DBS
systems, such as DIRECTVIUSSB, Primestar, and Echostar. As of June 1999,
the MCTV industry with a Herfindahl-Hirschman Index (HHI) of 923 can be

111

defined as "unconcentrated" under the Department of Justice (DOJ) Merger
Guidelines (see Table 7). However, this is largely due to the FCC’s introduction
of DBS systems into the concentration measure in an attempt to reflect the
dramatic rise of DBS services and mergers of largest DBS systems in 1999. This
is no surprise because DBS systems such as DIRECTV and EchoStar now rank
among the ten largest MCTV operators in terms of nationwide subscribership
along with eight cable MSOs. The inclusion of DBS in the measures of industry
concentration seems to be appropriate, but it also poses some regulatory

‘ challenges, particularly with respect to horizontal ownership limits. This issue will
be deferred to the last section of this chapter.

DBS aside, after a stable period of market concentration during the early
half in the 19909, cable MSOs started to increase horizontal concentration within
the wireline cable industry. As a result, the cable industry has exhibited a strong
movement toward internal concentration since the mid-19905 when US.
Congress was expected to pass new legislation, soon or later, that would allow
LECs to provide multichannel television services. Separately considered, the
wireline cable industry with the HHI of greater than 1300 can be categorized as

“moderately concentrated“ nationwide under the DOJ guidelines.

112

662626ng 36266666. .6666 .4. 66864 6.6 .666. .66 666866 .666. .6. .6969. 6866.66. .6666 .666
.6 LE“. .5506. .mEEEEmoE 82> .0 29:60 65 .2 98:65. 5 62:56:50 60 639m 05 Co EoEmmcmm< .macc< .00“. .68.:ow

 

 

 

 

 

 

 

 

 

 

666. 666. 666. 666 666 666 666 666 6.65%....

666 666. 66.. 6 .6. 666. 666 666 666 666 666 9.55.. .1:
66.66 66.66 66.66 66.66 666 4.66 .66 6.46 6.66 6.6 66 666
66.46 66.66 66.46 66.66 6.66 4.66 .66 6.46 6.66 6.66 66 666
66.46 46. .6 66.66 64. .6 6.66 6.66 6.66 6.46 4. .6 6. .6 6. 666
46.66 66.46 66.46 66.66 6.46 6.64 6.64 6.64 6.64 664 4 666
6664 46.64 64.64 66.64 6.64 464 6.64 6.64 6.64 6.64 6 606
64.66 66.64 .6. .4 66.64 .64 6.66 6.66 6.66 .66 6.66 6 666

66.66 64.66 4666 66.46 6.66 6.46 6.46 6.66 6.46 6.46 22.6 no.

 

 

 

 

 

4..

 

 

 

 

 

mag-mam. ”65.686596 ooc_> 6o 6665656 65 .2 66:65. 8.62.62 9.. E cosmLEcocoo 6 6.62-

 

113

Horizontal concentration within the cable industry has been achieved
through numerous mergers and acquisitions (M&A). A maturing cable market,
consolidation and more technology investments helped spur continuous record
years of mergers and acquisitions among cable, DBS, and lntemet access
providers in the second half of 19909. Brief review of key M&As during this time
period is warranted as they appear to reshape the multichannel television
industry.

In June 1998, AT&T announced its agreement to merge with Tele-
Communications Inc. Through this merger valued at about $48 billion, AT&T
plans to integrate its telecommunications business with TCI’s cable networks and
thereby build a facilities-based local residential telecommunications network. The
merging companies contend that the merger will expand and accelerate AT&T's
ability to compete with incumbent local exchange carriers (ILECs) in providing local
telephone service to residential customers. After a significant investment to
upgrade TCI's cable facilities to allow two-way voice and data communications, the
merger will allow almost one-third of American households access to the upgraded
high-capacity broadband networks (FCC, 1999C).

AT&T’s intention was clear: the best way to bypass the Bells and to grab
local customers was to use TCl's cable network business. TCI needed help
expanding into telephone and other two-way communications services. TCI also
owns stakes in other MSOs, including Cablevision Systems Corp., and it is

involved in a number of joint ventures with other large MSOs. AT&T considers

114

those stakes to present at least theoretical access to another 17 million homes
and 10 million TCI cable subscribers.

VVItI‘I the purchase of TCI, AT&T acquired big part of “last miles” for cable
telephony, but it would be far short of the nationwide local access that AT&T
craves. AT&T still needed to find a way to reach the remaining two-thirds of the
nation's households not passed by TCI. In 1999, AT&T solidified its dominant
position in the telecommunications world, with a series of bold transactions
giving the company 3 million MediaOne Group subscribers, a coveted telephony
deal with Comcast Corp., and a deep-pocketed partner in Microsoft Corp. Ending
a costly bidding war between Comcast, AT&T signed a definitive merger
agreement with MediaOne to buy out the fourth largest cable MSO for about $62
billion in May 1999. With the multi-faceted deal, AT&T gained access to
Comcast's 8 million homes to deliver AT&T-branded telephony services (Neel,
19996). The addition of MediaOne vaults AT&T to the top of the cable heap and
furthers its strategy to provide bundled video, voice and data services to a vast
number of consumers. AT&T now would have access to more than 35 percent of
US. homes (Farrell, 1999C).

AT&T pays substantial amounts to get direct links to homes. Some
estimates for AT&T’s local access and interconnection fees run as high as fifteen
billion dollars in 1998, about a third of its domestic wireline revenues. TCI’s local
transport may enable AT&T to cut those charges in half and thus double its net
income. Little surprise that some estimates suggest that AT&T plans to have

extensive and exclusive cable/phone penetration in four to five years. In that

115

case, gains from video services, let alone lntemet access, are just gravy. Seen
that way, AT&T got the basic advantage of lntemet access for a small marginal
cost. Moreover, the modifications required to add lntemet capacity to an existing
digital cable system are much lower than the estimates of the costs required for
upgrade of the digital network itself (Bar et al., 1999).

The purchase of TCI and MediaOne suggests that AT&T firmly believes
that there is a business opportunity for them in the local loop, and it is the
ultimate endorsement of the cable network into the home. Probably there would
be no better partner than AT&T for cable systems to provide the local telephone
service. AT&T could help to drive cable telephony (Gibbons, 1998c).

The surprising AT&T/MediaOne marriage will likely prompt a series of
other wrenching changes in the cable industry's strategic alliances over the next
few years. Of particular interest: the dismantling or restructuring of the often-
prickly Time Warner Entertainment partnership between Time Warner and
MediaOne, a merger of the @Home Network and Road Runner high-speed data
services and a sweeping local-telephone senrice marketing deal between the
new AT&T/MediaOne, Time Warner and Comcast.

At the same time, changes in ownership and reassignments of orbital
slots have altered the DBS landscape. FCC consented to the merger of United
States Satellite Broadcasting Co., Inc. (“USSB') and DIRECTV, and authorized
the transfer of USSB’s direct broadcast satellite licenses to DIRECTV. Also in
April 1999, Hughes, the parent company of DIRECTV, completed its acquisition

of PrimeStar's medium powered satellite business. Subsequent to that grant,

116

Tempo Satellite, Inc, a former DBS licensee that did not actually commence
service, was also allowed to transfer its license to DIRECTV. These changes
give DIRECTV increased channel capacity, facilitate “local-into-local" broadcast
signal carriage and, for the first time, allow for DBS service to Hawaii, which has

not previously been served by DBS.

Consolidation of the Regional Markets

Consolidations of cable systems have always given MSOs greater
leverage in dealing with equipment suppliers and programming distributors and
greater access to investment capital. However, this wave of mergers between
MSOs was largely prompted by the need to build "cluster" systems that are
geographically proximate. Clustering is a process by which MSOs consolidate
system ownership within separate geographical regions.

The trend toward consolidation between MSOs started in the early 19909.
Since 1995, particularly, cable MSOs have accelerated undertaking system
mergers, acquisitions, divestitures, swaps, and joint ventures with the objective
of creating regional “clusters“ of contiguous cable systems (see Table 8). It was
a clear indication of cable MSOs' preparation for upcoming competition, which
has been continuously and seriously, discussed among the policy makers since
the late 19809. During the period of 1994 through 1998, there were more than
400 such cable transactions. Most of these transactions resulted in the
expansion of existing clusters of cable systems or the creation of new clusters

(FCC, 1998).

117

 

6-0 x6884. 666-66 62 6666 60 .666 .6 .6666 .965. :36

6.066884. 66.66 62 .9666 mo .68.. .6666 .6854 6...“. 65669696 82> 6 662.66 9.. 6. 65.6.: 9.. c. 666.6668
6 6296 6.6 6 6666868 .6266 6 6.6:. 9.. c. .666. 6 666.. 006 c. 69.6. 8-66 6 .666. 6.166 6 .666. 6.166 6 .666. .64
-66 6 .666. 66-66 6 666. .6898 66.66.“. >6 6666 9.6 666506669666 >6 6.666 663. .66. 696.8694 .669. .86 .8566

 

 

 

 

118

 

 

 

 

 

 

 

 

 

 

 

 

Yov wow min 6. : mdn an r N. «n hnw RON hm Ego-r
Q? .N m... o. 6.6 o. _..m w wN «4 oomA
Nd 6 mm m on w m.v 0. me n mmToov
60 ON mm 4. 5m 5 wN m ON 0 mam-com
no mm Nb mm mm vm 42w mm 00 mm mow-cow
of mm 60 av 00. mm «4.04 mu 06 mm mar-Do.
. . . . 4 . .

.6659“. 62.63838
6.6.63.0 886.6 6.660 6.65. .o 665 65.535 6:6 LonEzz 6 636.6

 

As a result of acquisitions and trades, cable MSOs have continued to
increase the extent to which their systems form regional clusters. Currently, more
than 40 million of the nation’s cable subscribers are served by regional clusters
(see Table 8). Most cable companies are following strategies to cluster their
operations geographically so that they may operate most of the cable systems in
a particular city or region.

The upward trend in the total number of clusters serving at least 100,000
subscribers observed in 1994, reached a peak in 1996, and began to decrease
in 1997. Although the total cumulative number of clusters declined from 139 at
the end of 1996 to 117 at the end of 1997, the trend for clusters to increase in
subscribership or size appears to be continuing. This tendency toward larger
clusters clearly reflects greater economies of scale. In an empirical study that
estimated valuation models of cable systems, transactions to form a cluster
commanded a higher price, suggesting a premium for the expansion of the local
service area (Miller, 1997). System value was most sensitive to changes in
subscribership, total miles, and subscribers per mile. Whether or not the buyer
was one of the top 50 MSOs was also a significant explanatory factor in Miller's
models. Possible reasons for this premium could be that ownership by one of the
larger MSOs might generate tremendous value from a distribution perspective
such as enhanced access to cheaper capital, technology, programming, and
elevated revenue streams. Furthermore, combined with the top 50 M80 status,
the consolidation of system assets to achieve possible operating economies

stemming from regional clusters is valued at a premium by system buyers.

119

Large MSOs’ consolidation of local systems within a geographic area is
changing the competitive landscape of the cable industry in particular, the
telecommunications industry in general. AT&T continues to pursue its clustering
strategy and has made a number of substantial transactions with other MSOs in
furtherance of this strategy. AT&T plans to concentrate about 90 percent of its
subscriber base in the top 25 markets in the country. AT&T has been shedding
systems east of the Mississippi River through sales, exchanges and formation of
partnerships.

In the Chicago metropolitan area, for example, before Ameritech’s entry
there were five cable operators with large subscriber bases. The area had been
served by TCI, Time Warner, MediaOne, Jones lntercable, and Prime Cable.
Since 1996, TCI has made a number of swaps and acquisitions through which it
has gained control of the systems previously owned by Time Warner, MediaOne,
Jones lntercable, and Multimedia. This clustering strategy allows TCI (now
AT&T) to control virtually the entire Chicago metropolitan market. AT&T is also
eyeing Chicago—area systems owned by Prime Cable and Cablevision Systems
Corp. (Multichannel News, April 17, 1998).

MediaOne acquired 30,000 cable subscribers for $60 million in the Detroit
area from Time Warner Cable in June 1998. The Time Warner operations
included in the sale are suburban systems in Dearbom and Wayne, Mich, which
serve 25,000 and 5,000 subscribers, respectively. These cities fit nicely with the
other areas that MediaOne serves, where the M80 is offering advanced

telecommunications services like high-speed lntemet access and digital

120

telephone. MediaOne will have 310,000 subscribers in Detroit, making it the
second-largest M80 in that metropolitan area and the third-largest in Michigan.
With systems that have already been rebuilt to 750 megahertz, Dearbom and
Wayne are candidates for deployment of MediaOne Express, a high-speed
lntemet-access service that currently passes 52,000 area households (Estrella,
1998b).

In 999, Comcast Corporation and Jones lntercable, Inc. announced
agreements to swap certain cable systems with Adelphia Communications. With
these cable system swaps, Comcast continues to strengthen its Mid-Atlantic
super-cluster, encompassing cable systems in New Jersey, eastern
Pennsylvania, Maryland, Delaware, Washington DC and Virginia. These system
swaps, together with other pending acquisitions and investments will make
Comcast's Mid-Atlantic supercluster one of the largest cable system clusters in
the nation with interests in more than four million owned and managed cable
subscribers. In addition, they will enhance Comcast's and Jones' cable clusters
in the Midwest, in Michigan and Indiana, and also in West Florida (Comcast,
1999b).

TCI started to consolidate the San Francisco market. With the acquisition
of Viacom Cable in 1996, AT&T now serves 90 percent (1.3 million households)
of the San Francisco Bay area's cable subscribers; 90 percent (900,000
households) of the Seattle-Tacoma, Wash., area; and 60 percent (300,000

households) Portland, Ore., cable customers. In line with that, AT&T's San

121

Francisco Bay area systems rolled out such new services as telephony and high-
speed data access (Neel, 1996).

Cable MSOs have claimed that they are merging to create market clusters
to continue growth, to be more efficient, and to build scale economies in cable
operations, such as advertising sales and new technology implementation.
These companies seem to regard large-scale clusters of cable systems as an
important competitive strategy for facing the increasingly aggressive DBS,
telephone, and wireless cable companies. It is widely believed in the cable
industry that the bigger and the more efficient an M80 is, the more competitive it
will be and the less likely it is to need the assistance of a telco to deliver
interactive broadband media to consumers (Chan-Olmsted, 1996).

M803 are trying to form a “cluster" around major cities to reduce costs, to
improve operating and management efficiencies, to eliminate system
redundancies, and to attract more advertising. By consolidating systems in major
markets, MSOs can serve entire regions comprised of numerous local franchise
areas. This assures advertisers that they will get extensive regional market
coverage. The growing importance of advertising revenues for cable systems
has emerged as a major factor promoting regional consolidation.

But more importantly, regional clustering also enhances MSOs ability to
compete successfully in the cable senrices markets under competition. Cable
MSOs could obtain greater economies of scale from clustering as compared to
having cable systems that were noncontiguous and more geographically spread

out. In particular, the clustering strategy enables firms to consolidate facilities for

122

receiving and transmitting programming, reduce the number of repair crews,
have regional customer service centers, reduce management, and compete
more effectively for local advertising dollars (General Accounting Office, 1999b).

Furthermore, with clustering, cable operators may be able to achieve
efficiencies that facilitate the provision of cable and other services, such as
telephony. Clustering provides the critical mass of subscribers necessary to
support the huge capital investment needed to make system upgrades designed
to enable companies to enter other lines of telecommunications services, such
as lntemet access and local phone service. Clustering thus positions cable as a
potential competitor for local exchange services. It enables cable operators to
offer a wide variety of broadband services at lower prices to customers in a
geographic area that is larger than a single cable franchise area. For this reason,
clustering makes cable operators a more effective competitor to LECs whose
service areas are usually larger than a single cable franchise area. Clustered
systems increase cable operators’ ability to be more competitive across a range
of services and technologies such as video programming delivery, IP telephony,
and lntemet access services as “full service providers” in their service area
(FCC, 1997).

The industry's maturation and diminished prospects for cash-flow growth
also helped spark property sales in recent years. Several media companies
abandoned the field, including Times Mirror 00., Viacom Inc., Providence
Journal Co., Houston Industries Inc. and E.W. Scripps Co. With TV household

penetration rates leveling off, incremental revenue growth through new

123

construction and new subscription has been virtually exhausted. Faced with the
prospects of a slower-growing revenue stream and margins, operators have
chosen to cluster their holdings by region and other factors in order to achieve
efficiencies through lower overhead and cross-selling (Neel, 1997).

Clustering not only increases operating efficiencies, but it may also enhance
cable MSOs' competitive position in the video programming acquisition and
creation. Industry observers believe that along with local news, local sports holds
the most value for operators in the market. As the value and popularity of
regional sports programming has exploded over the past decade, cable
operators have, for the most part, relied on regional-sports networks to provide

access to local professional teams (Umstead, 1998). Recent examples include:

o Cablevision purchased the New York Rangers NHL and New York Knicks
NBA teams to fuel its SportsChannel New York (now Fox Sports New York)
and Madison Square Garden Network regional networks.

. Comcast Corp. bought the Philadelphia 76ers NBA and Philadelphia Flyers
NHL teams in 1996 to anchor programming for its Comcast SportsNet
servrce.

. In 1997, Marcus Cable purchased the Texas Rangers MLB franchise in an
effort to develop a new regional-sports network in the market.

0 In 1998, Adelphia Communications chairman John Rigas entered into an
agreement to purchase a majority interest in the NHL's Buffalo Sabres, as
well as to secure the team's cable rights for the MSG-owned Empire Sports
Network.

A cluster in a major city with the ownership of these regional sports teams
may provide the operator wit a significant competitive advantage in establishing a
superior programming portfolio vis-a-vis the overbuild and competing

multichannel operators.

124

Vertical Integration in the Multichannel Television lndustg

Vertical integration occurs when a cable operator has an ownership
interest in a video programming supplier or vice versa. The total number of
programming networks has grown and cable operators continue to consolidate
and develop new ownership interests. However, the proportion of vertically
integrated channels continues to decline. In 1999, there were 283 satellite
delivered national programming networks, of which 104 networks, representing
37 percent, were vertically integrated with a cable MSO (FCC, 2000). One or
more of the top six cable MSOs holds ownership interests in each of the 104
vertically integrated services. AT&T, the nation’s largest MSO, holds ownership
interests in 50 programming networks, or 18 percent of all programming
networks, through its subsidiary, Liberty Media. Time Warner has an ownership
interest in 24, or 8 percent of all programming networks. Cablevision, through its
programming subsidiary, Rainbow Media, owns 11 programming networks, or 4
percent of all national networks (FCC, 1998). Cox Communications has interests
in 23, or 8 percent of national programming networks. MediaOne, with
ownership interests in 12 national programming networks, controls 4 percent of
national programming. Comcast, with its six networks and control over two
national networks through Jones lntercable, has ownership interests in 3 percent
of all national programming networks.

Vertical integration is not only associated with the largest cable system

operators, but also the programming networks with the largest number of

125

subscribers. In addition, currently eight out of the top 15 video programming
networks ranked by prime time ratings are vertically integrated with a cable MSO.

It appears that a significant amount of MCTV programming is controlled
by approximately twelve companies, including cable MSOs, broadcasters, and
other media entities. The 12 companies are: ABC/Disney, General Electric, CBS,
News Corp, Time Warner, Viacom, Discovery, Rainbow Media, Liberty Media,
USA Networks, E.W. Scripps, and Comcast. For example, of the top 50
programming networks in terms of subscribership, 46 are owned by one or more
of these 12 companies. C-SPAN, C-SPAN2, WGN, and The Weather Channel
are the four unaffiliated programming networks among the top 50 programming
networks.

Sports programming warrants special attention because of its widespread
appeal and strategic significance for multichannel television. Industry observers
believe that along with local news, local sports holds the most value for operators
in the market. As the value and popularity of regional sports programming has
exploded over the past decade, cable operators have, for the most part, relied on
regional-sports networks to provide access to local professional teams
(Umstead, 1998). The 1999 FCC Report identifies 75 regional networks, 26 of
which are sports channels, many owned at least in part by MSOs.

These vertical relationships may have beneficial effects, or may deter
competitive entry in the video marketplace, and/or limit the diversity of
programming (Ahn, 1997). One of the most important efficiencies of vertical

integration between cable operators and program suppliers comes from the

126

reduced risk of program development. Developing new and innovative
programming is costly and risky. Not only does the vertical relationship help to
reduce costs of negotiating and enforcing long-term contracts between cable
companies and program suppliers, but the relationship also helps to ensure an
outlet for newly developed programming. For program suppliers, this minimizes
the risk that programs will be developed but not be marketable. From the cable
operators’ perspective, ownership ties to program suppliers help guarantee the
continued availability of programming. To be sure, the cable companies’ vertical
ties to program suppliers are largely responsible for the development of the

varied programming that now exists. (General Accounting Office, 1999b).

Concentration, Consolidation, and Comgtition

The preceding discussion suggests that cable companies realize benefits
from their size, ownership interests, and clustering. Consolidation in the cable
industry is not unexpected, parallels consolidation in other telecommunications
industry segments, and is “a competitive response to challenges from DBS,
LECs, and others. Although there are efficiencies realized from the ownership
relationships in the cable industry, there is also the potential for adverse market
effects from these relationships. Besides the joint and fixed nature of investment
capital, and incumbents’ market predation discussed in Chapter 3, overbuild
entrants face competitive barriers stemming from horizontal concentration at
both the national and regional levels. These barriers exist in two dimensions of

inputs for cable programming: first, cable entrants have to pay much higher

127

prices for the programming networks, leading to significant cost advantages for
the incumbent MSOs; second, new entrants may be denied access to certain
types of networks that are either exclusively licensed for incumbents or owned
and operated by large incumbent MSOs. Moreover, a clustered system can not
only enjoy extra revenue potential from local advertising, but it may also take
advantage of their market power at the local level to secure exclusive carriage of

certain programming.

Post-Entry Cost Advantages for Incumbents

Benefits are also gained by the large size of some cable companies and
from horizontal relationships with other cable companies. Larger cable
companies may enjoy reduced programming costs and also have costs savings
in management and other overhead functions (General Accounting Office,
1999b). From the standpoint of programming costs, increased concentration can
enable the M80 to reduce the price it pays for programming by increasing its
subscriber base, its bargaining power, and the magnitude of its quantity
discounts. These discounts are the results of the established MSOs’ first-mover
advantage.

Contractual agreements between a system operator and a programmer
are based on several factors, including the value of the network to consumers
and local advertisers; the increasing costs associated with the programmer’s
obtaining the cable rights to programming; and the cost of production. The net

impact of the same license fees on cable operators will depend on each

128

operator’s contractual agreement with programmers. These agreements include
varying discounts and incentives that enable cable operators to reduce program
licensing fees (General Accounting Office, 1999a).

There are concerns about competitive disadvantages that horizontal
concentration may create for overbuild entrants. Because programming is a
major expense category for multichannel television service, “the differences
between what entrants pay for programming and what the major incumbent
MSOs pay have important consequences for both entrant viability and the vigor
of competition where it does occur” (Dertouzos & Wildman, 1998, p.4). It is
commonly known that the largest MSOs are able to exploit their control over
access to millions of cable subscribers to obtain programming on more favorable
terms than smaller MSOs. An MCTV must deliver in excess of several million
subscribers to a programming network to qualify for the maximum discounts
(Dertouzos & Wildman, 1998; Ford & Jackson, 1997). An analysis of 19 basic
cable networks by Dertouzos and Vlfildman (1998) reveals that small MCTVs pay
significantly higher license fees compared to the industry average. Of course,
many small cable operators have been able to lower their programming costs by
joining the National Cable Television Cooperative (NCTC), which aggregates its
members’ subscribers to collectively purchase programming. Nonetheless,
NCTC does not always receive discounts that are comparable to those given to
similarly sized MSOs (General Accounting Office, 1999a).

The Dertouzos and Wildman study (1998) found that the mean of “volume

discounts” for large MSOs was about 45 percent off the top rate card prices.

129

According to Paul Kagan Associates, for example, small multichannel television
operators that receive no discounts would pay 62 percent more than the industry
average rate for CNN, 60 percent for NICK and MTV, 50 percent for the
Discovery Channel, VH-1, Sci-F i, and the Learning Channel, and 20 percent for
ESPN and TNT. For a new entrant with 100,000 subscribers, this programming
cost disadvantage amounts to $2.7 million per year in total or $27 per subscriber
per year. This is approximately 10 percent of basic service revenues (Dertouzos
& Wildman, 1998). Similarly, Ford and Jackson (1997) found that the average
size MSO pays about 11 percent more than the largest MSO for programming,
while the smallest MSO pays 52 percent more. Clearly, the implied cost
advantages from concentration can be substantial.

Such discounts exist because large cable operators have more
negotiating power than small cable operators. Programmers have an incentive to
have their programming available to the greatest number of subscribers possible
because, as the number of viewers increases, so do programmers’ revenues
from licensing fees and advertising revenue. Therefore, cable operators with a
large subscriber base can obtain greater concessions, such as volume
discounts, from programmers than small cable operators can (General
Accounting Office, 1999a).

The license fee disadvantage is exacerbated when a clustered cable
system can generate extra revenue from local advertising—commonly known as
“local avails”. Of course, the primary beneficiaries of local cable advertising sales

are cable systems in large urban areas, which is where major MSO system

130

clusters operate. The net cost of multichannel television programming may be
higher for small cable operators because they take in less advertising revenue
and are less able to take advantage of some of the other incentives that offset
programming costs. Small and large cable companies generally receive a portion
of the advertising time that is available within cable programming offered on their
systems. For example, programming networks typically provide cable operators
with 2 minutes of local advertising per hour during their programming, which
represents a revenue source for cable operators. However, the market size of
small cable operators limits their ability to sell advertising during the time allotted
for local advertising. For example, a cable system needs at least 5,000
subscribers to break even with the costs incurred to develop an advertising sales
and production department (General Accounting Office, 1999a).

Dertouzos and Wildman (1998) found that if the contribution of local
advertising is taken into consideration, the annual cost disadvantage to an
overbuild entrant with 100,000 subscribers amounts to nearly $3.9 million, or $39
per subscriber per year. Thus, it is obvious that a cost disadvantage of this size
alone could be a serious impediment to the viability of overbuild competition.
Because these programming cost advantages are realized only by the largest
MSOs, “they also represent a margin competitive entrants cannot cut into without
incurring substantial losses. Alternatively, they constitute a profit cushion that
allows incumbents to cut prices to levels that are unprofitable for entrants without

incurring losses themselves” (Dertouzos & Wildman, 1998, pp.24-25).

131

F urtherrnore, net licensing fees for small cable operators could be even
higher because small cable operators have difficulty taking advantage of, or
negotiating for, some of the other incentives available to both small and large
companies, such as marketing co-op and cash or other incentives. Although the
magnitude of these incentives is unclear, it is difficult for small companies to
utilize marketing support money because they have limited markets and few
media outlets on which to spend promotion funding. It is common practice in the
multichannel television industry for programming networks to provide cash or
other incentives to system operators in return for airing promotional advertising in
the local market. Many small system operators are unable to take advantage of
these incentives because they do not have a sufficient number of subscribers to
justify the investment necessary to sell, produce, and insert promotional
advertising into cable programming (General Accounting Office, 1999a).

A preemptive strategy, which might be preferred by the incumbent
monopolist, would be to limit entry by some means other than price. A nonprice
signal to potential entrants, such as a threat of postentry reaction, might be
preferable because then price need not be held below the potential entrant’s
average cost. By themselves, threats have never seemed too credible an entry
deterrent because they require little financial commitment and the threatened
post-entry conflict would harm the initial monopolist as well as the new entrant.
Once the potential rival has entered, joint profits are not maximized through price

wars (Smiley & Ravid, 1983).

132

The existence of first-mover advantage, however, opens up a wide variety
of strategic possibilities for an initial monopolist and alters one aspect of this
strategy (Smiley & Ravid, 1983). The initial monopolist may be able to preempt
the market if it can limit entry through a particular pricing strategy. Furthermore,
the initial monopolist can earn economic profits while setting price to preempt the
market. Thanks to post-entry cost advantage, the initial monopolist can threaten
to reduce price below the new entrant’s average costs, while at the same time
earning profits itself. Thus, first-mover advantages confer the incumbent
monopolist a very strong strategic option: “double squeeze.” The monopolist
may be able to preempt the market while earning economic rents and threats of
postentry reaction, directed at potential entrants, will be more credible when
learning occurs.

When Ieaming, a process in which average costs decline with increases in
cumulative production, occurs in a market, an initial monopolist can preempt the
market through the use of a limit price and still make a profit (Smiley & Ravid,
1983). Indeed, as long as its cumulative output exceeds the new entrant’s, the
initial monopolist can simultaneously earn profits and inflict losses on a new
entrant forever. The initial monopolist does not need to convince the potential
entrant that the monopolist will act irrationally if entry occurs and destroy both
rivals in an Armageddon type war; learning will allow the monopolist to destroy
the entrant and earn profits at the same time.

In order for any strategy to preempt a market successfully, a potential

entrant must expect entry to result in no economic profits. One such strategy is

133

to set price (establish a price trajectory) so that no economic profits would be
earned in the first period of a new entrant’s experience. If the potential entrant
expects this zero profit experience to be repeated in each period, entry will not
be a profitable investment.

The monopolist only need to set a price between the potential entrant's
average costs and its own. This price will simultaneously result in economic
profits for the initial monopolist and deter entry, if the potential entrant expects
the negative profit experience of the first period of production to be repeated.
And this expectation is a reasonable though not certain one, since if entry did
indeed occur, the initial monopolist could continue to price below the new
entrant’s costs and still earn economic profits except in the unlikely event that the
new entrants cumulative output were to exceed the incumbents cumulative
output. The new entrant is unlikely to have a greater cumulative output, since its
attempts to gain market share through cutting price can be matched immediately
by the initial monopolist; and in such a destructive price war, the initial
monopolist will lose less than will the new entrant, due to the incumbent's first
mover advantage (Smiley & Ravid, 1983).

Further, such a threat is more credible if it can be carried out without
violating the antitrust laws. Traditionally espoused criteria for predatory behavior
(Areeda & Turner, 1975) would allow a postentry policy that set price between
the incumbents and new entrant's costs, and that subsequently drove the new
entrant from the market. Although the credibility problem with threats of

predatory reaction is perhaps not completely solved, the first mover advantage in

134

a market in which learning occurs substantially alleviates the incumbents
dilemma.

In sum, when production is accompanied by first mover advantages (e.g.,
learning by doing), the price charged by the initial monopolist will be lower than
the static monopoly price. If the rate of entry is a function of price, the monopolist
will further lower price to limit entry.

Therefore, an overbuild competitor will inevitably compete at a cost
disadvantage to the incumbent MSOs in the supply of cable programming unless
it enters multiple markets simultaneously and very successfully. As a result, new
entrants are significantly hampered in their efforts to compete viably in the video
marketplace. Price increases of programming input represent cost increases to
local cable systems and hence compress their profit margins. Should incumbent
MSOs simultaneously lower the prices of cable service in the local distribution
markets, this form of predatory pricing would now squeeze profit margins for the
small, new entrants from above and below at the same time.

This double squeeze is more intensified if the incumbent is one of the
larger MSOs that has ownership interests in programming networks. The
extensive vertical integration in the cable industry magnifies the anticompetitive
impact of such price discrimination. Not only do these entrants have a higher
cost of programming than large MSOs, but the increased sums flowing into the
vertically integrated programming networks serve to subsidize the large
incumbent MSOs’ local system operations. In short, there is a compounding

effect associated with excessive price discrimination based on subscriber base,

135

which creates a competitive disadvantage for new entrants. It is a formidable
barrier to the development of a competitive multichannel television marketplace
(Ameritech New Media, 1999).

If these cost savings are passed on to consumers in the form of lower
cable prices, then the effects of concentration could be welfare enhancing.
Alternatively, if these discounts preclude competitive entry, they may serve to
diminish consumer welfare despite a reduction in price.

It is ambiguous, however, whether or not large MSOs transfer their lower
input costs into consumer prices. While Chipty (1995) found that a 1 percent
increase in national size would result in a 0.01 percent decrease in price, the
effect size appears to be trivial. Ford and Jackson (1997) found that MSO size
was positively correlated to price, although statistically insignificant. The
empirical evidence in Chapter 3 suggests that large MSOs charge higher prices,
which supports prior studies by the FCC (1994) and Jayaratne (1996). While little
consumer welfare is realized, horizontal concentration can result in programming
discounts large enough to potentially constitute an absolute cost advantage for
incumbent operators vis-a-vis potential entrants, erecting a barrier to entry.
Overbuild competition has been shown by numerous prior studies to reduce
prices by over 20 percent and the econometric analysis in Chapter 3 confirms
these findings. Considering the ambiguity of welfare-enhancing effects of
concentration, limits on horizontal concentration may be justified, especially
when ownership concentration restrains competitive entry. Large MSOs may

have lower costs, either due to economies of scale or due to bargaining power

136

with suppliers. Both these effects of horizontal size on firm costs provide
motivation for horizontal integration, although the latter has often been ignored in
studies of firm behavior. Dertouzos and Wildman (1998) assert that the
programming cost advantage of the largest MSOs over their smaller competitors
is almost solely a reflection of MSOs’ bargaining power over programmers. They
argue that it is difficult to explain price differences of this magnitude with the
standard efficiency and incentive reasons for quantity discounts because the
actual cost of negotiating network supply contracts is quite small. Chipty (1995)
tested a “bargaining power hypothesis” that cable MSOs may be able to
horizontally integrate, or increase their size, in order to enhance their bargaining
power over input supplier. She found that large MSOs have lower programming
supply costs for reasons other than scale economies, suggesting a bargaining
power effect on costs. Thus, the anticompetitive potential of absolute cost

advantages may not be unwarranted.

Issues of Program Access

In addition to the programming cost disadvantages, loopholes in existing
statutes allow incumbent MSOs to negotiate exclusive contracts with attractive
cable programming networks that deny these programming sources to
competitors. Leveraging their large subscriber bases and extensive geographic
coverage, incumbent MSOs find it profitable to outbid entrants for exclusive

rights to these networks.

137

Further, large incumbent MSOs tend to have ownership interests in
popular new networks such as the regional sports networks. With a few
exceptions, Cablew’sion Systems Corp.’s Rainbow Media Holdings Inc. and
Fox/Liberty Networks either own or are affiliated with more than 20 regional-
sports networks that have programming deals with most professional teams: 25
of the 30 MLB teams, 26 of the 29 National Basketball Association teams and 19
of the 26 National Hockey League squads. Fox and ESPN have tied up the rights
to most major college conferences for ratings-rich football and basketball. That
leaves very little marquee product for operators looking to start their own sports
senrices. Cable operators do not live by local sports programming alone; they are
major players in competing for professional-sports rights. Comcast Corp. bought
the Philadelphia 76ers NBA and Philadelphia Flyers NHL teams in 1996 to
anchor programming for its Comcast SportsNet service. Cablevision Systems
Corp. owns the New York Rangers NHL and New York Knicks NBA teams and
uses them to fuel its own regional sports networks such as SportsChannel New
York (now Fox Sports New York) and Madison Square Garden Network
(Umstead, 1998):

There are increasing complaints that incumbent MSOs refuse to license
them to competitors. The increasing trend of clustering in the cable industry may
worsen this situation. Certain federal laws and FCC rules have been designed to
ensure that vertically integrated cable companies make their programming
available to other market participants and to limit the extent of horizontal

concentration in the industry. Despite these safeguards, there are concerns

138

about the potential harmful effects of ownership ties and concentration within the
cable industry.

The program access rules adopted pursuant to the 1992 Cable Act were
designed to ensure that alternative MCTV operators can acquire, on non-
discriminatory terms, vertically-integrated satellite delivered programming. The
terrestrial distribution of programming, including in particular regional sports
programming, could eventually have a substantial impact on the ability of
alternative MCTV operators to compete in the video marketplace.

The Cable Television Consumer Protection and Competition Act of 1992
includes provisions aimed at, among other things, enhancing competition in the
multichannel television market. The FCC’s program access rules were designed,
in part, to ensure that vertically integrated cable operators generally make their
satellite subscription programming available to competitors. The program access
niles were very important in helping new competitors—particularly, the DBS
firms—to get a foothold in the subscription television market. Despite the
success of the program access rules, most of the “noncable” providers and
competing cable companies (who are “overbuilding” a cable system where an
incumbent’s system exists) express concerns about access to reasonably priced
programming networks that are owned by cable MSOs. Such concerns include a
“loophole,” whereby programs owned by cable companies that are delivered to
cable systems’ facilities through means other than satellite, such as through fiber
wires, are not covered under the current program access rules. Although there

have only been a few complaints filed with the FCC on this issue, there is

139

concern that such delivery of programs may become more widespread,
particularly as the clustering of cable systems increases. The FCC has recently
written that this practice needs to be monitored, but the Commission noted that
the program access rules as written would need to be clarified to provide clear
authority to the FCC over programming delivered in this manner. Therefore, the
program access rules should be broadened to include cable-owned
programming delivered to cable systems through means other than satellite.
The 1992 Cable Act directed the FCC to place limits on the concentration
of ownership of cable systems nationwide. Under the limit that the FCC has set,
generally no person or entity can own or have an attributable interest in a cable
system that reaches more than 30 percent of the homes with access to cable
nationwide. The FCC has stated that with this limit, it is unlikely that a cable
company, or a combination of two cable companies acting together, could thwart
entry by a new subscription network. However, there are concerns that dominant
cable operators are winning price concessions and may have significant
bargaining power vis-a-vis subscription networks even when there is no
ownership link. A subscription network needs its product to be carried by at least
one of the two largest cable companies in order to be economically viable—thus
creating a dependence on the larger cable companies and giving them
significant influence over the subscription network. In fact, program suppliers that
are not vertically integrated (such as MTV, A&E Network, and the Weather
Channel) may be very dependent on large cable companies. These independent

programming suppliers should generally be required to be made available to all

140

competitors, as is currently the case for programming owned by vertically
integrated suppliers.

Clustering also raises certain anticompetitive concerns. Clustering is
accomplished through the elimination of potential competition as operators of
adjacent cable systems would be the most likely overbuilder in the market.
These operators would be relatively low-cost potential wireline overbuilders—
because they could likely use their existing headend and parts of their existing
trunk lines to serve the new markets—compared to overbuilding a distant
wireline system. The potential cost saving is significant because the headend
and trunk lines comprise about 25 percent of the capital investment of a cable
system. Overbuilding from adjoining franchise areas, however, has rarely been a
significant means of entry into cable service markets. In recent instances where
overbuilding has occurred or is planned, the overbuilders have not been the
operators of existing adjacent cable systems. While these companies generally
do not compete against each other in local markets—that is, consumers rarely
have a choice of cable operators—this level of consolidation has more
significance for the program acquisition market (FCC, 1997).

BellSouth, Ameritech, and RCN report that they have experienced
difficulties obtaining programming (Ameritech New Media, 1999; BellSouth
Corp., 1999; RCN Corp., 1999). These new entrants point to the actual or
potential problem of migration of programs from satellite to terrestrial delivery in

order to avoid the program access rules. Ameritech and BellSouth also report

141

difficulty in gaining access to non-vertically integrated networks, and to cable
networks owned by the over-the-air broadcast companies.

The FCC has established rules prohibiting unfair and discriminatory
practices by vertically integrated cable operators. The program access rules seek
to promote competition and diversity in the multichannel video programming
market. Some industry observers express concerns related to the migration of
channels from satellite delivery to terrestrial delivery and price discrimination.
They assert that terrestrial migration weakens the intent of program access rules
and threatens the development of a competitive MCTV market. In 1998, the
Commission found that the record did not support claims of anticompetitive
impact from programming being moved from satellite to terrestrial delivery.
However, the FCC noted that terrestrial migration may in the future impact the
ability of alternative MCTV operators to compete in the video marketplace (FCC,
1998).

Several commenters advocate extending the program access rules to
non-vertically integrated programmers (Ameritech New Media, 1999). Alternative
MCTV operators indicate that certain programming networks not owned by cable
operators (e.g. fX, MSNBC) are not available to them. They contend that
exclusive agreements between unaffiliated programmers and incumbent cable
operators is a significant barrier to competition. Similarly, DBS operators assert
that cable operators are able to use their “overwhelming” buying power in the
programming market with both vertically integrated programmers, as well as

independent programmers at the expense of noncable MCTV (DIRECTV, 1999).

142

BellSouth asserts that consolidation and clustering can have a “chilling effect on
the willingness of programmers to sell their product to cable’s competitors.” The
incumbent cable operator generally maintains, however, that consolidation and
clustering result in cost savings and efficiencies that promote competition.

The National Cable Television Association (NCTA) disputes that an
anticompetitive effect results from the unavailability of terrestrially delivered or
non-vertically integrated networks and asserts that product differentiation and
exclusivity are legitimate, pro-competitive strategies that increases choices and
the amount of programming available to consumers. Non-vertically integrated
programmers respond that they actively seek widespread distribution and
audience share and thus have no incentive to disadvantage competing
distribution outlets. Independent programmers, like Viacom, argue that
exclusivity agreements allow them to gain distribution, recognition, and the
subscriber base necessary to successfully launch a start-up network and
therefore, actually add to program diversity. The FCC, in its 1998 Program
Access Order, declined to extend the program accesses rules to non-vertically
integrated programming networks, finding that the record did not support such
action.

Clustered systems may find it feasible and profitable to make the
terrestrial delivery of regional video programming services. Cable companies
under competitive pressure would be tempted to prevent competitors from
gaining access to vertically integrated regional programming. Regional

consolidation or clustering increasingly renders fiber-based networks more

143

attractive for the incumbent to deliver local cable programming to substantial
numbers of subscribers in a number of major markets including New York City,
Chicago, Columbus, Philadelphia, Boston, Minneapolis, Orlando, and Kansas
City. These fiber networks —- as opposed to satellite distribution — are now
being used by regional sports networks.

In New York and Philadelphia, where Cablevision Systems Corp. and
Comcast had increased its marketing power through system swaps and
clustering arrangements, the incumbents changed previously satellite delivered
sports programming to terrestrial distribution. Due to loopholes in program
access laws, such migration permits the incumbent MSOs to deny its competitor
access to MSG Sports Network, Fox Sports Network, New York, and Comcast
SportsNet. Similarly, the Tribune company recently migrated nearly 50 Chicago
Cubs games from the satellite-delivered WGN to the fiber-based ChicagoLand
Television Network (CoreComm Ltd., 1999).

Indeed, in Chicago, where AT&T amassed control over more than 90
percent of the entire market through regional consolidation, terrestrial distribution
of cable programming may become more economically feasible (Ameritech New
Media, 1999). Whatever the motivation of such a migration might be—be it
efficiency reasons or an attempt to evade program access obligations—the issue
of terrestrial distribution of programming could have a substantial impact on the
ability of alternative MCTVs to compete in the video marketplace (FCC, 1998), if
the dominance in a regional market is used to extract exclusive contracts for the

incumbent MSO. Indeed, Dertouzos and Vlfildman (1998) found that incumbent

144

MSOs are 25 percent more likely to carry regional sports networks than are
competitive overbuild systems. The trend toward regional consolidation and
terrestrial distribution will continue to increase as overbuild competition occurs in
major cities.

In summary, horizontal concentration and the accompanying monopsony
power of large MSOs will have a detrimental effect on viability of overbuild
competition. Needless to say, however, the financial viability of overbuild entry
also depends on other factors such as changes in demand, technological
breakthroughs, and potential for market expansion. This issue will be discussed

in the next chapter.

145

Chapter 5

CONVERGENCE AND COMPETITION IN THE TELECOMMUNICATIONS
INDUSTRY

As telephone networks provide video and data communication senrices,
cable modems carry high-speed data and voice switched messages, and
computer networks transfer voice and video data, the boundaries of the
traditionally separated industries of telephone, cable, and computer are blurring.
The interaction between technology and conditions of entry creates a new
dynamic of intensified competition and an increased flow of new products. These
factors fuel each other and intensify the process, creating an ever-accelerating
spiral of new competition and barrier disintegration (Litman & Sochay, 1993).

Until recently, RBOCs were not too concerned about cable operators
getting into their core voice business. The situation is quickly changing, largely
due to the technological advancement and strategic consolidation in the cable
industry. Through the acquisition of the two gigantic cable assets—TCI and
MediaOne—AT&T Corp. has made cable a credible threat to local telephone
companies’ core business. After the completion of merging with TCI, and now
the proposed merger with MediaOne, AT&T brings the biggest threat in history to
telephone companies’ local residential dominance. Combining MediaOne
properties with the former TCI systems and its affiliates, the Time Warner
properties covered by a joint venture agreement, and the agreement with

Comcast, AT&T will be able to reach more than 60 percent of US. households.

146

Reaching the remainder could be achieved through wireless access, purchasing
unbundled network elements or pure wholesale.

But the most serious threat from AT&T is not its sheer size. Instead, it is
Ma Cable’s ability to bundle a wide array of telecommunications services that no
other company can match. With AT&T’s entry into cable, the industry’s attention
to telephone services and bundled packages is expected to grow dramatically
over the next several years. Over time, the rules of engagement will be redefined
from one medium to a multimedia industry. The telecommunication industry is
rapidly moving in that direction, where one-stop shops offer all sorts of facilities-
based voice, video and lntemet products, and customers may receive savings
and convenience. This chapter begins with the economic implications and effects
of convergence on competition. The chapter then presents the current state of
integration of voice, data, and video services, followed by a brief comparison of
competing broadband services in technological and economic aspects. The
chapter concludes with the discussion of how large MSOs can take strategic
advantage of the multiproduct operations in dealing with increasing threats of

competition in the market.

Structural Effects of Convergence on Comgtition

Many analysts provide different views of convergence in
telecommunications. They appear to agree that convergence may take several
forms. Convergence is perceived to emerge as a result of technological

integration, business motivation, and regulatory changes (Baldwin, 1997; Katz,

147

1996; OECD, 1992). Each type of convergence has broad implications on
economic welfare. The most obvious way in which convergence affects
economic welfare is by expanding the set of potential service offerings.
Convergence of technologies, as well as business and regulatory models, allows
the provision of old services in new lower-cost, high-quality and more convenient
ways.

Clearly technological change has played a key role in driving
convergence. While technological change is important, other important forces of
change stem from a shift in the policy makers' attitude toward competition in
telecommunications markets. The old view held that, as a consequence of
natural monopoly cost conditions, competition was antithetical to the public
interest. Telecommunications providers were granted monopoly franchises and
were thought to be in need of protection from competition in order to ensure their
ability to invest in new facilities and their willingness to serve all segments of the
public. In addition to stifling competition, this policy stifled convergence. The new
view of telecommunications policy holds that competition can generate
significant benefits. Since the most likely entrants into many telecommunications
markets may be incumbents in other such markets, the new policy of promoting

competition may also promote convergence (Katz, 1996).
Conceptual Problems of Convergence

In evaluating economic implications of convergence, it is necessary to

refine the forms of convergence. To capture the proper effects of convergence

148

on competition, one must not dwell on the loose definition of convergence since
it will lead to not only a misunderstanding of the phenomenon but also ill-advised
prediction of market structure and industry evolution. For example, Michael Katz

(1996), a business professor at Berkeley, states:

“Under the new model hailed by many telecommunications executives
and policy makers, voice, video and data will converge. But converge
in what sense? Will it be convergence of the transmission networks,
the service providers or the customer premises equipment, or some
combination of these?...[F]rom the end user‘s perspective, this sort of
multimedia convergence could also be achieved through the use of a
single, integrating piece of customer premises equipment.
Convergence from the end user's perspective may also have little to
do with technology, but be much more concerned with billing and
customer support, as when a single provider offers one-stop shopping
even if that firm does not actually produce some or all of the services
being bundled. Today, for example, AT&T will sell you both long-
distance telephone services and direct broadcast television services.
While the consumer sees a convergence of service provider, the two
services are transmitted over completely separate networks and the
pieces of customer premises equipment used to receive the services
are entirely distinct." (p. 1081)

Apparently, this view obscures the appropriate assessment of
convergence from an economic perspective. First of all, convergence from the
end users’ perspective, while seemingly useful, does not provide any significant
help in understanding how the telecommunication industry will evolve as a result
of converging technologies, services, and corporate organizations. For example,
“convergence of customer premises equipment,” as provided by Katz as a form
of convergence, does not carry any meaning as to how the integration of voice,
video, and data services will influence cost, price, and supply of such services.

While it is a very important element in the development of telecommunications,

149

the integration of CPEs has little to do with the evolution of telecommunication
senrice industry and competition. Integrated CPEs may become available as a
result, not a cause, of convergence. Integrated CPEs may help enhance
consumer convenience and thus ignite a new equipment industry, but they would
cast few economic implications, if any, for the development of the
telecommunications industry in the future because the cost structure of senrice
suppliers would not be affected by the integration of CPEs.

Furthermore, the view from end users misleads one to the confusion of
convergence with bundling. “When a single provider offers one-stop shopping
even if that firm does not actually produce some or all of the services being
bundled,” in Katz’s view, it would be a form of convergence. But in fact, it is
merely a form of bundling strategies or marketing cooperation, not convergence.
It has no direct relation to the cost stmcture of the firms that provide the service.
Nor does it increase or decrease the supply of the senrice. If one uses the term
“convergence” in this sense, therefore, the term is abused and its economic
implications become obscured.

Bundling can take place with any products or services, be they related or
not. In other words, it is possible for a firm to bundle without changing cost or
supply functions of each component in a bundle. Convergence, on the other
hand, necessarily implies changes in cost and/or supply functions of products in
question. Thus, when RBOCs (or AT&T in Katz’s example) sell satellite dishes in
a coordinated effort with DIRECTV, it is not a form of convergence, but of

marketing cooperation that attempts to take advantage of bundling strategy. The

150

critical point is that unlike convergence, the RBOCs sales of satellite dishes do
not change cost of DIRECTV’s operations in any way, nor do they influence the
output levels of DIRECTV’s multichannel television services. The outcome of this
marketing arrangement would have little impact on the supply of multichannel
television service, and hence competition, because DIRECTV is available
regardless of RBOCs engagement.

Viewed from this perspective, convergence can best be defined as the
realization of subadditivity in production due to joint production. Thus, the
economic implications of convergence become closely related to the economies
of scope, giving rise to cost and output levels of production. Formally,
subadditivity is said to exist if the costs of separate production are less than the
costs of separate production for any scale of output or combinations of outputs
(Bailey & F riedlaender, 1982). Thus, whenever the costs of providing the
services of the sharable input to two or more product lines are subadditive, the
multiproduct cost function exhibits economies of scope (Panzar & Willig, 1981).

Typically, economies of scope arise from the sharing or joint utilization of
inputs. There may be special cost savings that accrue to a multi-product firm that
are unavailable to firms producing only one of the products but not both.
Typically, such economies are associated with the presence of fixed factors of
production, sharing of intangible assets such as research activities or other forms
of business know-how, networking, or otherwise reuse of an input by more than
one product (Bailey 8 Friedlaender, 1982). In short, the single firm can supply a

bundle of products demanded by the market at a lower total cost than some

151

combination of two or more single-product or low volume producers. These
byproduct savings may further entrench economies of scale and enhance the
overall barrier to entry (Litman, 1998). Viewed from the economies of scope,
convergence casts significant implications for the evolution of the

telecommunications industry.

Convergence and Market Structure

One major impact of convergence is through its effects on market
structure and competition. There are a number of ways in which convergence
can be expected to change the structure of the telecommunications industry and
the nature of competition in the markets. From a public policy perspective, one of
the most exciting developments is the possibility of increased competition as
incumbents in one market enter into other markets. The theory underlying this
viewpoint is that technological change has created economies of scope that cut
across formerly separate markets and thus create cross-market entry incentives,
while regulatory changes such as the Telecommunications Act of 1996 allow
firms to respond to these incentives.

Insights into the economic implications of convergence are provided by

Baldwin (1997), who writes:

“Because the advanced integrated communications company has such
a powerful, essential and comprehensive set of services, it is unlikely
that any government would permit a private monopoly provider.
Therefore convergence implies competition.

An integrated service provider would compete for the core business of
an incumbent non-integrated service, such as a plain old telephone

152

service, or cable-based video service. A single private company with

an integrated voice, video and data service can offer all of the

traditional services and some important new ones based on the

integration. If for no other reason, it is necessary for each traditional

provider of a non-integrated service to have a much broader business
plan as a matter of self-defense. A traditional single service provider,
from now on, has to fear another type of company in the same service

area, becoming an integrated service provider.” (p.10)

The view of convergence as a source of increased local service
competition posits that where once there were two monopolists, one in voice and
one in video, now there will be two multiproduct oligopolists—in addition to any
new wireless-based operators such as direct satellite broadcast services and
MMDS. To be sure, convergence holds the promise of bringing increased
competition to many telecommunications markets by allowing existing service
providers in what has traditionally been separate markets to cross over into one
another’s industries. Potential change is particularly great in markets for the local
distribution of voice and multichannel video. As a result of technological and
regulatory developments, which removes many legal entry barriers, cable MSOs
and LECs have been expected to begin entering each other's markets (Katz,
1996).

Since the technology will permit a single comprehensive system,
integrating voice, video and data service, there will be a business or economic
convergence. It is possible for a telephone company to offer video and data
services, or a cable company to offer voice and high-speed data, or a public

utility to offer all of these communication services. And, there could be some

brand new entities emerging to provide these services if existing businesses do

153

not accept the challenge. Since it is technically possible for one company to
offer all of these services, there is an incentive to do so (Baldwin, 1997).

Indeed, this expectation was grounded on the much debated issues of
LECs’ intention and potential in the provision of voice, data, and video services
over the same transmission facility. Until the repeal of cable-telephone
crossownership bans, LECs had continuously contended that the opportunity to
provide video programming should increase LECs’ incentive to construct
integrated broadband network facilities to carry that programming to the home.
This, in turn, could lead to the introduction of a wide range of innovative voice,
data, and video services for subscribers (NTIA, 1991). Telephone companies
argued that it would be more efficient to deliver voice, data, and video services
on an integrated basis and that the potential economies of scope might be
substantial. They asserted, for example, that because a broadband system has
me bandwidth capacity to carry both high-speed data and video services, and
because telephony, high-speed data, and video all share the same transport
medium in an integrated switched broadband network, an economy of scope
would exist for each of these service capabilities. Telephone companies further
asserted that economies of scope were likely to arise from the sharing of capital
costs and expenses, marketing costs, R&D, and billing and accounting functions
among the three types of services carried (NTIA, 1991).

Convergence is believed to have two economic consequences with
respect to market structure. It may give rise to a significant shift in economic

perspectives of telecommunication services. The integration of voice, data, and

154

video services will strike down the most fundamental barriers to entry in the
traditional telecommunications industries. For the established telecommunication
senrice providers such as telephone and cable, convergence may help them
realize cost savings from joint production to the extent that the integration of
such senrices is technologically feasible and economically profitable. If
substantial economies of scope exist in the integration of voice, data, and video
services, it is highly likely that those telecommunication companies invade each
other's market sooner or later, resulting in the change of market structure from
monopoly to competitive markets. For new business entities that consider
entering this industry from scratch, convergence may play a role in two opposite
directions. First, the enlarged market opportunities will most likely enhance
revenue potential for new entrants. Therefore, integrated services can give these
companies extra revenue sources that have not been exploited by any
established telecommunication service providers. For example, high-speed
lntemet access service alone can significantly enhance a possibility for a brand-
new entrant to recoup its costs and make profits. Certainly, the new entrant need
not have to recoup the costs entirely from its video services. The new,
unexplored data services market may confer the new entrant some safety
cushion that would not have been available for firms specializing in video
services market. Thus, convergence will induce more entry in the multichannel
television industry.

On the other hand, one should not overstate the likely extent of

competitive entry, especially into local distribution markets. The changes in

155

technology and public policy notwithstanding, significant barriers to entry still
remain. In addition, there are economic forces associated with convergence that
may reduce competition rather than strengthen it. Increased economies of scope
and the increased use of multi-product bundling strategies may lead to the
erection of new forms of barriers as well as greater concentration of the
multichannel television industry. The converging telecommunications industry will
require an increase in capital investment for the new entrant necessary to
provide such new senrices. This, in turn, may increase absolute cost barriers to
entry, which will lead to unfavorable conditions for small-scale entry.

But to what extent will convergence create pressures for industry
consolidation and increase the difficulty of entry in the long run? To answer this
question, one must examine economies of scale and scope, as well as the
magnitude of sunk costs. Economies of scale and scope can affect industry
structure in several ways (Katz, 1996). Viewed in static terms, a wireline local
telephone network and a wireline multichannel video distribution network have
been considered to have strong characteristics of natural monopolies. And there
clearly are natural monopoly elements present in a combined voice and video
network. An important question for industry structure is whether the production
technology used by a combined network will have greater economies of scale
that increase minimum efficient scale relative to market demand. The effects on
entry should also be taken into account. If all providers offer integrated packages

of voice, video and data services, a potential entrant might also have to come

156

into the industry as a multiservice provider. This is likely to raise the sunk costs
of entry, thus increasing its risk.

Such barriers may be reinforced with post-entry cost disadvantages for
the new entrant discussed earlier in chapter IV. Moreover, entry—particularly
facilities-based entry—will take time, even when it is by firms already established
in related markets. Further, competitive entry will not occur evenly in terms of the
targeted user groups. High-volume customers located in urban areas will see

competition much sooner than will low-volume, rural customers.

Current Development of Service Convergence

The next generation lntemet marketplace will be driven by the deployment
of ubiquitous, "always-on” networking with broadband content into the home.
Home networks permanently connected to the lntemet, with screens in several
rooms, are a possible part of this vision. Interactive video conferencing and low
cost lntemet telephony are also parts. But what really distinguishes this phase is
the final convergence of TV and PC, of entertainment, education, and work at
home, the seamless linking of the home into the larger electronic community.
Broadband means many different kinds of content and communication patterns
concurrently; ”always-on" makes the home a permanent part of the network. The
next generation communication applications and patterns of lntemet use will not
necessarily be restricted to the home and will be adapted throughout the
economy. But the mass market will play a key role in shaping the third

generation lntemet and e-commerce evolution because it will bring a population

157

of broadband users large enough to constitute a critical mass able to sustain the
development of third generation applications (Bar et al., 1999). This section
presents a summary of three new services that will significantly alter the
landscape of telecommunications: cable modems, IP telephony, and DSL

services.

High-speed lntemet Sen/ice Over Cable

Historically, cable networks were constructed to provide traditional video
programming services that required only one-way transmission of signals. Until
recently, the typical one-way cable system provided approximately 50 channels
of analog video. The network was a full coaxial system designed with a
centralized headend and lines called “trunks” leading from the headend to nodes
placed in the residential neighborhoods. Distribution lines emanated from these
nodes which carried the signals through the residential neighborhood. A coaxial
wire called a “drop” line then carried the service from the distribution line to the
customer's television set. The distribution and drop lines represent the cable
industry's "last mile” of plant into the consumer's home. A traditional 350 MHz
coaxial cable systems included many amplifiers to boost the signal along the way
to subscribers’ homes.

Although telephone and cable incumbents already have facilities serving
the last mile, traditional telephone and cable plants are not ideally suited for
broadband. Broadband, however, opens the possibility of new facilities to serve

the last mile to the home. As lntemet usage continues its dramatic rise, the

158

demand for broadband services grows. The market demand to bring high-speed
data, video and voice to residential and business customers is reflected in
increased levels of investment and faster deployment schedules for various
technologies.

According to the National Cable Television Association (NCTA), the cable
industry's spending on the deployment of two-way broadband via high-speed
cable modems in 1997 alone totaled $6 billion. One estimate is that 63 percent
of all cable systems will be broadband-ready by 2001. TCI has committed to
spend $1.8 billion to upgrade its plant, in part to provide broadband services.
TCI's upgrades are expected to be 60 percent completed by the end of 1999 and
90 percent completed by the end of 2000. Another major cable operator,
Comcast, has spent over $1.2 billion in the past three years to upgrade its cable
systems, largely to be able to offer broadband. In addition, Microsoft's
investment of $1 billion in Comcast, the investments of $210 million each by
Microsoft and Compaq in Road Runner, and AT&T’s purchase of TCI for $48
billion all appear to be motivated in part by a desire to enter broadband via cable
television systems (FCC, 1999a).

Major MSOs are eager to get their network upgraded because they view
lntemet access as a high-revenue, high-margin product. Cable companies will
have to diversify their product and find other revenue sources to survive in the
increasingly competitive environment. They invest so much to upgrade their
systems because they want to develop new, revenue-generating products, such

as interactive television, digital cable, and high-speed lntemet devices.

159

Cable MSOs deployment of high-speed lntemet service was prompted
largely by an increased level of competition in the multichannel television market.
As RBOCs and others enter the video industry, cable MSOs feel lntemet access
services would allow them to generate another revenue stream in the short term
and to maintain a competitive edge in the long term. In many competitive
markets, cable operators responded to telephone entry by increasing channel

capacity and introducing new services.

Access to the In t er cable generally hasbecome easier in the past

, r- _ .4 w)..." ' "' "Mc—afl~m-"-

thre’gyga'rs. To deliver data services over the cable network, cable operators
using a two-way broadband architecture typically allocate one television channel
for downstream traffic and one channel for upstream traffic. Cable operators
using a one-way broadband network typically allocate one television channel for
downstream, while the upstream path is provided over a telephone line. At the
cable headend, a cable modem termination system (CMTS) communicates
through the allotted channels with cable modems located in subscriber homes to
create a virtual local area network (LAN) connection (Kinetic Strategies, 1999).

In addition to the advantage of speed, lntemet over cable also offers end
users a connection that is “always on,” as compared with the more widely-used
dial-up services. Furthermore, most lntemet over cable providers offer
proprietary content. These lSPs are also known as online service providers or
OSPs .

As of this writing, virtually all of the major cable operators offer broadband

access in some areas, and they are steadily expanding service areas to meet

160

demand. Currently, however, service is not available in all markets. Notably,
cable broadband access will continue to become more widely available as cable
system infrastructures are upgraded. In markets where cable broadband is
available, the industry is hopeful that the eventual standardization of cable
modems will increase subscription levels. To that end, an industry consortium
called CableLabs has adopted hardware and software interface standards called
Data Over Cable Service Interface Specification System (DOCSIS) to support
the delivery of data services over the cable infrastructure. This standard should
contribute to lower-cost modems, less complex and time consuming installation
procedures, and potentially, self-installation by subscribers.

Cable operators have adopted an aggressive schedule to upgrade their
networks to provide broadband services. The cable operators are also partnering
with a number of cable lSPs that provide comprehensive networking and
systems integration services to support broadband access. For example,
Excite@Home and RoadRunner offer their own high-speed data backbone and
regional data centers with local caching equipment. Other companies, such as
High Speed Access Corporation (HSA) and ISP Channel, are offering basic
turnkey lntemet packages specifically designed for small cable system
operators.

The most recent development in cable modem services is summarized in
an FCC staff report (Lathen, 1999). As of August 1, 1999, cable modem service
was available to 32 million homes. This was equal to 30 percent of all cable

homes passed in the US. and Canada. The penetration rates for cable modem

161

service averages 3.5 percent as of 1999. The FCC report also found that the
total subscriber count could surpass 1.5 million by the end of 1999.

Many analysts believe that cable modem deployment will take off in the
coming years. The projections for residential cable modem subscribers range
from 4 to 6 million by 2002, and over 11 million by 2005. Particulaiiy, the
development of the DOCSIS standards for high-speed data delivery over cable
will dramatically accelerate cable modem deployment, because DOCSIS
compliant modems will increase the success of retail distribution channels, as
well as simplify the installation process.

With lntemet usage exploding, the number of cable modem users in North
America has passed the 1 million mark and all top 5 MSOs in the US have now
signed up more than 100,000 data customers apiece (see Table 9). Led by Time
Warner Inc. and MediaOne Group, cable operators signed up about 250,000
lntemet users in the first quarter of 1999 only. Both Time Warner and MediaOne
now have more than 150,000 data subscribers, while Cox, AT&T and Comcast
are estimated to have more than 100,000 high-speed lntemet subscribers.

The two leading providers of lntemet over cable are Excite@Home and
RoadRunner. Both are OSPs, offering proprietary content as well as access to
the lntemet. As of August 1, 1999, the Excite@Home subscriber count was
estimated at 670,000. RoadRunner’s subscriber count was estimated to be
350,000. A number of new cable ISPs, such as HSA, Prolog, and ISP Channel,
have partnered with cable operators to offer their versions of high speed lntemet

access. In a number of markets, both Road Runner and Excite@Home are

162

becoming the second biggest lntemet service providers, trailing only America

Online Inc. (Breznick, 1999a).

Table 9 CABLE MODEM CUSTOMER RANKINGS
(As of September 30, 1999 )

Time Warner 245 000
MediaOne 173 000
Shaw 147 000
Cox 140 474

125 400

AT&T 1 13 600
Comcast 11 900

41 500
Videotron 36 000
Cablevision 31 474
Other 240 000
Total 1 ,403,333

 

Source: Kinetic Strategies, Company Reports

Cable Telephony

In addition to high-speed data, large MSOs have begun to offer local
telephone service. Voice services from cable operators have grown robustly
since 1996 due to MSOs deployment of fiber-optic cable, aggressive marketing
strategies and a desire by operators to become "one-stop shops" for multiple
communications services. A recent report published by market research firm
Frost & Sullivan predicts that voice services will be a major cash cow for cable

operators over the next five years (Cho, 1999c).

163

This positive outlook stems from what they expect to occur as a result of
the AT&T’s acquisitions of TCI and MediaOne. It was local access and bundling
of voice and video services that AT&T believed as having the greatest potential
when it decided to acquire TCI and MediaOne (Vittore, 1999). AT&T already is
giving regional Bell operating companies their most fierce competition for the
residential voice market.

MediaOne has been rolling out telephony since late 1997. The MSO
offers telephony in six markets: Atlanta; Los Angeles; Pompano and
Jacksonville, Fla.; Richmond, Va.; and suburban Detroit. The MSO began
offering its digital-telephone product in the metropolitan Detroit communities of
Plymouth, Northville and Canton, Michigan with 55,000 homes passed (Hogan,
1999). The towns mark the first where MediaOne offers all three of its advanced
broadband services: high-speed lntemet access, digital video and telephony. In
Atlanta, MediaOne is offering consumers in neighborhoods that it has rewired
with hybrid fiber-coaxial 750-megahertz lines a handful of incentives and
features, including free phone installation and the ability to keep their existing
phone numbers (Paikert, 1998).

One of the more aggressive MSOs on the telephony front, Cox provides
digital telephone service to more than 40,000 subscribers in six markets with a
5.9 percent penetration rate (Farrell, 1999a). In 1997, Cox Communications Inc.
launched its ‘Digital Telephone” service in Orange County, California, followed

by Meriden, Connecticut, and Omaha, Nebraska (Applebaum, 1997). In 1998,

164

Cox introduced “Digital Telephone” in two new markets—San Diego and
Phoenix.

In general, competitive local exchange carriers (CLECs) such as
MediaOne and Cox undercut the incumbent local exchange carrier (ILEC).
Interestingly, pricing on the first line for Cox Digital Telephone services is
consistently about 20 percent lower—the same price reduction in the overbuild
areas for the multichannel service—than the incumbent RBOC, with the second
line about 50 percent lower. The discounts apply to Cox cable customers. In San
Diego and Orange County, that translates to $9.99 per month for the first phone
line, versus $11.25 from incumbent Pacific Bell. A second line costs $4.99 per
month (Hogan, 1998s).

TCI has introduced its own telephone service in Hartford, CT, Fremont,
CA, and Arlington Heights, IL. TCI and Ameritech New Media are embroiled in a
heated race to sign and keep cable subscribers in Arlington Heights, a suburb
northwest of Chicago. The two companies also have competed for telephone
customers in Arlington Heights since TCI began a soft launch of its local phone
senrice and high-speed lntemet service there in 1997. Earlier in 1997, Arlington
Heights became only the third test market for a TCI local phone service called
PeopIeLink and for its high speed lntemet-access senrice, @Home Network.
TCI’s (now AT&T) phone service uses hybrid fiber-coaxial cable, and the
operator offers high-speed lntemet access on the same plant (Hogan, 1997).

Residential telephony via lntemet protocol (IP) is still in its infancy, but

technology is at hand that can give cable service providers a clear advantage

165

over narrowband-service providers as lP-voice applications become
commonplace. With IP now embedded in backbones, end-to-end IP connectivity
provides the opportunity to process voice calling without having to deal with
gateways and traditional switches (Dawson, 1999b).

The idea of offering competitive residential telephone services has
captivated cable operators for years. Although reliable telephony equipment for
hybrid fiber-coaxial (HFC) networks is commercially available, significant
economic and operational barriers have discouraged most MSOs from widely
deploying it.

Now, with lntemet Protocol (IP) networks emerging as viable platforms for
the delivery of voice traffic, MSOs hope to use their high-speed data networks to
support packet telephone services instead of deploying standalone HFC
telephony equipment. The rationale is clear: Deploying separate telephony and
high-speed data architectures creates capital, operational and spectrum
inefficiencies, since a cable operator must purchase two hardware platforms,
allocate a pair of upstream and downstream channels for each service, integrate
two operations support systems, and assign staff to manage each service
offering (Kinetic Strategies, 1999b).

There are tremendous implications of IP-telephony trials over cable. Using
IP, cable operators hope to create an integrated multi-service communications
platform that operates on a lower cost structure than existing circuit-switched
alternatives, enabling aggressive service price discounting without sacrificing

margins. Besides undercutting competitors, MSOs hope the flexibility of IP

166

networks will allow them to deliver a host of unique value-added features, such
as integrated voice mail and e-mail messaging and the real-time provisioning of
additional phone lines without rewiring a home.

The ultimate goal of many MSOs is to also offer long-distance IP
telephony over their packet backbone networks. For example, a residential cable
IP telephony customer served by Comcast in Philadelphia might call another
cable IP telephony customer served by MediaOne in Los Angeles. The packet
calls could be carried nationwide at very low cost without ever touching a
telephone company network. MSOs are currently evaluating options to enter into
backbone interconnection arrangements that would make such a solution viable.
Apparently, IP telephony will hit the long distance market first, bringing a
challenge for traditional interexchange carriers including AT&T. But it will also
undermine the potential growth opportunities that LECs have been pursuing for a

long time.

Local Exchange Carriers’ DSL Deployment

Telephone companies are starting to realize the growth prospects of the
data business. In recent years, digital and data services revenue increased 30-
40 percent a year. In addition, a large portion of telephone companies revenue
comes from data. In fact, Ameritech data revenues overtook voice revenues in
1998 and the LEC expects exponential growth over the next few years as
customers are expecting carriers to provide end-to—end data solutions and total

converged solutions. Ameritech’ plans to transform its revenue structure is so

167

dramatic that in 2010 the LEC hopes to make 99 percent of its revenues from
data, leaving only 1 percent from voice (Quinton, 1999b). But most of the data
services being offered right now by the RBOCs were in the business sector,
rather than on the residential side, where most cable modems are deployed
(Cho, 1999d).

Being that data services are becoming such a large portion of the Bells'
revenue mix, it is natural for them to get even more aggressive in their
deployment of ADSL lines. But Bell companies cannot survive in the long run for
several fundamental reasons. Plain old telephone services (POTS) alone can no
longer buttress the Bells’ business. They must create new markets for revenues
including data and video. The telephone companies cannot maintain their
monopoly power in voice any more. Nor would they capture a significant market
share in data and video. This is partly due to technological advance in
telecommunications, but more importantly due to LECs’ inability to provide
integrated services in a bundle.

Recognizing the significance of the residential broadband market, the
telephone companies are aggressively entering the consumer market for high-
speed lntemet access service by deploying ADSL in an effort to counter the
threat of cable modems. About three years after launching cable modem service
in selected regions across the US, cable operators no longer have the high-
speed data market all to themselves. In city after city, regional phone companies,
competitive local exchange carriers (CLECs), wireless cable firms and other

telecommunications companies are introducing their own high-speed lntemet-

168

access services to compete against cable modems. While largely charging
higher prices than cable operators, they hope to lure residential and business
customers with promises of dedicated bandwidth, greater service reliability and a
wider choice of lntemet senrice providers (ISPs).

Technology affording such increased bandwidth for the conventional
twisted pair telephone lines is known as “digital subscriber line” and takes many
forms (collectively xDSL). To date, the most prominent forms are ADSL
(asymmetric DSL) and HDSL (high-speed DSL). DSL technology is capable of
increasing the capabilities of the incumbent LECs’ existing copper plant, and
may be capable of offering many advanced services. DSL uses digital signal
processing techniques to make possible the provision, on existing copper loops,
of high-speed data communications without interfering with the carriage of voice
service. DSL allows a copper loop to be used simultaneously for high-speed data
service and ordinary voice service, and keeps the data capability available 24
hours a day.

Although DSL service is available in several versions, there are two
general categories, symmetrical and asymmetrical. Symmetrical versions offer
the same data rates upstream and downstream and are best suited for business
applications such as video-conferencing. Asymmetrical versions offer different
data rates upstream and downstream and are ideal for residential users who
receive a lot of data but do not originate or send much (e.g. lntemet surfers).

One such version is called asymmetric digital subscriber line (ADSL). As ADSL

169

does not interfere with the basic voice service, the user can simultaneously
browse the lntemet or watch a movie while talking on the telephone.

ADSL typically delivers up to 8 megabits per second (Mbps) downstream
and 1 Mbps upstream (See Table 10). A "skinnied down” G.Lite version delivers
1.5 and .5 speeds. VDSL modems can run at symmetrical 52 Mbps speeds
depending on how close telco’s central office gets to the customer. It is all
dedicated point-to-point bandwidth per user. So it does not pose congestion

issues as a shared modem solution may have (Barthold, 1999b).

Table 10 Classes of DSL Technologies

 

 

 

 

 

 

Maximum Data Rate Max. Distance from
Acronym Full Name Central Office to
Downstream Upstream End-User (feet)
HDSL High-data-rate 1.5 Mbps 1.5 Mbps 12,000
DSL
SDSL Symmetric DSL 768 kbps 768 kbps 10,000
VDSL Very-high-data- 51.8 Mbps 2.3 Mbps 4,000
rate DSL
RADSL Rate—adaptive 8 Mbps 1 Mbps 18,000
DSL
ADSL Asymmetric DSL 1.5-8 Mbps 640 kbps 18,000
G.Lite DSL Lite 1.5 Mbps 384 kbps 22-25.000

 

 

 

 

 

 

 

Despite the promise of DSL to deliver broadband access to businesses
and consumers, there are several technical issues with regard to the widespread
implementation of DSL. One of the primary inhibitors is signal attenuation, also
known as the distance limitation. Attenuation describes the dissipation of signal
strength as it travels over the copper line. DSL utilizes a higher frequency that is
more susceptible to attenuation than ordinary voice transmission. Consequently,

the various DSL technologies detailed in Table 10 have distance limitations

170

ranging from 4,000 to 18,000 feet from the telephone company’s central office.
“These limitations may ease as technologies improve, but as a practical matter,
DSL is currently limited to locations within a three-mile maximum loop from the
central office” (Lathen, 1999).

The adoption of DSL-lite (or G.lite) should further accelerate the pace of
broadband deployment in the residential market, because it increases the
coverage area of DSL beyond 18,000 feet from the central office and allows
more homes to receive high speed lntemet access over their existing copper
lines. In addition, G.lite can be installed (plug-and-play) by the customer. This
plug-and-play feature is an important competitive factor, because it allows for off-
the-shelf retail availability and modem pre-installation in PCs—two goals the
cable industry hopes to achieve with its DOCSIS cable modem standard. The
plug-and-play feature also lowers labor costs, since it reduces the time
technicians spend installing the service at the home or business (Ellis & Dawson,
1998).

The ILECs’ aggressive deployment of DSL can be attributed in large part
to the deployment of cable modem service. Although the ILECs have possessed
DSL technology since the late 19808, they did not offer the service, for concern
that it would negatively impact their other lines of businesses. The deployment of
cable modem service, however, spurred the ILECs to offer DSL or risk losing
potential subscribers to cable. Prior to cable modem deployment, the ILECs had
little incentive to deploy DSL and the consumer had no choice for high-speed

lntemet access. In fact, the deployment of DSL could have an adverse impact on

171

the telephone companies’ T1 business. T1 is a form of high-speed access that
was sold primarily to business customers. With a price range of $300 to $3000
per month, the T1 business generated high profit margins for the telephone
companies. Since the price point of DSL was lower, ranging from $50 to $1000
per month (depending on the type of DSL), the deployment of DSL service would
undercut the T1 business (Lathen, 1999). A contrasting factor between carriers
that were moving aggressively into the consumer market and those that were
holding back was the extent to which they were willing to risk cannibalizing
existing service markets, such as the market for ISDN (Dawson, 1998c). Telcos
will not abandon ISDN and T1 services, but as they launch high-speed lntemet
service more and more, there will be more price pressure on telco's own existing
services like ISDN and T1 services.

Various telecommunications providers, from incumbent local exchange
carriers (lLECs) to competitive local exchange carriers (CLECs), recently have
adopted aggressive deployment schedules for digital subscriber lines (DSL). The
pace of telco DSL rollouts has escalated on schedule since 1998, including the
first commercial deployment by BellSouth Corp. in September 1998 and a major
expansion throughout California by Pacific Bell. Other telcos like USWC, GTE
Corp. and Ameritech Corp. have all moved fonrvard with large-scale ADSL
deployments in 1998. There were approximately 160,000 DSL lines in service at
the end of the second quarter 1999. This represents a 300 percent increase
since the fourth quarter 1998 and a 100 percent increase since the first quarter

1999. Analysts predicted that over 30 million telephone lines would be qualified

172

to support DSL services by the end of 1999 (See Table 11). Interestingly, one of
the first Bell companies to launch ADSL, Ameritech Corp., is proving to be one of

the most cautious in expanding its market base.

TABLE 11 Number of DSL-Ready Lines

RBOCs 1998 1999
Ameritech nla nla
Bell Atlantic 2.0 7.0
Bell South 2.0 4.0

SBC 3.3 10.0
US West 3.6 5.2
GTE 5.0 6.0
Total: 15.9 32.2

 

Source: Donaldson Lufl<in & Jenrette—Wireline Communications (June

1999) cited in Lathen (1999), Broadband Today, A staff report to William

E. Kennard, Chairman of FCC.

lLECs also have entered co—marketing and co-branding agreements with
established lntemet access companies such as AOL, MindSpring and EarthLink.
For example, Bell Atlantic and SBC have agreed to provide volume-discounted
DSL transport service to AOL in order to tap its 19 million-customer base and
brand name. BellSouth recently reached a similar agreement with MindSpring,
which has a 1.2 million customer base. EarthLink has agreements with GTE and
Sprint to offer DSL senrices nationwide (Lathen, 1999).

The ability to cost-effectively offer advanced services over VDSL rests in
part on the fact that telcos have ample amounts of fiber already in place within
the VDSL transport range. Moreover, the decreasing costs of ADSL equipment

used for the delivery of high-speed—data services over telco lines—now in the

173

$300 range for customer modems—is driving down the costs of the ATM and
modulation chips that will be used in VDSL set-top boxes (Dawson, 1999a). As a
result, what once seemed to be an ever-receding possibility for the use of
telephone wires in delivering high-speed data has suddenly become an imminent
reality involving not just the telcos, but an expanding base of competitive local-
exchange carriers and lntemet-service providers. The activity on the DSL front
posing the competitive pressure for cable can be found in market launches that

are under way around the country.

Comparison of Broadband Services: Technolggy and Economics
The market for high-speed lntemet service is starting to explode. There is

an emerging battle between cable modems and ADSL for getting high-speed
access to the home and business. The battle between cable and telephone
companies for lntemet surfers is heating up, as LECs started to aggressively roll
out ADSL service at much lower prices.

In assessing the diffusion of high-speed lntemet access services, an
objective way, if not the most convincing, is to compare the relative merits of
cable and DSL technologies. First, RBOCs’ high-speed lntemet services never
match cable modems in terms of price. High-speed data service through ADSL
requires the use of an ADSL modem, which is sold at $200. The service charges
are around $40 per month, not including charges for lntemet access. Early 1999,
Bell Atlantic has formed an agreement with lntemet-access giant America Online

Inc., making the telco’s asymmetrical-digitaI-subscriber-line service available to

174

AOL subscribers for an estimated $42 per month. SBC Communications Inc.
announced that it plans to roll out ADSL in markets in seven states—including
California and Connecticut—for $49 per month with lntemet access. The two
announcements are significant, as they address what many have said is one of
ADSL's biggest problems—that it is just too expensive. But with these offerings,
the two telcos are narrowing the gap between ADSL and cable-modem-service
pricing, which is generally sold at $40 per month. SBC’s $49-per-month price is
only valid for customers who commit to service for one year, and who use SBC's
ISP. For customers who opt for a month-to-month service or who use an outside
ISP, the price climbs to $81 (Farrell, 1999b). But despite the fine print, the new
pricing strategy signals the evolution of ADSL from mainly a business service—
priced anywhere between $60 and $200 per month—into one where the main
target is the residential user. And that change was accelerated by the success of
cable modems.

In technological aspects, ADSL data transfer rates are far below those of
cable modems. Perhaps the best measure of competitiveness for a various high-
speed lntemet service would be monthly “bits per buck” ratio, as in the last
column of Table 12. The "bits per buck" ratio, created by the FCC, is a short-
handed way to illustrate today's performance/price trade-off that a typical end-
user would face. Specifically, "bits per buck" here refers to the kilobits per
second of data a user would receive for each dollar spent, based on recurring
monthly costs (not including one time costs for, 9.9., CPE purchases and for

installation). The higher the kilobits per second for each dollar spent, the better

175

the value for the end-user. Clearly, broadband deployments in the lead today,
such as cable modems, benefit from having lower price points and higher data
speeds. ADSL and ADSL-lite, the telephone company’s version of high-speed

access, offer lower ratio because of lower data speeds and higher price points.

176

9.66.3 .6 .6628”. .6...
.66 .oz .380 00. 6-8 toaom 00“. .86. .o .o< 6co_.mo_c:EEooo_o.. o... .o won cosoow 0. 2.6365.“ E95230 scam o.m.o.ooo< o. 30% 03.669. van .5566“.
20E... oca 0.965660”. 6 5 6:60.523 :4. o. 3:52.60 63:60.5...5826... oooca>o< .0 2.553300 0.... 65:30:00 End... .o .26.: o... c. .83.. 00“. 605cm

 

 

 

 

 

 

 

 

 

.8982
.66 80.
So .6 6658.98 66.980 2:6“.
(.2 (.2 (.2 (.2 .86» (.2 $2 .695 6.5.2 5.9 o..._o.om - 2523
6.6... >63.~ c.
06.6 ”69336 no.2:
.00. 6886 c3533 can 60...... 60.56600 i 6620...;
.6. 666.6 s 88.8.. 8 6.6.66» 866 8.6 .8. 866 6.66 9.66 36.2 . 692.2 66626.
63305 ~10 66.34. -
.6523 6:0 66-66 60.2.. -
.3. 663cm .03 coon .» moot—6E 6.68 62:66:50 6629.2.
0.06 cos .6 5 60.5.9... on on» 08.» com» .88» on .25 c. 6.55. 6.. mos... 6.5695...
6686
2383:... .22 Ba 82...:
.8. 3686 .8. =32. 6... 3.33 .866 .830
0.6.. mam» c. 33.9... 0» own 5 .9... on on 5.2» 085.83 3.68 oEom 62.5. n .6092 6663.2 ozoo
3:6 .8.
$2 $2 <2 $2 866 8 866.83 6.65 c. an: . 8x. .86: £369..
6626
c3533 .83 new 602...:
.06. 66806 .63 69336 .636 .6030
6.3 866 5 3.5.0.... 0» 006.06» coma 8.» 83.88 6.68 oEom mans. 6.. aux. «Ow... ..mo<
L
.3. 66306 contowfivw Mower-MW $250
0.0. 063 c. 33.9... on 06.6.06» con» om» 83.83 02260662 669.. o9~ 05.23 i 05.28
66.6.0 60.....5
6.. 666.6 86.66» 666.666 866 8.6 - 666 <2 6.62 .8: 39. 66. com... 206.
85>
850 .666. 895 6992
v. 83 on» own com» on (.2 623.5662 668. on 605 .68.... 6:0...th

 

 

 

 

 

 

 

 

 

 

 

 

moooqm
$600 50> o>w 225 3.2.6263.
2.6m ....o:m 66...... .23. 8.36m 25:05. moo . c2556... Emo...w:3oo

.2. 6.5.. . 5:56.80 5585 25:05. . . 55380 853.50 . .oEzmcoo 323;. b...n6__m>< 3.9.8.2 9626:.
25.55. o. .30 5:56:00 2 .600 o. .30 o. .30 o. .30 o. .600 2250 .663: 33.6w 322.com;

 

wmeDwZOU ...<.._.zmn=mm¢ Oh wthO w hzm2>04awo >0040210mh Nv 03.6».

177

Most cable modern systems rely on a shared access platform, much like
an office LAN. Because cable modem subscribers share available bandwidth
during their sessions, there are concerns that cable modem users will see poor
performance as the number of subscribers increases on the network. Common
sense dictates that 200 cable data subscribers sharing a 27-Mbps connection
would each get only about 135 Kbps of throughput—virtually the same speed as
a 128-Kbps ISDN connection. Based on this characteristic, in fact, some analysts
contend that DSL will catch up cable modems in terms of speed (Barthold,
1999b; FCC, 1999a; Lathen, 1999). An FCC staff report (Lathen, 1999), for
example, argues that transmission speeds over cable networks degrade as more
subscribers are online because it is subjected to individual subscriber's action.
They argue, therefore, that as traffic mounts, data speeds may fall to the level of
ADSL or lower.

This contention implies that lntemet access services over cable networks
will have a negative network extemality problem (i.e., congestion) as the number
of subscribers increases. If so, cable operators will have a terrible dilemma
because the senrice is doomed to grow only to a certain level that can
accommodate a limited number of users. The arguments are seriously flawed,
however. This scenario may become true only if cable operators will take no
action upon the increase in the number of users. Unlike circuit-switched
telephone networks where a caller is allocated a dedicated connection, cable
modem users do not occupy a fixed amount of bandwidth during their online

session. Instead, they share the network with other active users and use the

178

network's resources only when they actually send or receive data in quick bursts.
So instead of 200 cable online users each being allocated 135 kbps, they are
able to grab all the bandwidth available during the millisecond they need to
download their data packets—up to many megabits per second.

If congestion does begin to occur due to high usage, cable operators have
the flexibility to allocate more bandwidth for data services. A cable operator can
simply allocate an additional 6 MHz video channel for high-speed data, doubling
the downstream bandwidth available to users. If the demand keeps increasing,
the system may allocate more channels, even half its channel capacity, for data
transmission. It would, of course, depend on the relative strengths of demand
levels for video, voice and data service. Another option for adding bandwidth is
to subdivide the physical cable network by running fiber-optic lines deeper into
neighborhoods. This reduces the number of homes served by each network
segment, or node. The segmentation of nodes increases the amount of
bandwidth available to end users within each node (Kinetic Strategies, 1999). In
fact, AT&T Broadband & lntemet Services even plans to test a "micronode"
architecture that pushes fiber to possibly every 50 homes passed, instead of
every 500 or 1,000 (Menezes, 1999). Eventually, cable operators can adopt fiber
to the home (FT'I'H) where every home is fibered from the central office
(headend) to the home drop as the demand for data services increases. The
beauty of this evolutionary scheme is, of course, its flexibility.

On the other hand, telcos have many issues to deal with as they push

ahead with DSL rollouts, even before they get to the point of agreeing on a

179

standardized approach. What makes DSL challenging is its technological
problems. There are distance issues [from home to central office] and issues of
how to run ADSL over digital loop—carrier systems, not to mention the amount of
equipment that has to be installed.

Because of these limitations in delivering DSL to every customer, LECs
have to fall back on ISDN (Integrated Services Digital Network). Although ISDN
could be a solution where DSL is not available, it may hurt LECs own business
(Barthold, 1999). There is limited availability of "last mile" competitive broadband
network infrastructure. In many areas, Cable is still the only broadband option for
the mass market. The access alternatives for business are considerably better
than for households. Indeed, the Local Exchange Carriers, incumbents and
competitors alike, have aimed DSL deployment at business customers. Larger
businesses in major commercial centers may also have fiber optic connections
from a CLEC. Alternatively, wireless broadband access services are emerging in
most major urban centers. However, most businesses in the United States do
not have access to the cable infrastructure, and running cable to a business
results in customer charges of thousands of dollars.

For the residential market, the cable modem and DSL are the only
broadband options today and it is clear that wireless broadband lntemet is not
now and is not likely to be available in the very near future. As a result, the
network alternatives for any given household are limited. The phone lines serving
a large share of US. households are simply unsuitable and will remain

unsuitable for DSL services. Reading of the evidence varies, but at least 20

180

percent, perhaps 40 percent of the local loops in the country will not presently
support DSL at anything near Cable-modem speeds, if at all (Bar et al., 1999).
Ameritech, for example, estimates that DSL will not work on 45 percent of its
loops today, and may never work on 20 percent of them. In the same vein, Bell
Atlantic's new DSL service can reach up to 80 percent of telephone subscribers
in the Bell Atlantic region. AT&T estimates that “most ILEC loops can be
upgraded to support ADSL. Vlfith digital loop carrier lines included, an estimated
60-80 percent of RBOC access lines are ADSL qualified” (FCC, 1999a).

Certainly, the share of local loops unsuitable for DSL will fall significantly
over time, but these limitations are unlikely to be overcome easily or soon. DSL
is therefore unavailable to a significant portion of the American territory today, or
anytime soon. Consequently, the benefits of broadband DSL are unlikely to be
available to many Americans in the near future.

This explains why LECs are working on a different architecture. BellSouth
Corp. is preparing to experiment with a fiber-to-the-home delivery system in
Atlanta. BellSouth will use the asynchronous transfer mode data-networking
protocol, enabling it to offer significantly higher data speeds and potentially wider
availability than digital-subscriber-line technology. Running fiber to the home
advances BellSouth’s plan to spend $200 million to replace existing copper lines
with fiber to the curb for 200,000 homes in metro Atlanta and metro Miami's
Dade and Broward counties as a platform for advanced data, video and voice
services. The Baby Bell also installs fiber as its preferred infrastructure in new

subdivisions. The concept of pushing fiber even deeper into the network has

181

been a hot topic both for phone companies and their cable-system rivals, as they
look for cost-effective ways to deliver more bandwidth for supporting advanced
consumer services (Menezes, 1999). But high costs and a general absence of
incentive to push the technology envelope in this direction have stymied the
deployment of fiber-based broadband-telco networks (Dawson, 1999a).

BellSouth believes that fiber to the home (FTI'H) is its ultimate platform
for satisfying customers' voracious appetite for bandwidth—an appetite that is
growing at exponential rates. BellSouth's FTTH setup includes optical-access
network equipment developed by Lucent Technologies, such as a network-
terrnination unit installed inside the home for converting optical signals from the
fiber to high-speed Ethernet data, which can be used by subscribers’ personal
computers.

Nor are telco prices where they need to be, even when the service offered
is only 256 kbps. At $40 per month plus $19.95 for lntemet access, the 256-kbps
version of the multispeed Megabit service is about $15 per month higher than it
should be to compete with cable modems (see Table 12). Telcos also have to
worry about what they can offer beyond speed in efforts to capture a large
consumer base. Here, they are well behind the cable industry, where much effort
has gone into developing consumer-friendly interfaces and multimedia-enriched
content that capitalizes on the power of high-speed access (Ellis & Dawson,
1998).

182

The Strategic Uses of Convergence

Nevertheless, it is not cable’s relative technical strengths against DSL that
will determine the prospects of competition in the broadband service industry.
The reality is that companies with the best products will not always win, as
chance events may cause ‘Iock-in' on inferior technologies. Quite independently
of any inherent technological advantage that cable or DSL may possess, the
rapid cable modem roll-out means that cable will have a massive, certainly
difficult to dislodge, and perhaps enduring deployment lead at least in the
residential market. In network markets such as this, initial leads often result in
enduring dominance that can even preempt later competition (Bar et al., 1999).
The gap between cable modem and DSL will likely endure because of strategic
uses of converging services. This section discusses why cable MSOs will
dominate the converging telecommunications market from the strategic
perspectives. Cable MSOs will have three main strategic advantages: bundling

capability, first-mover advantages, and dynamic resource allocation.

Bundling of Services
No analyses would be complete if it does not capture one critical feature
of today’s telecommunications business, namely convergence. As many analysts
predict, a key growth area in telecommunications that has limitless opportunities
is to combine voice, data, and video services. Convergence may play a role in
influencing entry decisions in two opposite directions. It may enhance the viability

of overbuild competition through the expansion of market opportunities that have

183

not been explored thus far. This newly found market (e.g., high-speed lntemet
access over cable) will greatly expand revenue potential for an incumbent and an
entrant alike, giving rise to a new possibility for the entrant to exploit.

At the same time, however, convergence may further impede entry by
increasing the necessary capital investment in the diversified operation for the
new entrant. If the entrant is to compete with the incumbent that operates at
multiple product spaces, the entrant will be forced to provide at least the same
product categories as the incumbent. Theoretically, there is no reason for the
entrant to match all the products offered by the incumbent. The entrant can
carve out a niche, single market without entering all the submarkets in which the
incumbent operates. But this would obviously put the entrant at disadvantage for
many reasons. First, the specialized entrant will not be able to achieve
efficiencies stemming from economies of scope. To the extent there is a
subadditivity factor, this would result in post-entry cost advantage for the multi-
product firms. Second, bundling can be a very powerful strategy for the multi-
product firms. If a firm monopolized one product, the firm can effectively leverage
that monopoly to preclude competition in another market by using a bundled
pricing strategy. Further, the ability to drive or preclude a rival from a market
using bundled pricing exists even if the rival can offer a superior alternative to the
monopolist’s non-monopolized product. Microsoft's bundling of its browser with
its own operating system offers one prominent example of such cases. In the
multichannel television industry, it has been sometimes claimed that an MSO

who owns exclusive rights to an anchor channel can preclude competition in

184

thematic channels (e.g., comedy or SF channels) by bundling their own thematic
channels with the anchor (Aron & Wildman, 1999).

In this converging marketplace, only the one who can provide multiple
services efficiently and seamlessly will survive. In order to be a player five or ten
years from now, telecommunications companies would have to offer services in
a package. In the days of bundled services, the stakes are higher in the fight to
get and keep market share for a full package of telecommunications services.
That is why major cable companies are transforming themselves from multiple
system operators (MSOs) to broadband service operators (BSOs). As the name
tells, “MSO” implies that cable services in the past wer largely confined to
“broadcast” of signals in geographically disperse areas. But advanced cable
systems today provide diverse services in package, including interactive data
and voice as well as the conventional one-way video services over “broadband”
networks. Cable operators are no longer counting people in terms of basic cable
but as revenue generating units. AT&T and MediaOne have billed themselves as
a "broadband" company, and these cable MSOs have spent heavily on system
upgrades to two-way 750-MHz fiber optic wires that can be used for high-speed-
data transmission and telephony. In the bundling opportunity, moreover, brand-
name recognition goes a long way toward keeping subscribers loyal. And,
unquestionably, AT&T enjoys high brand recognition.

Given the acceleration of companies trying to offer a full range of
broadband services, it is clear that there is a competitive advantage in offering

multiple senrices and that the cable TV providers have, at least, a short-tenn

185

lead on offering the entire package. With their core business virtually intact in the
multichannel television market, major cable MSOs are well poised to take off.
While there are certainly additional costs to make digital cable interactive, less
than 5—8 percent of the total bandwidth on a digital cable system is used for high
speed data services; the rest remains available for profitable video services. (Bar
et al., 1999).

Bundling may give significant advantages to broadband service operators
(BSOs). It is also expected to offer some benefits to consumers. Bundling may
be in the interest of both the business and the consumer. Thus, multi-service
firms will likely attempt to provide a bundle of services whenever possible. The
ability to include television offerings in its bundles certainly makes it easier for
AT&T to create distinctive packages. AT&T could, and apparently intends to,
offer integrated bundles of phones service (both local and long distance),
multichannel television, mobile services, and ISP (Bar et al., 1999). With bundled
package, most cable operators have noted no significant churn or downgrades
among their multiple-product customers. The anchor is, of course, cable
modems (Haugsted, 1999b).

Some market analysts estimate that merely the prospect of bundled
services creates approximately $150 in new value per subscriber for a cable
system, irrespective of value created by the anticipated revenue from each
individual senrice offering. There would be competitive advantages in the
package of services created, advantages in pricing those services, and

advantages in a single bill. Indeed, the consumer's preference for one bill is

186

believed to be strong enough to reduce switching, even without price reduction
for the services in a bundle (Bar et al., 1999).

Cable operators across the country have already been testing and
launching multiple-service offerings. Comcast Cable Communications Inc.
launched its first Comcast-branded long-distance phone service in Detroit in
1998, marking another entry in cable's slow but steady invasion into competitive
telecommunications markets. Comcast provides the long-distance service at a
discounted price for customers who also subscribe to Comcast cable. The
discounted rate is also available to Philadelphia-area Comcast cellular-phone
customers. The MSO revealed that the company chose the Detroit market for the
launch because it is fighting a competitive battle with Ameritech New Media in
the Detroit area for cable senrices. Comcast, like most cable companies that
provide "bundled" or "packaged" services, awards discounts to customers who
select multiple services from the company (Hogan, 1998f).

Cox Communications Inc. offers lntemet, digital-telephone and digital-
television services in select markets, such as Orange County, California. Cox
customers, too, receive discounts for multiple services (Hogan, 1998f). Cox
initiated the high-speed lntemet and data service in Orange County, Calif., in
1996, making the affluent suburb of Los Angeles the first market where
subscribers have access to telephony, the lntemet and digital TV provided by
Cox. On the @Home front, Cox has experienced a rapid growth in a subscriber
base with less than 1 percent churn. Cox is focusing especially hard on customer

service in connection with @Home. On the telephony front, Cox is fighting based

187

on price. It offers about 20 percent discount for single-line phone service
compared with US West. Customers get a stiff discount for a second line. In
addition to unlimited local calls, the offering is also being packaged with Cox
long-distance service (Estrella 8: Heam, 1998).

With multiple products for bundling, as apparent in Cox, BSOs can
choose different strategies for different market segments. For a price-sensitive
product like telephone, they cut the incumbent’s prices. For a complex and
technical product like cable modems, they may focus on customer service. This
flexibility in marketing, along with cost advantages due to scope economies,
would equip multi-product firms with a critical competitive advantage in the
increasingly competitive telecommunications industry.

Further, a bundled package of multiple services from a single provider
also implies convenient payment from the consumer's standpoint. Many cable
operators and analysts agree that integrated billing will become crucial as
customers start to demand it and telephone-company competition supplies it.
Currently, most operators that have moved into the multiple-product realm are
bundling cable and lntemet-related services, leaving a separate telephony bill
(Haugsted, 1999b).

Despite the challenges involved in moving to integrated billing, doing so
offers payoffs to telecommunications service providers. The service rewards
time-constrained customers, and it could help to reduce churn among customers
who do not want to break up voice senrice and video packages (Hogan, 1998f).

This explains why telephone companies are not just launching their own

188

competing video services. SBC and Bell Atlantic are partnering with DIRECTV,
which has built a strong national brand of its own, to co-market the direct
broadcast satellite service and to integrate its video programming onto the
telephone bill.

Can competitors create equivalent alternative bundles? In bundling, cable
systems have certain advantages. Cable’s competitors, be it RBOCs or DBS,
cannot create equivalent packages simply because they miss at least one key
component for bundling—the broadband wires already connected to 95 percent
of all US households. Local telephone networks mainly consist of copper wires.
By contrast, cable companies mainly use coaxial cable, which has much greater
capacity (or "bandwidth") than copper wire. Combined with AT&T’s fiber-optic
trunk, which has even more bandwidth than coaxial, the cable industry will
certainly become dominant at the era of convergence, which will necessarily
require a broadband connection to the home. As the telecommunications
industry moves toward offering customers more advanced high-bandwidth
senrices—such as interactive video and high-speed lntemet access—this
additional capacity will be critical for companies that want to diversify their
services.As a result, RBOCs will suffer from their lack of video entertainment.
DBS providers will certainly play a critical role in the video programming
distribution market, but they will not be able to compete with cable for high-speed
data service, at least in the near future. Cable’s ability to bundle multiple

services, particularly with the monopolistic combination of video and high-speed

189

data senrice, implies that no competitors, including RBOCs, will be able to offer
comparable alternative bundles for the consumer.

Local exchange carriers try to offer their video services to offset cable's
move into their territory. They recognize that if they do not offer bundled
senrices, AT&T and TCI will take away their share soon. That is why LECs are
taking the initiative and the leap of faith to go out and compete in some places
such as Detroit, Columbus, Atlanta and Phoenix, where LECs rolled out a
complex high-speed DSL voice, video and data offering. US West, for example,
compete with Cox in Phoenix with both fiber-to-the-node high-speed video-
capable DSL and fiber-to-the-curb (F'l‘l' C). GTE is testing the technology in the
Southeast (Barthold, 1999b). Bell Atlantic is countering MediaOne's telephony
launch in Boston by expanding its DSL services that have been available since
April 1999 and beefing up its satellite-TV offerings in the market. The RBOC
expects to begin offering customers in Massachusetts direct-broadcast satellite
service from DIRECTV. But LECs’ strategic options would never be close to
those of cable-based broadband service providers. They may resell video
products, but cannot use DBS as a leverage. It may prove to be more burden
than an asset in terms of profit-generating power. The very essence of most
competitive advantages is that they cannot be readily assembled through
markets. As the recent development in organizational theory of firms
demonstrates (see, for example, (T eece, Pisano, & Shuen, 1997), the fimi’s
competences cannot be replicated by a portfolio of business units amalgamated

just through formal contracts as many distinctive elements of internal

190

organization simply cannot be replicated in the market. To overcome this, LECs

may have to buy a significant stake in the DBS operation in the future.

First Mover Advantages for Cable BSOs

With the heightened focus on broadband technologies, however, the
current state of narrowband access often gets overlooked. Due to the ubiquity of
the switched telephone network system, the vast majority of residential
consumers continue to access the lntemet through analog modems. In January
1999, 65 percent of lntemet users were still using analog dial-up modems with
an average speed of access of 33 kbps. It is projected that dial-up will remain the
principal means of accessing the lntemet in the near future.

Nevertheless, cable is now well positioned to take the lead in the high-
speed lntemet access service market. In the short term, we can expect two
distinct broadband “footprints”, with little overlap. The Cable modem footprint
generally covers residential areas and clearly dominates in many suburbs, while
DSL targets predominately the business sector (Bar at al., 1999).

As of 1999, cable modem service has a considerable lead over DSL in
terms of total number of subscribers in the residential market. Some analysts
believe that DSL will close this lead soon, with the adoption of new technologies
and standards that will improve the performance of DSL. Most recent FCC report
predicts, for example, that given the rapid pace of DSL deployment in recent
years, the subscriber levels for DSL would be nearly comparable to cable by

2007. The FCC cites Lehman Brothers” estimates that cable will capture 14.5

191

million subscribers and DSL will capture 10.5 million subscribers by the end of
the year 2007. Another estimate comes from Donaldson, Lufltin & Jenerette
(DLJ) cited in Lathen (1999). According to the DLJ estimates, today’s lntemet
usage is based predominately on dial-up connection (96%), followed by cable
modem (3.8%) and DSL (0.8%). Rapid technology improvement and network
modernization, coupled with strong demand for broadband transmission, will
transform the landscape of “World Vlfide Wait” rather dramatically. By 2005,
about 50 percent of online users will still have to rely on dial-up connection. But
35 percent will use cable modem and the rest,15 percent, DSL (See Figure 5-1).
Considering the current pace of high-speed lntemet access deployment, this
comes as no surprise. Despite some variation in these estimates, they all show
the unequivocal trend that cable modem will drive the next generation lntemet
use and become the major, if not the dominant, platform for the high-speed
lntemet access. Even the most conservative estimate of these studies
anticipates that cable modems will have a 50 percent bigger market share for
high-speed lntemet access in 200.

Further, these estimates of DSL deployment include competitive local
exchange carriers as well. In fact, the competitive threat from CLECs and lSPs
significantly reduces ILECs market prospects with DSL services. In addition to
the ILECs, a new type of CLECs has begun to focus on the high-speed lntemet
market. Covad Communications, Rhythms NetConnections Inc., and NorthPoint
Communications Inc. all have raised billions of dollars in their initial public

offerings. These companies intend to target DSL services to small and mid-sized

192

businesses, as well as to residential customers (Lathen, 1999). Accordingly,
LECs’ share out of these DSL estimates would shrink to the extent how these
non-LEC players will capture online users’ attention in the future.

In spite of their validity in deployment rates in the near future, however, all
these estimates provide only for a static view of the high-speed lntemet access
market. It is certain that the adoption of a particular service will be influenced by
user's earlier decision and that industry players will try to take advantage of their
strategic assets at hand. In that regard, it is critical to acquire a data customer

today, not tomorrow.

193

92.66- .1. 66606:,“ ...o. - -. ..moImrl 6360+ 1-63.560 , .

xoom xoom moow Voom moom NOON room ooom mmmw

I. I: i ii: .Il . “I..- o
x.... . . ..... ho... o o M
. .. ...... ....-.... . . .66
. e .9 \ m %
O O \ _
. n .2. 6..
. .x m
\\ O i 00 w.
~ 0 a
. . n
I 0‘ _.,
-. . s a
0| 0 N
. -8.
ow...

on. 63223.. .6 Ext... .coosacoo $2.5m
.686 -666.. 362:: 65:5 6 $6386 . 2:6...

194

Looking at Figure 1, one notices that the forecasting of online universe,
like most forecasting of technology adoption, is based on a static view of the
potential growth in the high-speed lntemet access market. As the two solid lines
for cable modem and DSL growth in Figure 1 depict, these estimates imply that
the two services would increase their market position in a linear fashion. But in
reality, considering the close substitutability of cable modem and DSL, their
growth rates would take a very different path from a linear trend depicted by
these two solid lines. It is reasonable to assume that because cable modem and
DSL will compete for the same consumer dollars for broadband services at
home, the growth rate of each service will be obviously interdependent on how
the other service performs. Thus, sooner or later, it is highly likely that as one
service takes off, the other service will suffer from the take-off of its competitor.
Consequently, the growth pattern will more likely exhibit a familiar S-curve for the
successful diffusion as depicted in the dotted lines while the unsuccessful
senrice may end up with a complete extinction of use (Markus, 1990).

In light of the current status of broadband services, cable modems will
most likely be the successful choice of technology for reasons as follows. It may
be too early to judge how DSL will compete with the cable modem service in the
high-speed online access services. But low churn and customer satisfaction
suggest that cable’s first-to-market advantage will be a formidable challenge for
ADSL. The cable industry, which is now estimated to have more than 1 million
customers for high-speed data in the US, is well-positioned to take advantage of

being the “first mover” in the high-speed lntemet access market.

195

 

The first-mover advantages are related either to their demand structures
or to their cost functions. On the demand side, early entrants may be able to
mold the cost structure of customers. This can occur in three main ways. First,
the customers’ perceptual space may evolve in a manner that favors the initial
position of the pioneer. Second, customers may develop switching costs as they
accumulate experience with the pioneer's product. Thus, the first-mover can
have an advantage simply by being first, so that buyers have tried its product
and thus incurred the sunk set-up costs if any exist, or by trying it they have
removed uncertainty over its quality. Third, “network extemalities” may establish
the pioneer's product as the industry standard. In this case, customers enjoy
lower costs or greater benefits when using the standard product, which allows
compatibility with the largest base of external users (Lieberman, 1987).

The most typical form of first-mover advantages occurs when there are
set-up and switching costs associated with a product. Switching costs can take
the form of transaction costs from switching brands, learning costs, or seller-
induced costs (Mueller, 1997). With respect to broadband services, the switching
costs have two sources: the physical architecture of the network and the logical
architecture of the network. The physical architecture of the network creates
prohibitively high switching costs and hampers a customer’s ability to switch
between broadband access service providers using different physical delivery
vehicles. Different requirements for inside wiring, different terminal equipment,
non-refundable connection charges, and different computer set-ups in many

cases are among the factors that can easily push the physical cost of switching

196

from cable to DSL—where both are available— up to $600. Since most industry
surveys indicate that consumers are not willing to pay large sums for broadband
access, they are even less likely, one would presume, to pay high sums to
switch. The logical architecture of the network also creates important switching
costs. Information access and transmission systems become embedded with
one’s current provider. Consider the several costs of switching from one
broadband system to another. First, many everyday communication activities are
tightly entangled with one’s lntemet provider, so that shifting providers may
range from the inconvenient to the tme burdensome. Consider computer
software. Once one has become accustomed to one company’s software, one
must incur the costs of Ieaming another to be able to switch to another
manufacturer's product. With narrowband lntemet access, the inconvenience is
typically limited to getting a new e-mail address and modifying a few dial-up
settings. However, because broadband lntemet supports a wide range of new
communication activities, switching among broadband access providers would
be much more cumbersome (Bar et al., 1999).

Considerable switching costs (the cost customers would incur to switch
from one broadband access method to another) combine with early deployment
lead for broadband cable to allow the credible exercise of market power. The
competitive barriers resulting from these switching costs will help cable maintain
its significant deployment lead into the foreseeable future. Hence, even in the
limited areas where cable and DSL broadband access are both available,

competition between different infrastructures is highly imperfect. Once customers

197

 

make an initial decision for either cable or DSL, or later perhaps for wireless
when it is available, they must live with that decision for a while (Bar et al., 1999).

From the supply-side perspectives, first-mover advantages are related to
the structure of production costs. To the extent that some costs like R&D and
advertising are sunk, an incumbent firm faces lower costs than a potential
entrant, because the incumbent can ignore the sunk costs when choosing its
optimal output and price combination where the potential entrant must incur
both. The incumbent firm may thus be able to choose price and quantity
combinations, which fail to cover the potential entrant’s total costs (Mueller,
1997).

Furthermore, if it takes time to install capacity, and there are economies of
scale in production, then the first firm in the industry has more time to expand
and achieve these scale economies. Large firms have a greater incentive to
invest more on cost-reducing innovations (e.g., R&D) and thus their costs should
fall faster than those of small firms, leading to further increases in size. These
size-related advantages with cost-reducing innovations will then be reinforced by
any leaming-by-doing cost advantages of the first movers in a market. Since
experience accumulates with production, leaming-by-doing cost reductions
should depend on the cumulative output of the firm, and time. Together, they
add up to a potentially significant set of cost advantages that the largest firms in
a market can be expected to have. The first companies to establish a significant
production lead over their rivals can be expected to have lower and falling unit

costs, and thus will be in a position to extend their lead over time. There is

198

considerable evidence that suggests the potential for significant first-mover
advantages, even when the first-mover has not chosen the optimal product or
production process design from the set of initial candidates (Mueller, 1997).

Cable's early lead in deployment, coupled with substantial first-mover
advantages are likely to give AT&T/@Home considerable power in many
conceivable scenarios for future market dynamics in the high-speed data service
industry. In some set of markets—likely to be a significant set given the
limitations on DSL—cable will be the only broadband option. Either DSL service
will be technically impossible, or cable's initial lead in deployment will result in an
unwillingness of the local RBOC or competitors to spend resources deploying
DSL in that area. In this effective monopoly, consumers are likely to be harmed:
they will pay the fees for access that an unregulated monopolist can charge, and
they will suffer from limitations on the kinds of services offered and the degree of
experimentation offered imposed by the single access provider (Bar et al., 1999).

In some markets where asymmetric duopoly exists (i.e., cable on one
hand, DSL on the other), the typical residence will possess two active wires
capable of carrying broadband video services subsidizing high speed data
services. Consumers seeking broadband service will have a choice between the
cable-blessed access provider allowed to operate over the cable line, and the set
of ISPs and local exchange carriers buying access over the telephone line from
the local incumbent phone company. In this case, broadband cable will operate
three markets simultaneously. They are the traditional multichannel video

senrices, the broadband access market that includes high-speed data, and the

199

other services it supports such as interactive video teleconferencing and the
phone/fax access network market. An upgraded cable television network not only
hastens the provision of broadband access to households, it also permits a
second line to the home for phone/fax senrice competition and therefore
accelerates the emergence of competition in the historically forbidden market for
local telephone companies. Furthermore, cable companies have aggressively
deployed digital video services to compete with Direct Broadcast and RBOCs,
reaping extra revenues from that deployment. Those revenue sources from video
as well as voice service bring them considerable strategic clout to defend the key
battlefront, i.e., the lntemet access market. Holding a franchise monopoly for
cable television thus creates a solid foundation for cable to defend the market for
broadband access. If economies of scale and scope in broadband for the
consumer market are significant, the present headstart of the cable companies
will give them an insuperable first mover advantage and leave them with the kind
of dominance they still enjoy in their core market for multichannel video program
distribution. This is why some analysts believe that eventually cable television
companies will have an 80 percent market share in the broadband industry
(FCC, 1999a, fn. 149). The inevitable consequence is cable's "joint dominance"
in the video and lntemet access markets, which will seriously undermine LECs’

competitive position in the increasingly converging telecommunications market.

200

 

Dynamic Resource Allocation Through Cross Subsidization

AT&T and cable MSOs are not precluded from entering any market, while
RBOCs are. Precluded or not, the thought of a world powerhouse such as AT&T
combining with a dominant MSO such as TCI can be frightening. Consider the
implications of the merger of America Online and Time Warner. Major cable
MSOs’ capabilities to provide a bundle of services, backed by AT&T's acquisition
of TCI and MediaOne will profoundly affect RBOCs’ future. AT&T's acquisition of
the two major MSOs will give it a dominant position in the market for providing
consumers high-speed access to the lntemet (Kuhl, 1998). Being the incumbent
in telephony is not an advantage anymore. LECs are still worried about local dial
tone and long-distance and wireless and all the other things that they are
preoccupied with. RBOCs cannot offer the same service any cheaper than cable
operators under the current regime of regulation.

Cross subsidization may come into play in this context. Large MSOs will
be able to transfer their profits from monopoly markets to fight in the competitive
markets. Unless the new entrant operates in comparable scale throughout the
incumbent's geographic markets, cross subsidy can jeopardize the financial
profitability for the small-scale entrant. Large MSOs can also utilize their market
power in the multichannel television market to help enhance the competitiveness
of emerging services. For example, a cable operatorcan subsidize its new data
and voice services with cash flow generated from video programming operations.
It is not uncommon for the cable operator to increase cable service prices to help

establish a new business like telephony. When Cox offered telephone service in

201

Omaha, Nebraska at lower price than US West did, the LEC took issues with
Cox as to whether the MSO was truly undercutting U S West on price, since the
MSO was raising the cost of its cable service by $2 a month at the same time.
The LEC blamed that Cox used their cable subscribers to subsidize their entry
into the telephone business (Estrella & Heam, 1998). Cross subsidy may also
happen to help high-cost areas (e.g., rural) through the transfer of profits from
the low-costs areas (e.g., urban). Whenever necessary, therefore, large MSOs
can enhance their market position through the strategic use of their ability to
dynamically allocate resources. The financial prospect for the new entrant may
not be so promising, ultimately impeding entry and the vigor of competition.

There have been very significant battles in court since LECs entered into
the cable business. In particular, Ameritech New Mediaraised an issue when it
introduced the so-called “AmeriChecks" promotion in three Midwest states.

The AmeriChecks plan was offered in three states, and Illinois was the
only one to side with the regional Bell operating company. On the other hand, the
State Supreme Court of Ohio found that the program undermined competition,
while Michigan struck down the plan for discriminating against certain phone
customers.

The Ohio Supreme Court ruled in July 1999 that Ameritech Ohio cannot
use coupons to give its subscribers americast cable television service rebates on
their Ameritech local phone service. The case stems from a promotion launched
in the summer of 1997, in which Ameritech New Media issued AmeriChecks to

its cable customers in Illinois, Michigan and Ohio, entitling them up to $120 off

202

their Ameritech cable bill, local phone senrice, security monitoring or cellular-
phone bill (Quinton, 1999a).

But that promotion prompted complaints to state regulatory bodies. Local
cable associations claimed that the AmeriChecks promotion was subsidized by
the phone service, even though ANM insists that the checks were paid for by
ANM, and not by the company’s telephone division (Hogan, 1997). The Public
Utility Commission of Ohio (PUCO) found that “Ameritech attempted to aid its
cablevision affiliate by ordering, in effect, lower telephone service rates for its
customers who would use americast.” Giving those customers “the exclusive
benefit of a price break on their telephone service” constituted an unreasonable
preference under Ohio law, PUCO found. That finding was upheld in a 4-3
decision in the Ohio Supreme Court (Quinton, 1999a). Following an adverse
Ohio ruling, Ameritech customers in that state could not use their AmeriChecks
for local phone service.

Michigan became the second state to outlaw an Ameritech Corp.
marketing plan aimed at local cable viewers. In 1998, the Michigan Public
Service Commission (MPSC) issued a cease-and-desist order after finding that
the telco's controversial 'AmeriChecks' promotion violated the Michigan
Telecommunications Act. Acting on a complaint by the state's cable operators,
the MPSC found that the program discriminated against those consumers who
did not sign up for cable senrice. The MPSC order forced Ameritech to stop
accepting the AmeriChecks coupons for local phone service (Estrella, 1998c). In

its decision, the MPSC recognized that incumbent local-exchange carriers have

203

considerable resources at their disposal. As a result, they are in a position to use
these advantages not only to gain entry into unregulated markets, but also to
undermine competition once they gain entry (Estrella, 1998d).

However, in May 1999, the Illinois Commerce Commission (ICC) ruled the
same AmeriCheck promotion did not violate Illinois telephone regulations. Illinois
became the first state to rule in favor of Ameritech Corp.'s controversial
"AmeriChecks" marketing program after Michigan and Ohio found the same
promotion illegal a year ago.

The ICC found that Ameritech had not violated the state's Public Utilities
Act by offering up to $120 in vouchers to consumers signing up for its Ameritech
New Media cable service. In the ruling, ICC accepted Ameritech's argument that
the law only prohibits charging a greater or less or different compensation for
products or senrices rendered. "Because Ameritech receives the full tariffed price
for its service, the payment of all or part of customers' bills by a third party
cannot be found to violate the act, whether the third party is a customer's
roommate, an interexchange carrier or New Media," the ICC concluded.

As these cases illustrate, current regulatory regime significantly hampers
LECs’ resource allocation ability. Like the Cox case, cross-subsidization should
not be blamed on its own. It is sometimes inevitable: It is sometimes desirable. In
fact, many forms of cross-subsidies in the telephone industry have been largely
conducive to universal service policies in the US and the resulting ubiquity of

telephone services across the country. So the question is not the use of it, but

204

the effects in promoting principal policy goals set forth with respect to the
telecommunications services.

In that regard, the current regulations of prohibiting cross-subsidization fail
to envisage the changing technological and economic environments in
telecommunications. The rules and regulations governing the telephone industry
are based on classic theories of scale economies, but they have not placed
much attention on the increasingly important characteristics of
telecommunications business: economies of scope. Thus, the prohibition of
cross-marketing, as in the case of “AmeriChecks," significantly impede telephone
companies from gaining a customer base, thereby heightening barriers to entry
and market expansion. The availability of cross-subsidization would be critical on
telcos’ decision of whether or not they will enter the video industry and how long
they will successfully sustain the new business operation in deficit. This is
particularly true, considering the only means of funding the development of the
advanced telecommunications systems is by cross-subsidization (Baldwin,
1997)

Thus, the requirements of separate subsidiary for video services may in
fact block LECs ability to achieve efficiencies resulting from economies of scope.
Likewise, the ban of cross-subsidization may have led to the foreclosure of

potential entry into the multichannel television industry by LECs.

205

Chapter 6

SUSTAINABILITY OF OVERBUILD COMPETITION

This chapter further discusses the feasibility and sustainability of overbuild
competition based on two analyses: likelihood of overbuild survival, and q-ratios.
First, an investigation of changes in market structure from a sample of selected
markets was conducted. Second, Tobin’s q-ratios for the cable industry, a
traditional measure of market power, were estimated and the interpretation of
such estimates will be presented. The chapter begins with a review of prior

studies that addressed financial viability of overbuild competition.

Sustainability of Overbuild Comgtition in Simulation Studies

Economic theory is clear to the benefits of competition. Despite some
shortcomings in methodology, empirical evidence from the cable industry is
clear. Unlike rate regulation, findings on competitive effects are consistent. Its
effects are very large. Prior studies provide substantial and consonant evidence
that overbuild competition has led to reductions in price and enhancement in
quality. In fact, the effects of competition are so large and unequivocal that one
might even suspect the attractiveness of competitive entry in cable television
from a potential entrant’s standpoint. If this were a strong indication of market
predation by incumbents, the prospective entrant would reconsider entering the
cable television market.

For overbuild competition to be viable, it is necessary to resolve the issue

206

of natural monopoly in cable television. With an extraordinarily large number of
observations (all 4,800 cable systems in operation in 1981), Noam (1985)
investigated the extent of economies of scale in cable television operations. He
found overall elasticity of scale was 1.096, suggesting a 10 percent increase in
size is associated with a unit cost decrease of about 1 percent. According to
Noam, economies of scale did not appear to exist primarily in the technical
distribution aspects of cable television. Instead, they were observed for the
output definitions that include a strong element of marketing success. Owen and
Greenhalgh (1986) also found modest economies of scale with respect to
system size. They used cost information contained in proposals that were
submitted in municipal franchise bidding competition for 34 cities across the US.
for the period from 1979 to 1982. Their analysis of cost structure for the
provision of cable television revealed that overbuild competition would generate
a 14 percent penalty in unit costs per subscriber. Owen and Greenhalgh argued
that “although this is hardly negligible, it is within the range of monopoly markups
that might be expected in the absence of competition or effective regulation"
(1986, p.76). Despite the cost penalty from lost scale economies, therefore, they
concluded that consumers might be better off with overbuild competition.
Similarly, Owen and Gottlieb (1986) argued that the characterization of cable as
a natural monopoly is artificial and inaccurate.

A few studies confront the question of whether a competitive market
structure is feasible. To date, at least two approaches have been employed in

the assessment of overbuild competition. Johnson and Reed (1992) and Owen

207

and Greenhalgh (1986) took the engineering-costs approach, while Smiley
(1986) and Hazlett (1995b) used a simulation. Owen and Greenhalgh (1986)
found that the economies of scale in dimensions relavant to the feasibility of
overbuild competition are not so large as to entirely rule out the possibility of
effective actual or potential competition in the cable industry. But they also added
that “while [cost subadditivity] does not rule out the desirability of competition,
any [overbuild] competition will be fragile” (Owen & Greenhalgh, 1986, p.78).

An engineering-economics study by Johnson and Reed (1992) reached
more pessimistic conclusion. Based on the comparisons of costs for integrated
broadband, narrowband, and cable television networks, Johnson and Reed
found that an integrated broadband network is much more costly than the sum of
the separate narrowband and cable television networks. Therefore, they argue
that in constructing a hypothetical lBN in a large community, a company would
be better off by building two separate networks because the additional cost of
building an integrated network would fall short of the additional revenues. In a
situation where a cable system is already serving a community, according to
Johnson and Reed, the telephone company or a cable operator is not likely to
overbuild the existing video network. The entrant would face severe difficulties in
seeking to provide video service because of competition from incumbent cable
operators. The key conclusion of the Johnson and Reed study is that elimination
of the legal entry barriers such as cross-ownership bans and an exclusive
franchise would not greatly add to competitive pressures because telephone

companies have no special economic advantage over cable operators in entering

208

the residential video market.

A financial simulation of overbuild competition reached a similar
conclusion. Smiley (1986) has conducted an assessment of social welfare due to
direct cable competition. His cost-benefit analysis approximates the gain in
consumer surplus (e.g., price reductions of a duopolistic market) and the
increase in costs due to duplication of cable lines and central facilities. From a
simulation on benchmark cases for a typical cable system, Smiley found that
while a pure monopoly produced less total welfare than either a simultaneous
dual entry or a partial sequential entry scenario, a fully sequential overbuild
would not likely occur because the second firm earns negative profits. Smiley
(1990) states that “if it is difficult for the new entrant to differentiate its service
from that of the incumbent, entry is less likely to occur because prices are
expected to be lower in the post-entry equilibrium" (p.132). In the Smiley
simulation, the sequential entry was found to result in a consumer price that was
18.3 percent lower than that of a monopolist. Smiley concludes that the second
entrant’s best response in this case is not to enter, suggesting the same
equilibrium as in the monopoly scenario.

The Smiley study was criticized by Hazlett on three main grounds: first,
the framework lacks any plausible public choice model of the municipal decision
to award cable franchises; second, it fails to allow profit-maximizing firms to
internalize rationally the costs of duplication; third, social costs are mistakenly
overestimated in the Smiley model (Hazlett, 1990). Hazlett argues that, absent

municipal franchise restrictions and predatory behavior by incumbents, potential

209

and actual competition would have been much more widespread and would be a
significant factor in constraining rates and improving service. He presents some
anecdotal evidence from two case studies showing that oveibuild competition is
feasible.

However, in an industry with over 10,000 cable systems, and given the
idiosyncrasies of local regulators and maverick operators, as Smiley (1990)
noted, it is possible to find anecdotes to support almost any proposition. Open
entry policy notwithstanding, Smiley argues, entry would be rare because of
many factors that might undermine the economic viability of overbuild
competition.

Hazlett responded to Smiley with another financial simulation study
(Hazlett, 1995b). He analyzed the returns available to a hypothetical entrant in a
typical U.S. cable market so as to evaluate both the claim of natural monopoly
and to assess the importance of entry barriers. This hypothetical market was
assumed to have average density (93 homes passed per mile), average
penetration (62%), average revenues ($29.93/subscriber/month), and average
cash flow margin (42.9). Notably, the entrant was assumed to achieve a 50
percent market share.

Hazlett based his simulation exercise on two basic scenarios. In the
benchmark case, Hazlett hypothesized no changes in market price and output
levels; two rivals simply split the market evenly with 31 percent penetration
apiece, at current price levels. In this scenario, the entrant is able to attain an 18

percent rate of return. Hazlett argues, therefore, that “the market demand for

210

cable is easily able to sustain two competitors at current prices in the typical US.
market” (Hazlett, 1995b, p.215). In the second scenario, market price and output
levels were modified to reflect the effects of competitive entry in the real world; a
20 percent decrease in price accompanied by a 20 percent increase in
penetration. The entrant was again assumed to capture a 50 percent market
share, implying a penetration of 37 percent. Hazlett’s finding in this case was that
the typical U.S. cable market would allow an entrant to earn a 16 percent rate of
return, assuming a 5 percent annual revenue growth. A 16 percent return,
according to Hazlett, shows that “entry is financially viable in typical cable
markets, even when prices significantly decline" (Hazlett, 1995b, p.216).
Although much of the debate over cable competition has centered on the
issue of the financial feasibility and sustainability of overbuild competition, the
question still remains unclear. The present study does not attempt to resolve the
disputes. Its goal is much less ambitious. Given the importance of strategic
behavior involved in competitive struggles, it is believed that an ex post empirical
investigation may lead to the more realistic assessment of competitive entry, and

the long-haul prospects of market structure, conduct, and performance.

Is Overbuild Competition Sustainable?
The price discounting that is persistently reported in competitive cable
markets indicates that market power is exercised in monopoly situations; if other

video services were close enough substitutes to cable television service, then

211

prices would be fully constrained before additional cable entry. Instead, strong
evidence exists that prices respond, and sharply, to competitive entry.

The empirical investigation in Chapter 3 presents evidence of how difficult
overbuild competition is in cable television. It would be useful to discuss the
findings of this empirical analysis in the context of feasibility and sustainability of
overbuild competition. To the feasibility issue in cable television, a careful review
is warranted for two simulation studies that show opposite predictions.

The hypothetical outcomes from Smiley (1986) and Hazlett (1995b) can
be cross-checked against the empirical estimation conducted in the present
study since both simulations were based on similar system characteristics in
many respects. In both simulations by Smiley and Hazlett, the baseline market
size was 50,000. Smiley used 90 homes passed per mile with 62 percent
penetration, while Hazlett adopted 93 homes per mile. The sample for the
present study shows very similar system characteristics with the average market
size of 75,000 homes, a 68 percent penetration rate, and 92.5 homes passed
per mile (see Table 3 ). Thus, the assumptions in the Smiley and Hazlett
simulation models are indeed a pretty good approximation of the overall
characteristics of the typical cable system. What Hazlett’s model fails to capture,
however, is a dynamic that overbuild competition will bring about in the market
under attack. While Hazlett argues that a 5 percent annual revenue growth is
conservative because it is far below recent experience, that is precisely what
happened recently in the cable industry. Due partly to Congressional concerns of

price hikes after the March 1999 sunset of rate regulation, and partly to the

212

increasing threats of competition, the cable industry adopted an industry-wide
"limit pricing” policy in 1999. AT&T, the nation’s largest cable operator, urged
other MSOs to refrain from increasing cable rates more than 5 percent in 1999,
which was found to have worked out in the most recent FCC study (FCC, 2000).

The most critical defect in the Hazlett simulation is its static view of
competitive interaction between the incumbent and the entrant. While it is
theoretically possible for the entrant to carve a 50 percent—maybe even
higher—market share, it would be extremely difficult to achieve that goal if the
entrant’s price is the same as the incumbent’s, unless the former can
significantly differentiate its programming from the latter. Predatory price cutting
represents only one type of the strategic instruments available for the larger
MSOs. To split the market evenly with the incumbent becomes a formidable task
when there are other strategic advantages for the incumbent that were discussed
in Chapters 4 and 5. It suffices here to mention exclusive programming for the
incumbent, significant programming cost advantages for the bigger MSOs, and
the bundling capabilities of the more advanced cable systems.

Nevertheless, the only dynamic consideration in the Hazlett model is a 20
percent price reduction upon entry. But in the absence of significantly
differentiated products, the entrant would have to offer a much cheaper price
relative to the incumbent in order to capture a 50 percent market share. Or
otherwise the entrant would have to provide a significantly higher level of service

quality than that offered by incumbents. Emmons and Prager (1997) state that

213

“when a second firm decides to enter the market, it must offer a higher-quality

product (at a lower price) to attract customers away from the incumbent” (p.744).

Are New Markets Big Enough to Sunnount Double Sgueeze?
Incumbent cable operators typically respond to overbuild entry by adding

channels, upgrading their networks and offering new—and in some cases, free—
services. Since Ameritech entered the Detroit market following the 1996
Telecommunications Act, incumbent cable companies such as Comcast and
MediaOne have offered free premium channels to customers who sign a
yearlong contract. Some are reducing or freezing rates; others have added new
channels (Solomon, 1997). Hazlett himself provides numerous examples of how
costly it is to successfully overbuild: “One explanation for Telesat’s initial
overbuilding success is that Telesat charged significantly less while offering
significantly more channels than did the incumbents” (Hazlett, 1990, p.93).

Along with predatory pricing, such aggressive nonprice competition would
inevitably result in more squeezed profit levels for the entrant. Contrary to the
Hazlett simulation, therefore, economic theories and empirical evidence suggest
that the estimate of a 16 percent rate of return is certainly overestimated, if the
entrant is committed to competing for a 50 percent market share. Furthermore,
Hazlett's prediction that the second entrant will be able to attain as high as a 16
percent rate of return is very questionable if one comes to the understanding of
significant cost advantages that large incumbent MSOs enjoy. In addition to price

squeeze through market predation, the entrant must overcome a substantial

214

differential in costs—both programming and operating—between large
incumbent MSOs and an upstart entrant.

Once firms have expended the fixed costs necessary to enter the video
distribution market, they might have an incentive to expand output in order to
lower unit costs and to help recoup fixed investment. One way to do this might
be to position their services as closer substitutes for those of cable systems and
compete more strongly with those systems on price terms. Cable overbuilders
have pursued a strategy based on a certain degree of price competition with
incumbent cable systems. They appear to generally provide programming
choices that are very similar to the ones provided by incumbent cable systems
and try to draw customers away by offering lower prices (FCC, 1995).

When competitors offer close substitutes as is the case in many duopoly
cable markets, the entrant must match the incumbent’s price to gain half the
market. In fact, it is highly likely that the entrant would have to undercut the
incumbent's price to lure cable subscribers away from its competitor. Oftentimes,
this means the entrant would have to offer significant discounts, unless it is able
to differentiate its own services. Hazlett shows this in his strong argument for
open entry policy (Hazlett, 1990). Thus, incumbent’s price reduction in response
to entry makes it quite a task for a new entrant to gain a 50 percent market share
in the first place. It may be unrealistic to assume that the new entrant would cut
out 50 percent of the market at the same price level with the incumbent. A recent
survey of consumer preference conducted by the city of Tacoma, WA (Tacoma

Telecommunications Study) discovered consumer inertia; only 29 percent would

215

be willing to switch to a new entrant if both firms offer the same programs at the
same price. More than two thirds (68%) responded that they were not likely to
switch (Tacoma Public Utilities, 1997). Furthermore, when there are significant
first-mover advantage and other types of entry barriers, such as programming at
the local level and scale economies at the MSO level, it seems almost
impossible for a new entrant to share the market evenly with incumbent.

One may have different views of whether or not more than 20% price
reduction by the incumbent cable operator violates the Sherman Acts. However
categorized—be it illegal predatory pricing practice or legal aggressive
competitive price reduction—the effect on entry and its feasibility would be the
same: overbuild competition is not likely to be sustained at this price level.

Furthermore, overbuild entrants will most likely face substantial
programming and operating handicaps that stem from horizontal concentration
may create. Besides the joint and fixed nature of investment capital, and
incumbents’ market predation discussed in Chapter 3, overbuild entrants face
competitive barriers stemming from horizontal concentration at both the national
and regional levels. One of these barriers is higher programming and operating
costs. Small systems have to pay much higher prices for the programming
networks, leading to significant cost advantages for the incumbent MSOs.
Because programming is a major expense category (typically 25-30%) for
multichannel television service (FCC, 1998; Kagan Media Inc., 1997), the
"differences between what entrants pay for programming and what the major

incumbent MSOs pay have important consequences for both entrant viability and

216

the vigor of competition where it does occur" (Dertouzos & Wildman, 1998).
Moreover, a clustered system can not only enjoy extra revenue potential from
local advertising, but it may also take advantage of their market power at the
local level to secure exclusive carriage of certain programming.

Combined with predatory pricing, this post-entry cost advantage allows
the initial monopolist to threaten to reduce price below the new entrant’s average
costs, while at the same time earning profits itself. Thus, first-mover advantages
confer the incumbent monopolist a very strong strategic option: “double
squeeze.”

Entry into the cable industry may be feasible and sustainable when a
significant market growth is anticipated or new entrants can achieve substantial
cost savings from new technologies or new entrants are able to differentiate their
products. It appears that DBS providers have pursued more of a product
differentiation strategy, with the focus on their ability to offer digital services that
include programming and operational features currently unavailable from many
cable systems. On the other hand, overbuild competitors such as Knology and
RCN take a different approach. Their strategy is to offer enhanced services (e.g.,
cable modem and telephone services) together with the “plain vanilla” video
service. Municipal cable systems often share some elements of the network
facilities with their sibling utilities companies (e.g., electricity, gas, water)
operated by the same parent local government. Local exchange carriers appear
to take yet another distinct approach in the pursuit of video programming

distribution.

217

Convergence is believed to have two economic consequences with
respect to market structure. It may give rise to a significant shift in economic
perspectives of telecommunication services. The integration of voice, data, and
video services will strike down the most fundamental barriers to entry in the
traditional telecommunications industries. For the established telecommunication
service providers such as telephone and cable, convergence may help them
realize cost savings from joint production to the extent that the integration of
such services is technologically feasible and economically profitable.

On the other hand, increased economies of scope and the increased use
of multi-product bundling strategies may lead to the erection of new forms of
barriers as well as greater concentration of the multichannel television industry.
The converging telecommunications industry will require an increase in capital
investment for the new entrant necessary to provide such new services. This, in
turn, may increase absolute cost barriers to entry, which will lead to unfavorable
conditions for small-scale entry.

In sum, the sustainability of overbuild competition is a function of post-
entry price reduction, cost differences between incumbents and entrants, and
increased revenue potentials due to market expansion. The question, then,
becomes whether or not the entrant can overcome such financial difficulties with
enlarged revenue pies that will be generated in the converging
telecommunications markets. Obviously, this will depend on consumer
expenditures on such new services as voice and lntemet access services over

cable. All these factors considered together, the financial prospect for the new

218

entrant may not be so promising, ultimately impeding entry and the vigor of
competition. Worse yet, the overbuild entrant must survive not only the
competitive struggle against the incumbent cable system operator, but also the

success of DBS senrice.

The Success of DBS and Its Impact on Overbuild Eng

The recent success of DBS may reveal the significance of competitive
interaction in cable television. While direct-to-home (DTH) technology (DBS and
HSD) suffers a number of handicaps in competing with wireline cable systems,
its technological characteristics permits DBS to escape the most devastating
strategic response from the incumbent operator. Once it starts operation, a
single transmitter may blanket an entire region of the country, not to mention a
television market. This makes selective price-cutting a much more expensive
strategy for the incumbent to employ than when fixed cable plant precisely
defines the geographic scope of the overbuild. Because market predation
against DBS is practically impossible, the incumbent cable systems had to
accommodate its entry. This explains the nature of competition between DBS
and wireline cable television: the primary battlefront is on product differentiation
rather than on price. Despite the critical handicaps due to the lack of local
network signals, DBS managed to compete with the wireline industry with much
more superior digital signals and enlarged programming options for the
consumer. lmmunized from predatory pricing, DBS systems were able to steadily
increase its footprints in the MCTV industry without incurring too much financial

loss in the early period of service operation.

219

Buoyed by continued strong growth and increased investor confidence,
the DBS industry is zeroing in on 10 million subscribers and threatening to reach
20 million customers by as early as 2005 (Breznick, 1999d). In fact, bullish
industry analysts now expect DBS to surpass a 15 percent and even a 20
percent penetration of US. TV households over the next five to 10 years, despite
the accelerating national rollout of digital cable.

Analysts and industry executives believe satellite TV sales will keep
surging because of plummeting DBS equipment prices, greater picture and
sound quality, more channel offerings, improved promotions, increased
marketing budgets, superior service, high customer satisfaction and rising cable
rates. A typical cable system now charges $30 to $35 a month for extended
basic service, as opposed to $30 a month for DIRECTV's standard Total Choice
package (Breznick, 1999d). Industry experts believe the passage of "local-into-
Iocal" legislation by Congress, granting DBS providers the right to carry local
broadcast stations back to their home markets, will help even further, particularly
in large metropolitan areas.

That does not necessarily mean great, or even any, subscriber losses for
cable, however. Cable is holding its ground or suffering only minor customer
losses as the multichannel video universe keeps expanding and the larger cable
systems switch to digital. DBS does not have to kill cable to be a great business.
They can co-exist (Breznick, 1999d)

In recent years, the direct broadcast satellite business has seen a

tremendous upturn in stock prices. Like the cable industry, the DBS industry has

220

been consolidating, which is trimming the number of service providers. DIRECTV
merged with USSB and acquired Prime Star, forming the third largest MCTV
operator that has more than 9 million subscribers. The sheer subscriber base
now would give significant cost advantages for DIRECTV to compete with the
cable systems, securing its long-ten'n viability in the multichannel television
industry. Completing the successive mergers and acquisitions with USSB and
PrimeStar, DIRECTV commands over ten million DBS subscribers. The
acquisition of USSB gives DIRECTV eight more high-power frequencies, split
between two orbital slots, and allows the DBS provider to offer unified packages
of basic and pay networks.

At the same time, EchoStar Communications purchased frequency
spectrum from News Corp. and MCI Worldcom. EchoStar, with more than 3
million subscribers, also became the eighth largest provider of multichannel
video programming senrice in 1999. Many analysts believe this rationalization of
the business coupled with what appears to be favorable legislation pending in
Congress will continue to propel DBS growth. The growth potential for DBS
stocks is strong in long term. Advancements in technology, which will enable
DBS providers to sell two-way data and true video on demand, will also spur
consumers' desires for DB6 systems. Alliances such as DIRECTV and America
Online Inc. should also help propel DBS growth (Neel, 1999b).

DBS seemed to gain a big momentum in 1999. Possessing more than 10
million subscribers between them, both DIRECTV and EchoStar gained new

subscribers at a breakneck pace, thanks to the elimination of such rivals as

221

cable-backed Primestar Inc., increased marketing, plummeting equipment prices,
and rising cable rates (Breznick, 1999e).

Internal consolidation also enhances the competitive position of the
survivors in the battle against the wireline cable service. Once DBS passes the
survival test and picks up subscriber bases, its programming costs begin to
precipitate. These cost savings, in turn, allow DBS operators to subsidize the
equipment vendors in their effort to distribute satellite dishes at local retail stores.
With more than 12 million subscribers, DBS seems to gain a big momentum in
1999.

Seeking to sustain its record growth, DIRECTV also expanded its
marketing and distribution agreement with SBC Communications Inc. to cover
the Baby Bell's 18 million residential phone customers. Previously, the SBC deal
only covered apartment complexes and other multiple-family dwelling units
(Breznick, 1999e).

Also in 1999, the US. House and Senate ovenivhelmingly passed the
Satellite Home Viewer Act to allow satellite companies to offer broadcast signals
(Glick, 1999a). This legislation permits the satellite operators to beam local
broadcast stations back to their home markets throughout the US, improving
their position against cable.

Overall, there still remain many challenges for DBS to become a true
substitute to cable service. DBS service is available from high power DBS
satellites that transmit signals to small DBS dish antennas installed at

subscribers' premises. While satellite service is popular in states like Montana

222

(22 percent penetration) and Vermont (20 percent), one may question whether
DBS can be viewed as a robust substitute to cable if DBS failed to offer local-
broadcast signals, required upfront equipment and installation payments and
required subscribers to obtain additional equipment to receive service on more
than one TV set (FCC, 1998). The DBS industry must overcome these
challenges to compete more aggressively against cable.

It is clear, however, that some competitive barriers appear to be
diminishing. DBS equipment costs fall and broadCast signals become available in
many markets. Indeed, the advertising for DBS services indicates that DBS is
now targeting a broader range of consumers.

The lack of network television signals, the primary disadvantage of DBS,
was partly resolved in 1999. On November 29, 1999, a revision of the Satellite
Home Viewer Act (“SHVA”) was enacted. Under the Satellite Home Viewer
Improvement Act of 1999 (SHVIA), satellite providers are allowed to retransmit
network and network affiliate signals into local markets. The original SHVA
legislation, passed in 1988, granted only a limited exception to the exclusive
programming copyrights enjoyed by television networks and their affiliates. This
exception recognized that some households are unable to receive network
station signals over the air and allowed direct-to-home ("DTH”) satellite video
providers to retransmit network signals, but only "to persons who reside in
unserved households” (also known as “white areas”). In 1998, several federal
court decisions that resolved litigation between broadcasters and DBS program

distributors over violations of SHVA resulted in the termination of network signals

223

to approximately 1.5 million satellite customers. Subsequently, the House and
Senate passed separate bills authorizing DTH providers to retransmit network
and network affiliate signals into local markets. Immediately after the law was
signed, EchoStar began transmitting local network packages, selling for $4.99 a
month, to subscribers in 13 markets: Denver, New York, Los Angeles, Chicago,
San Francisco, Boston, Washington, DC, Dallas/Ft. Worth, Atlanta, Miami,
Phoenix, Pittsburgh and Salt Lake City. The local channel package consists of
the local affiliates of ABC, CBS, NBC, and FOX. DIRECTV began immediate
service to two markets: New York and Los Angeles. In addition to local network
channels, DIRECTV will include a national PBS in each local package, which
sells for $5.99 a month. During the six-month period beginning on the date of the
enactment of SHVIA, DBS providers may retransmit local signals without
consent. Thereafter, DBS operators will be subject to retransmission consent
mles similar to those established for cable operators (FCC, 2000).

The inability to fully serve urban areas is another barrier to competition
faced by DBS operators. DBS is a predominantly rural service and faces
obstacles to competitive access to the suburban and urban markets.
Approximately 40 percent of DBS subscribers do not have access to cable.
Signal interference may continue to limit DBS access for some viewers. For
instance, DBS antennas must face south to receive an acceptable quality signal
from the satellite, which transmits the video programming service. This means
that renters and property owners who do not have south facing exclusive use

areas cannot opt for DBS service, and therefore must rely on cable operators.

224

Finally, because DBS service relies on a telephone return path and can provide
only limited interactivity, DBS is at a competitive disadvantage to cable with its
broadband conduit to and from the home.

The very success of DBS is of great significance in assessing the
prospects of overbuild competition. As DBS strengthens its attractiveness with
local network signals and decreasing equipment costs, it would create more
difficulties for new overbuilds to compete for survival. Wherever DBS service is
available, overbuild entrants will find them as the third multichannel television
provider, not as the second. In other words, the viability of overbuild competition
will depend not only on the incumbent cable system, but substantially on the
availability of DBS service and its substitutability in a given area. In the light of its
technological handicaps and diffusion patterns over the years, DBS is expected
to continue growing in rural areas where subscribers are not well served by the
wireline cable systems and do not have the line-of-sight problems. On the other
hand, overbuild entry would almost always be limited to the more populated

areas, if it does occur.

Survivor Analysis
It might be helpful for understanding the sustainability of overbuild

competition if past experience is to give any lessons. Ironically, Hazlett (1990,
1995) provides very persuasive evidence of predatory pricing in the cable
television industry. Facing an aggressive overbuilder (T elesat) in Orange County,

Florida, in the late 19803, two incumbent cable operators dramatically responded

225

by cutting their rates in half, precisely in those regions where the incumbents
were overbuilt (Hazlett, 1990). In Sacramento, California, the incumbent poured
sales effort such as free TV sets and a $1 monthly rate for basic service into a
small area where the entrant would first compete, resulting in losses of $15 per
subscriber a month (Hazlett, 1995a). The main thrust of his arguments is that
“the constraint on entry is not so much the presence of subadditivity, as the
strategic behavior by first movers and franchising agents" (Hazlett, 1990, p.113).
Despite Telesat’s aggressive efforts in Florida, however, rumors continued to
persist that Telesat’s primary interest was not in operating cable systems, but in
greenmailing competitors (Thompson, 1989). Indeed, even before the 1990
Hazlett study appeared in publication, Telesat announced in August 1989 to sell
its two Florida systems to competitors (Multichannel News, 1989, August 14,
1989, p.22). In Sacramento, too, the entrants were unable to survive market
predation by the incumbent, deciding to exit from the market (Hazlett, 1990;
1995a).

An examination of how overbuild competition has fared for the last decade
is presented in Table 13. The 26 duopolistic markets used in the sample of the
Merline (1990) study were reexamined to check changes in market structure for
the last ten-year period. Out of 26 competitive markets, only 12 remained
duopolistic in 1999. More than half of these 26 markets have changed back to
monopolistic ones. A US. Senate report reveals similar results: mergers have
occurred between the two existing or proposed systems in 62 of the 132

communities in which second franchises have been advanced or were under

226

study (cited in Johnson, 1994). The lesson from such high mortality is clear and
simple: overbuild competition in cable television is more likely to fail than to

succeed.

227

Table 13 Changes of Market Structure From 1990 To 1999

(It) . Mule

Operators in l‘)‘)ll

Operators in Will)

 

 

 

Troy, AL: Storer v. Troy Cablevision Duopoly: AT&T v. Troy Cablevision
Mesa, AZ Dimension Cable v. Cable America Duopoly: Cox v. Cable America
Chula Vista, CA Cox Cable v. Ultronics Duopoly: Cox v. Ultronics

 

Sacramento, CA

Pacific Select TV v. Sacramento
Cable

Monopoly: Comcast

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Cape Coral, FL Telesat v. Cablevision Industries Monopoly: Time Warner

Citrus Co., FL Telesat v. Cablevision of Florida Duopoly: Time Warner v. Adelphia

Orange Co., FL Cablevision of Central Florida v. Duopoly: Time Warner v. Adelphia
Cablevision Industries

Orlando, FL Cablevision of Central Florida v. Monopoly: Time Warner
Cablevision Industries

Brunswick, GA Rentavision v. Star Cable Monopoly: Century (Charter) —> Time

Warner (2000)

Cumming, GA Cable USA v. Cable TV of GA Monopoly: Prestige Cable TV Inc.

Vidalia, GA TCI v. South] and Monopoly: Northland Cable

Warner Robins Cox Cable v. Watson Community Monopoly: Cox Cable

AFB, GA Television

Boone Co., KY Jacor v. Storer Monopoly: InterMedia -) AT&T (2000)

Frankfort, KY Consolidated TV Cable v. Monopoly: Frankfort Electric & Water
Community Cablevision Plant Board

Glasgow, KY TeleScripps v. Glasgow EPB Monopoly: Comcast v. Glasgow EPB —)

Glasgow EPB (2000)

Ann Arundcl North Arundel Cable v. Jones Duopoly: Joncs lntercable (Comcast) v.

Co., MD Intercable Intermedia

Monroe, MI Toledo Blade v. River Raisin Duopoly: Monroe Cablevision v. River
Cable Raisin Cable

Omaha, NE Cox Cable v. Metrovision Duopoly: Cox Cable v. US West

Hillsboro, NC Cablevision Ind. v. Carolina Cable Monopoly: Time Warner

Paramus, NJ Cablevision v. UA Cablesystems Monopoly: Cablevision Systems

Cleveland, OH Metro TEN v. North Coast Monopoly: Cablevision Systems

Allentown, PA Twin-County Trans Video v. Duopoly: RCN v. Service Electric Cable
Service Electric Cable TV TV

Pottsville, PA Service Electric Cable TV v. Duopoly: Time Warner —) Service
Warner Cable & Wire Tole-View Electric Cable TV v. Wire Tole-View

Henderson, TN Multivision v. CableAmerica Monopoly: InterMedia

Carrollton, TX Storer v. Planned Cable Systems Duopoly: AT&T v. Charter Com.

Sandy, UT TCI v. Insight Cablevision Duopoly: AT&T v. Insight Cablevision

 

 

—-) Monopoly: AT&T (2000)

 

 

Sourecs: Merline (1990) for 1990, Television and Cable Factbook for 1999.

 

228

 

Analysis of Tobin’s _a_

Another method to investigate the sustainability of overbuild competition is
the use of Tobin’s q-ratio. The q-ratio is equal to the market value of an asset
divided by its reproduction cost. The notion is that in a market in competitive
equilibrium, with no monopoly rents, the market value of a firm’s assets should
tend to equal their reproduction cost. This is so because potential buyers in a
competitive industry are unwilling to pay more than the replacement cost of the
finn’s tangible assets. Therefore, a competitive industry is expected to have the
unit value of q-ratio (q = 1), while monopolies will show much greater q-ratios. If
the q-ratio is greater than one, then another firm would find it profitable to enter
the market. Such entry would increase market supply, force prices and profits
down toward a competitive level, and hence reduce the market value of the
incumbent firm or the industry. The q-ratio provides a useful measure of market
power and hence, profitability of an industry because of its simplicity. The FCC
(1994) and other studies (Hazlett & Spitzer, 1997; Shooshan, 1989) have used
the q-ratios to gauge the degree to which cable operators actually exercise
market power.

To construct q-ratios for an industry, it is necessary to collect the
replacement costs and market values of an asset. For the cable television
industry, the FCC has compiled a large number of system transactions each
year. Table 14 presents three widely used measures of market value for cable
systems. For the present purpose, one needs to take a look at price per

subscriber.

229

In the early 1990s, each cable subscriber was worth slightly more than
$2,000 in system trades. But the rate roll-backs in 1993 and 1994 led to the
reduction of cable system trade values in the ensuing years. The shock of the
FCC rate cuts was absorbed in 1996, but the market price balked again in 1997
because of the Telecommunications Act of 1996, which struck down one of the
most significant entry barriers in cable television by repealing telco-cable cross-
ownership bans. Thus, these two events appeared to strongly influence the
financial prospects for the cable industry, which is reflected by the low market
price of a cable system in 1995 and 1997, as measured in dollar value per
subscriber and cash flow multiples. But by 1998, the market price of cable
systems resurged to $2,900 per subscriber. In the early half of 1999, the cable
industry witnessed an explosive leap in system transactions, with the sales price

of $3,800 per subscriber.

Table 14 Market Valuation of Cable Systems

1993 1994 1995 1996 1997 1998 1999

 

 

Jan.-
June
Number of Systems 96 64 142 98 110 1 14 60
Sold
Dollar Val. Per $2,160 $1,869 $1,829 $2,115 $2,044 $2,877 $3,849
Subscriber

 

DollarVal. Per Home $1,258 $1,123 $1,174 $1,299 $1,281 $1,776 $2,428
Passed

 

 

Cash Flow Multiple 11.3x 10.3x 9.7x 11.4x 9.2x 13.1x 16.3x

 

 

 

 

 

 

 

 

Source: FCC, Annual Assessment of the Status of Competition in Markets for the
Delivery of Video Programming, Second Annual Report (1995) through Sixth Annual
Report (2000)

230

 

On the replacement costs side (denominator) of the q-ratio, the average

construction cost per subscriber has been evaluated by many different sources.

Among others, the most reliable estimates are reported in Table 15. No doubt,

these estimates of replacement costs are based to some extent on one’s

subjective judgment, but in aggregation they can provide reliable information. In

fact, the estimates of industry observer, economists, and engineers do not vary

widely, as seen on column 2 of Table 15. The differentials appear to stem from

general inflation in construction costs. So, it seemsreasonable to conclude that

the replacement cost was about $600 per subscriber in the late 1980s, while

approximately $800 to $900 in the late 1990s. Combining these estimates with

the average market price in each year from Table 14 results in the q-ratios for

the cable industry in the 1990s.

Table 15 Estimates of Tobin’s q for the period of 1993-1999

 

 

 

 

 

Source COSI per 1993 1994 1995 1996 1997 1998 1999
Subscriber

Paul Kagan $600 3.6 3.1 3.0 3.5 3.4 4.8 6.4

(1988)

Smiley (1986) $619 3.5 3.0 3.0 3.4 3.3 4.6 6.2

Egan (1996) $700 3.1 2.7 2.6 3.0 2.9 4.1 5.5

Ciciora, Farmer, $900 2.4 2.1 2.0 2.4 2.3 3.2 4.3

& Large (1999)

 

 

 

 

 

 

 

 

 

Although the q-ratio was originally crafted to measure market power and

the accompanying profitability of an industry as an alternative to rate of return

and has been so used, it can also be used to indirectly estimate the feasibility of

231

 

de novo entry in an industry. Given its shortcomings in measuring such attributes
(see Lindenberg & Ross, 1981; McFarland, 1988; Owen & Wildman, 1992;
Smirlock, Gilligan, & Marshal, 1984), the q-ratio is perhaps much more
appropriate to measure the feasibility of entry in an industry. The logic is simple. /
If an industry exhibits a high value of q-ratio, not only does it demonstrate the
levels of excess profitability available to the producer, but it also suggests the
difficulties of de novo entry in that industry. When entry is open, a potential
entrant would have two options; buy out or enter frOm scratch. If de novo entry is
believed to be feasible, (i.e., profitable), the potential entrant would not be willing
to pay more than the replacement cost to purchase existing assets. It is better to
build them from scratch. In this case, the industry will have a q-ratio that is close
to unity because there is little reason to pay such premium. On the other hand, if
de novo entry is not profitable for whatever reason, the potential entrant would
have two choices. It can choose not to enter or be willing to pay more than the
replacement costs, if the potential entrant is committed to the industry.
Regardless of the decision outcomes, the industry will show a high q-ratio. The
point is that the more difficult de novo entry, the greater the q-ratio.

Looking at Table 15, one immediately notices that the cable industry
exercises its market power and enjoys high profitability, when viewed from the
conventional interpretation of the q-ratio. The other side of the coin is that it is
very difficult to enter this industry from scratch. Depending on the estimate of the
replacement cost, q-ratios for the cable industry vary, but an indisputable fact is

that the market for cable systems supports little of the feasibility for overbuild

232

entry. Among the various estimates of q in Table 15, the best estimate of the
replacement cost in the cable industry for the present time would be those by
Egan (1996) and Ciciora, Farmer, and Large (1999). Then, the best estimate of
the q ratios for 1999 would be somewhere in the range of 4.3 to 5.5. These
estimates are very close to earlier estimates. Paul MacAvoy, a Yale economist,
submitted a comprehensive analysis of q to the FCC 10 years ago. His “best
estimate” of q for the cable industry as of 1989 was 4.3 (FCC, 1990, Appendix
E). FCC’s own estimates for 1993 and 1994 ranged from 3.95 to 5.23 (FCC,
1994c).

These high level of q-ratios and no significant changes between 1989 and
1999 suggest two conclusions. First, the size of q-ratios indicates that the cable
industry is reaping excess profits as of 1999. Despite the federal govemment’s
incessant efforts in the past 10 years to encourage competition in the cable
industry, they failed to constrain the exercise of market power by the cable
operator. Second, there must be some impediment to the production of cable
service. It appears that the elimination of legal entry barriers would not
necessarily lead to pervasive entry because of other than legal and political entry
barriers in the cable industry.

As alluded to earlier, the relatively lower q-ratios during the period from
1995 to 1997 are the consequence of the FCC’s rate roll-backs and open entry
policy codified in the 1996 Telecommunications Act. But the most striking
change in q ratios in the 19903 has occurred since 1998 in which the q-ratio

started to climb significantly. This resurgence is partly due to the rediscovery of

233

revenue potential in the cable assets that may be created through the provision
of new service offerings like digital cable, lntemet access, and IP telephony. But
equally important, the market for cable systems has learned from the previous
years that despite the much hyped open entry policy, overbuild entry by local
exchange carriers and others would not materialize as easily and quickly as
anticipated and neither would it pose insurmountable competitive threats to the
incumbent should entry ever happen. To be sure, when intangible assets
account for much of the market value, then q-ratios may be a biased estimate in
measuring the viability of de novo entry. An exclusive cable franchise and
goodwill would be an excellent example. The high q-ratios in the cable industry in
the past perhaps reflected the market power granted by the local franchising
authorities and federal laws that prohibited overbuild entry. The high q-ratios
even after 1996, however, appear to indicate that overbuild is not a viable means
of entry. Further, because the numerator of q, a finn's market value, reflects a
flrrn’s expected future profits with an automatic adjustment for risk, the market for
cable systems does not seem to support the notion that overbuild entry is
feasible and profitable in the future.

The preceding discussion based on econometric results, survivor analysis,
and Tobin's q may explain why there has been no nationwide rush into the cable
market even after the most significant entry barriers were struck down. And the
market predation and the squeezed profitability may be the cause for Ameritech
and SNET’s decision to stop pursuing further franchises for cable television

(Jones, 1999). No rational businessmen would decide to enter an industry that

234

promises less than the cost of capital. Neither would they pay such stiff
premiums as 4 times the replacement cost to buy cable systems, if they believe
de novo entry via overbuilding is indeed profitable. This is especially true when
they see a much more attractive industry looming in the horizon. With limited
resources, local exchange carriers would opt for the more promising new market,
namely high-speed lntemet access, in confrontation with incumbent cable
systems’ price warfare.

When there are promising business opportunities in the marketplace, it is
usually businessmen, not economists, who will first notice and exploit such
opportunities. In an open market where no legal entry barriers exist, it is actual
entry, not conjectures through simulations, that would tell the true story about the
profit potential in that market. If overbuild competition were indeed feasible and
viable, there should have been wide-scale entries since 1992 and much more so
since 1996. The Cable Television Consumer Protection and Competition Act of
1992 specifically prohibits franchising authorities from writing exclusive
franchises. Under the Act, franchise authorities may not “unreasonably refuse” to
award an additional franchise. The Telecommunications Act of 1996 repealed
the line of business restrictions imposed on local exchange carriers under the
Modified Final Judgment. The 1996 Act adopted an open entry policy. But with a
notable exception of Ameritech and SNET, local exchange carriers have not
been enthusiastic about overbuilding the cable market. Where they did overbuild,
they faced sharp price reductions by incumbents. As of October 1999, it looked

like Ameritech and SNET were preparing retreat.

235

Chapter 7

CONCLUSION

Prosgcts of comgtition in the multichannel television market

Cable television systems now pass virtually every home in the US. and
provide multichannel television service to approximately 66 percent of them. The
34 percent of residences that choose not to subscribe to cable service do so, for
the most part, for reasons other than lack of money. Incumbent cable television
systems continue to dominate the multichannel television market, although
competition for them is growing. Nevertheless, competition between two wires is
not flourishing. The increased level of competition in the multichannel television
industry is attributed mainly to DBS.

Recently SBC decided to retreat, at least for a while, from the
multichannel television industry by holding any further expansion of Ameritech
New Media and SNET’s cable business. Some cable industry officials view
SBC's decision to halt Ameritech cable operation as a prelude to the RBOC
getting out of the cable business (Jones, 1999). The decision also suggests that
there are formidable entry barriers to the multichannel television industry.

The overbuilding business has always been risky. The changes in
technology and public policy notwithstanding, significant obstacles to competition
remain today. Most importantly, sharp price reductions in overbuild markets
significantly influence entry decisions in the multichannel television industry.

Market predation compresses profit margins for the incumbent as well as an

236

entrant. Large MSOs, however, may still enjoy substantial cost advantages and
profits because of lower input costs, associated with both programming and
operating. Thus, an entrant suffers “double squeeze” where predatory pricing
pushes post-entry price down from above while higher input costs further
squeeze profit margins for the entrant from below. Moreover, there are economic
forces associated with convergence that may further impede entry into the
multichannel television industry. If an entrant is compelled to provide a bundle of
senrices, i.e., not only video but also data and voice, to compete with the
incumbent’s product space, convergence would mean extra capital requirement
for the entrant. Together, the combination of these factors may reduce
competition, rather than strengthen it in the multichannel television industry.

Considering all these factors, overbuild competition may not materialize
as expected. Competition between the two monopolies will come first in the high-
speed lntemet access service. Telcos are more likely to focus on data and voice
services. The cable industry will move toward interactive video and data
services. The first front that telcos and cable companies will meet nationwide will
be data service. Video service will remain under monopolistic control by cable
operators. The only significant inroads to the multichannel television industry
would be provided by DBS services. Voice service will remain under lLECs
control for the foreseeable future but the cable industry will gradually expand into
voice service market. In this regard, cable operators will be ready more quickly
than telcos to provide a “bundle of integrated services”. To the extent that

consumers would like to get their telecommunications senrices from a single

237

provider, therefore, cable operators will have significant strategic advantages
over their telephone competitors. The findings of the preceding chapters

explaining each of these factors are summarized below.

Entry and Market Predation

The present research conducted an econometric analysis of overbuild
competition, using a sample of 247 observations encompassing 124 monopoly,
91 duopoly, and 32 municipal systems. The effects of competition and
nonprivate ownership variables are all large, negative, and statistically very
significant in both equations. First, overbuild competition leads to a price
reduction by 20.7 percent. Second, municipal cable systems offered basic cable
service at prices that were 25.7 percent lower than those charged by monopoly
operators. Third, competitive threats from a “potential entrant" lead to a price of
basic cable service that is 7.7 percent lower than that of monopolists. This
differential may be an indication of “limit pricing" by incumbents in an effort to
prevent actual, full-blown entry.

Of particular interest is the differential between duopoly and municipal
rates. When compared to basic service rates charged by monopoly systems,
competitive systems charge 20.7 percent less and municipal systems 25.7
percent less. The difference between these latter two groups (5%) may shed
light on the existence of market predation and the feasibility of overbuild

competition in the long run.

238

In the service quality dimension, an operator facing overbuild competition
offers, ceteris paribus, 9 to 11 percent more channels than a monopolist,
depending on the model estimated. This would translate to 4 to 5 additional
channels on a duopolist cable system since the typical monopolist offers 41
cable-only channels or 52 basic channels in total.

Another interesting set of findings involves multiproduct offerings on the
modern cable systems. Cable systems that also provide lntemet access service
over cable are predicted to charge 4.6 percent more for video service than
systems that do not offer such a service. Cable operators seem to take
advantage of cross-subsidy through the transfer of their monopoly profits from
the plain old cable television operation to market such new services as high-
speed lntemet access services. On the other hand, cable price is significantly
cheaper when the system sells a bundle of multiproducts including telephone
services as well as lntemet access service over cable. This may indicate the

existence of economies of scope in the joint production of video, data, and voice.

Concentration and Competition

As far as competition is concerned, there are three ownership issues that
characterize the cable industry: horizontal concentration and relationships
among cable operators; vertical relationships between system operators and
program suppliers; and clustered cable systems, whereby cable operators

consolidate ownership within geographical areas.

239

Cable television still is the dominant technology for the delivery of video
programming to consumers in the multichannel television marketplace, although
its market share has continued to decline since the mid 1990s. Consolidations
within the cable industry continue as cable operators acquire and trade systems.
The seven largest operators now serve almost 90 percent of all US. cable
subscribers. The four largest cable MSOs (AT&T, Time Warner, MediaOne, and
Comcast) have served approximately 54 percent of all MCTV subscribers since
1995. The largest eight MSOs (CR8) take up more than 70 percent of the
multichannel television market. The only prominent change is DIRECTV with a 9
percent nationwide market share that ascended to one of the largest four firms in
the multichannel television industry in 1999. With DBS excluded, internal
concentration within the cable industry continues to increase.

Through numerous acquisitions and trades since the mid 1990s, cable
MSOs have continued to increase the extent to which their systems form
regional clusters. MSOs are trying to form a "cluster” around major cities to
reduce costs, to improve operating and management efficiencies, to eliminate
system redundancies, and to attract more advertising. But more importantly,
regional clustering also enhances MSOs ability to compete successfully in the
cable services markets under competition. Clustering also provides the critical
mass of subscribers necessary to support the huge capital investment needed to
make system upgrades designed to enable companies to enter other lines of

telecommunications services, such as lntemet access and local phone service.

240

Although there are efficiencies realized from the ownership relationships
in the cable industry, there is also the potential for adverse market effects from
these relationships. Significant size differences between incumbents and
entrants result in first-mover advantages for the entrenched MSOs. Substantial
discounts for license fees available only to larger MSOs, combined with market
predation, may undermine healthy competition.

The difficulties in overbuild business are more intensified if the incumbent
is one of the larger MSOs that are vertically integrated into programming
distribution stage. The extensive vertical integration in the cable industry
magnifies the potential anticompetitive impact of price discrimination. Loopholes
in existing statutes allow incumbent MSOs to negotiate exclusive contracts with
attractive cable programming networks that deny these programming sources to
competitors. Leveraging their large subscriber bases and extensive geographic
coverage, incumbent MSOs may find it profitable to outbid entrants for exclusive
rights to these networks. There is the actual or potential problem of migration of
programs from satellite to terrestrial delivery in order to avoid the program

access rules.

Convergence of Telecommunication Services

Convergence can best be defined as the realization of subadditivity in
production due to joint production. Thus, the economic implications of
convergence become closely related to economies of scope. Viewed from this

perspective, one major impact of convergence is through its effects on market

241

structure and competition. To be sure, convergence holds the promise of
bringing increased competition to many telecommunications markets by allowing
existing service providers in what has traditionally been separate markets to
cross over into one anothers industries. On the other hand, as discussed earlier,
potential entrants will be required to raise more capital, if they are to compete
with the multiproduct cable MSOs.

Local exchange carriers have shown interests in multichannel television
markets, but the extent to which they will rely on their own facilities (as opposed
to resale of those of satellite operators) remains to be seen. Cable companies
eventually will go after the local telephone markets in major urban areas, but
doing so has proved more difficult than many industry optimists predicted. In
many cases, cable companies will first try specialized services, such as cable
modems, to establish themselves.

Of particular interest is the emergence of new industry, namely the high-
speed lntemet access service market. Quite independently of any inherent
technological advantage that cable or DSL may possess, cable modem’s early
lead in deployment means that cable will have first-mover advantages at least in
the residential market. Cable MSOs will have three main strategic advantages:
bundling capability, first-mover advantages, and dynamic resource allocation.

The ability to include television offerings in its bundles certainly makes it
easier for AT&T to create distinctive packages. AT&T could, and apparently
intends to offer, integrated bundles of phones service (both local and long

distance), multichannel television, mobile services, and ISP. Cable's early lead in

242

deployment, coupled with substantial first-mover advantages are likely to give
AT&T/@Home considerable power in many conceivable scenarios for future
market dynamics in the high-speed data service industry.

Recently, cable companies have aggressively deployed digital video
senrices to compete with Direct Broadcast and RBOCs, reaping extra revenues
from that deployment. Those revenue sources from digital video as well as voice
service bring them considerable strategic clout to defend the key battlefront, i.e.,
lntemet access market. Holding a franchise monopoly for cable television thus
creates a solid foundation for cable to defend the market for broadband access.
If economies of scale and scope in broadband for the consumer market are
significant, the present headstart of the cable companies will give them an
insuperable first mover advantage and leave them with the kind of dominance
they still enjoy in their core market for multichannel video program distribution. It
is very likely that the cable industry will dominate the high-speed lntemet service
in the future. The inevitable consequence is cable’s “joint dominance” in the
video and lntemet access market. And this would seriously undermine LECs'
competitive position in the increasingly converging telecommunications

industries.

Sustainability of Overbuild Competition
Combined with predatory pricing, the post-entry cost advantage allows the
initial monopolist to threaten to reduce price below the new entrant's average

costs, while at the same time earning profits itself. Thus, first-mover advantages

243

confer the incumbent monopolist a very strong strategic option: “double
squeeze.”

Furthermore, the proper assessment of the viability of overbuild
competition must take DBS into consideration. The viability of overbuild
competition will depend not only on the incumbent cable system, but
substantially on the availability of DBS service and its substitutability in a given
area. As DBS strengthens its attractiveness with local network signals and
decreasing equipment costs, it would create more difficulties for new overbuilds
to compete for survival. Wherever DBS senrice is available, overbuild entrants
will find them as the third multichannel television provider, not as the second. In
sum, the DBS factor will further exacerbate the viability problems for overbuild
competition.

This finding is reinforced with the survivor analysis. The 26 duopolistic
markets used in the sample of the Merline (1990) study were reexamined to
check changes in the market structure for the last ten-year period. Out of 26
competitive markets, only 12 remained duopolistic in 1999. More than half of
these 26 markets have changed back to monopolistic ones. A US. Senate report
reveals similar results: mergers have occurred between the two existing or
proposed systems in 62 of the 132 communities in which second franchises
have been advanced or were under study. The lessons from the high mortality of
overbuild competition are clear and simple: overbuild is more likely to fail than to

succeed.

244

Further, high q-ratios found in the late 19903 may confirm the difficulties
that overbuild competition encounters in the marketplace. The high level of q-
ratios and no significant changes between 1989 and 1999 suggest two
conclusions. First, the size of q-ratios indicates that the cable industry is reaping
excess profits as of 1999. Perhaps more importantly, high q-ratios suggest that it
is very difficult to enter the cable industry from scratch. The impediments may be
the combination of such factors as market predation, cost disadvantage, and

increased capital requirements for new services.

Issues and Policy Re_commendations

Regulation of Prices and Service Areas

As discussed in Chapter 3, cable service rates keep rising and they are
expected to continue in the near future where “effective competition” is absent. It
is so particularly because of the rate sunset since March 1999. Overbuild has
been growing in absolute number, but it has not spread out nationwide. Further,
overbuild competition from LECs now seems to face a stalemate considering the
recent halt of franchise negotiations and construction after SBC’s decision in
October 1999. It remains to be seen whether cable operators will show
opportunistic behavior as they did in the late 19803. But it is highly likely that
cable systems exempt from "effective competition” will increase rates as much as
they can.

In markets where overbuild competition has partially occurred, regulation

of rates and service areas becomes a very troublesome issue. Congress

245

established the uniform rate requirement in the 1992 Cable Act to prevent cable
operators from having different rate structures in different parts of one cable
franchise, and to prevent cable operators from dropping the rates in one portion of
a franchise area to undercut a competitor temporarily. Section 623(d) of the
Communications Act requires that: "A cable operator shall have a rate structure,
for the provision of cable service, that is uniform throughout the geographic area in
which cable service is provided over its cable system." In implementing the 1992
Cable Act, the FCC permits, consistent with the requirement of a uniform rate
structure, a cable operator to establish some differences in rates between
separate categories of subscribers. For example, nonpredatory bulk discounts to
multiple dwelling units (MDUs) are permissible if offered on a uniform basis.
Multichannel television distributors can realize significant efficiencies and cost
savings by senrice to MDUs and other high-occupancy buildings, and this way it is
possible to avoid foreclosing the prospect that those savings might be passed on
to consumers in those dwellings. Allowing cable operators to respond to
competition in individual MDUs gives consumers the benefit of lower prices from
incumbent cable operators.

The 1996 Act retains the uniform rate requirement for cable operators not
subject to effective competition but authorizes affected cable operators to deviate
from their uniform rate structures in response to competition at MDUs. The 1996
Act clarifies that the uniform rate requirement does not apply where the cable
operator is subject to effective competition and does not apply to programming

offered on a per channel or per program basis. The 1996 Act also exempts bulk

246

discounts to MDUs from the uniform rate requirement, and prohibits a cable
operator from charging predatory prices to an MDU. For the purpose of the 1996
Act, a bulk discount is a volume discount, available to all residents of the MDU.
Congress' objective, that cable operators have the flexibility to offer discounts to
MDUs, is satisfied if the discounted rate is offered to all residents of the MDU.
Thus, the bulk rate exception does not permit a cable operator to offer discounted
rates on an individual basis to subscribers simply because they are residents of an
MDU, but rather requires a bulk discount agreement negotiated by the property
owner or manager on behalf of all of the tenants.

Incumbent cable operators have heavily utilized this exemption to retain
subscribers in overbuild areas and to block further overbuilding throughout their
franchise area. The dilemma is evident. Without the exemption, consumers may
not benefit from competition in the short term. With the exemption, the entrant
has to suffer significant loss due to the incumbent's selective price cutting only
within those overbuild areas. Balancing the costs and benefits of this policy is a
tough task, but it seems that the long term prospect of competition is dimmed if
the exemption is continued. Clearly, consumer benefits from complete overbuild
must outweigh consumer welfare available with the exemption. Furthermore, the
problem of the exemption is more acute considering the pattern of overbuild
entry. A new entrant may have no intention of serving all of the communities at
the initial stage, but it may target only affluent, high-density properties. When the
entrant has to build its network from scratch, a strategy of attacking high-density

properties first is appropriate. They appear to be MDU-rich communities, as

247

opposed to single-home, lower-density communities. It the entrant starts to serve
affluent communities, which is very likely, and the uniform rate requirement is
exempt so that incumbent cable operator can respond to the entry, the benefits
of this competition go to the rich at the expense of the poor. In other words, the
incumbent’s price reduction would be supported by cross subsidization from the
poor communities where competition is absent, to the rich where competition is
present. It would be a very troublesome situation that has happened, and might
happen, in many overbuild areas. Thus, a uniform rate requirement should apply
regardless of the existence of effective competition.

The dilemma may be alleviated if “universal service” requirement is
enforced. Without the requirement of universal service in a franchise area, the
entrant will likely take advantage of "cream-skimming" and economic "redlining"
may occur. And this is where the distortion of welfare transfer arises. However,
the universal service requirement may be enforced with flexibility since it could
deter entry from the beginning. As discovered in Chapter 3, “potential
competition” can still constrain the incumbent’s market power to increase rates.
The incumbent will not be able to increase rates substantially even when only
small portion of its franchise areas has been overbuilt. It is not necessary for a
community to be completely overbuilt to realize competitive benefits.

Should there be some form of rate regulation again in the future, if the
rising trend continues? Instead of relying on ineffective regulations of prices, we
must find other solutions to check the price and to ensure enhanced customer

service. Maybe the DBS industry provides the answer in certain markets.

248

Because of impracticality of widespread overbuild competition, Smiley (1990)
noted Schumpeterian competition is more likely to come from other technologies,
such as direct broadcast satellite than from conventional cable overbuilds.
Schumpeter talked about a different form of long-ten'n competition that
threatened the possessors of monopoly power. He called this competitive force
"the gale of creative destruction."

Will DBS be the gale? Although the DBS is still in its infancy and there are
substantial entry barriers, as revealed with LECs’ recent withdrawal from
pursuing cable franchises, entry barriers and strategic control are never totally
impregnable. But a number of technical and economic factors makes DBS very
weak substitute for cable television in some regions, particularly in major cities.
This is the most important handicap that DBS must overcome to bring

competition in the Schumpeterian sense.

National Ownership Limits

The cable industry is urging federal regulators to relax cable-system-
ownership rules in light of the rapid development of direct broadcast satellite as a
robust multichannel-video competitor. NCTA, for example, contends that one
cable operator should be allowed to serve "well beyond 35 percent" of all
subscribers with multichannel-video services. NCTA argues that cable operators
are no longer the make-or-break bottlenecks in the programming-distribution
market that they were in 1992, when Congress passed a law requiring the FCC

to adopt horizontal constraints (Heam, 1998c).

249

With the purchase of TCI in 1998 and MediaOne in 1999, AT&T now
reaches more than 40 million households. After failing to penetrate local phone
markets by reselling incumbent services, AT&T is planning to spend lavishly to
upgrade TCI’s wires in order to offer voice telephony and high-speed lntemet
access, along with video programming and long-distance telephone service.
AT&T Corp., looking to protect its multibillion-dollar cable-television investment,
is urging the FCC to ease up on cable-system-ownership rules as a way of
promoting local phone competition.

In many respects, the old rules seem to be inappropriate for today's
converging telecommunications market. In the world of convergence, size
matters. While cable systems are potentially capped at reaching no more than
30 percent of homes, the Baby Bells are unencumbered by regulations setting
horizontal-ownership limits. Bell Atlantic Corp., after its acquisition of GTE Corp.,
will control 38 percent of the country's 170.5 million local phone lines, and SBC
Communications Inc., after its purchases of Ameritech Corp. and Southern New
England Telecommunications Corp., will control 40 percent (Heam, 1999).
Viewed this way, the 30 percent cap would be counterproductive because cable
companies need to extend their reach to at least match the footprints of
regionally concentrated local phone companies. Thus, if one views convergence
as a means of competition that cable and telephone invades each other’s
market, keeping the cap at 30 percent is no longer justified.

When cable and telephone companies begin to provide an array of

services to consumers—with advanced set-top boxes that provide a gateway to

250

cable networks, the lntemet and voice services—counting cable subscribers for
the purpose of enforcing a horizontal cap could be difficult. "A consumer with one
of these devices receives a broad range of senrices, and this makes 'counting'
the consumer as a cable subscriber overly simplistic” (Heam, 1999).

On the other hand, as far as the multichannel television industry is
concerned, MSO size has been associated with efficiency gains and market
power. Far from the economic justification of such large size usually provided by
large MSOs, the econometric analysis in Chapter 3 suggests that consumers do
not appear to receive benefits of lower price from large MSOs. Consumer
advocate groups, hence, urge the FCC not to relax the 30 percent cap, saying
that the benefits of cable-MSO consolidation "have not reached the public”
(Multichannel News, August 17, 1998). Large MSOs’ market power or
“bargaining power' vis-a-vis programming networks has been found in prior
studies and it may undermine the vigor of competition as discussed in Chapter 4.

Thus, the ownership issues become more complex in the converging
telecommunication services industry. More research is necessary to resolve
these conflicting effects of ownership rules. It is important to remember that the
cable ownership rules should be redesigned to prevent further consolidation of
cable’s traditional video senrice, while encouraging AT&T and its smaller
competitors to continue investing in the broadband pipeline.

There is also growing concern that the constitutionality of the FCC's new
cable ownership rules could be challenged in court, freezing the entire AT&T-

MediaOne deal. That happened in 1993, when Daniels Cablevision Inc. sued the

251

commission, saying the federal statute it used to limit the number of subscribers
an MSO can pass violated cable operators' First Amendment rights (Glick,
1999b). The 30 percent cap has been stayed by the FCC indefinitely in response
to this decision. In 1998, the FCC reaffirmed the 30 percent cap and refused to
lift the stay (Heam, 1998c). But, the FCC must show that why there should be 30
percent limit, instead of 40 percent or maybe 50 percent. There seems to be no
clear evidence to keep or reject the 30 percent cap. Apparently, this rule needs

more scientific evidence than hunches.

Programming Access Rules

There were a series of disputes about access to programming between
the cable television industry and its emerging telco competitors. Under the 1992
Cable Act, vertically integrated programming entities that are more than 5
percent owned by an MSO cannot withhold or refuse to license programming to
another video network operator.

Under current law, most cable companies that own networks are required
to sell programming to any company that wants it. But Ameritech and other
companies have complained that many networks not covered by the rule are
entering into exclusive contracts with incumbent cable companies, blocking them
from offering popular programming, such as ESPN's Classic Sports Network

The loophole in the current law has to do with the mode of program
delivery. The FCC program-access rules apply only to satellite-delivered

programming. They do not apply to all producers of programming. Current law

252

does not apply to broadcasters. It does not apply to non-vertically integrated
cable operators. In addition, cable programming owned by cable operators that is
satellite-delivered may not be withheld from competing distributors of
multichannel video programming. In practice, this provision has been used in
making exclusive agreement between M803 and programming networks that are
delivered via other than satellite (e.g., terrestrial distribution over fiber optic
network).

Cable operators are taking advantage of those exceptions to strike
exclusive deals with programmers and deny program access to competitors.
DIRECTV has filed a program-access complaint against Comcast Corp. over the
MSO's refusal to sell Comcast SportsNet, the Philadelphia regional sports
network that Comcast distributes by microwave. The No.1 DBS operator said
that in Philadelphia, Comcast has used a loophole in the access laws to get
around selling programming to DBS providers. By transmitting through terrestrial
means, instead of by satellite, Comcast is not bound by the current law (Maggi,
1998). Bell Atlantic said it has proof that such activity has limited its profitability.
The telco, embroiled in a battle with Rainbow Programming Holdings, an affiliate
of Cablevision Systems, claims that it has lost potential subscribers in Dover
Township, N.J., because Rainbow has denied it access to SportsChannel New
York (Levine, 1997a).

In May 1999, Cable overbuilder RCN Corp. filed a program-access
complaint at the FCC, accusing Cablevision Systems Corp. of illegally

withholding sports programming from RCN's New York City system to benefit

253

Time Warner Cable in Manhattan. RCN filed the complaint after Ieaming that
Cablevision refused to sell it MSG Metro Channel, which Cablevision is partially
using as a sports "overflow" channel. Overflow occurs when New York's many
area sports teams are playing at once and can't all be seen live on Madison
Square Garden Network or Fox Sports New York. RCN alleged that Cablevision
agreed to sell MSG Metro to Time Warner in Manhattan, but not to RCN, which
has about 50,000 Manhattan subscribers. Cablevision can withhold MSG Metro
because it is not satellite-delivered. RCN has access to satellite-delivered MSGN
and FSNY, where the bulk of the games appear. "As a result of the loss of the
overflow programming that has been moved to MSG Metro, RCN has been
besieged by complaints from its subscribers, and it has suffered damage to its
business reputation prompted by the loss of such programming," RCN's
complaint said (Multichannel News, June 7, 1999).

To level the playing field, therefore, these competitors insist change to the
program access laws. Ameritech is asking Congress to extend the exclusivity
ban to broadcasters that own cable networks and to independent cable
programmers such as Viacom Inc. The Chicago-based Baby Bell also wants the
law to guarantee that programming delivered by means other than satellite
cannot escape the exclusivity ban (Heam, 1998b). Consumer groups urge the
FCC to cover terrestrial delivery under the rules. In Chapter 4, we discussed that
cable MSOs’ regional clustering gives operators an incentive to migrate
programming off satellite. Given the incentives and abilities of incumbent MSOs,

the FCC needs to be more vigilant in enforcing the program access laws.

254

Cross-subsidization

As illustrated in the “AmeriChecks” and Cox cases in Chapter 5, current
regulatory regime significantly hampers LECs’ ability of resource allocation.
Cross-subsidization is not the evil on its own. The question is not the use of it,
but the consequences in promoting principal policy goals that cross subsidization
is designed to achieve. The current rules and regulations governing the
telephone industry are based on classic theories of scale economies, but they
have not placed much attention on the increasingly important characteristics of
telecommunications business: converging services and the resulting economies
of scope. The failure of adopting this change is reflected in the prohibition of
cross-marketing, as in the case of “AmeriChecks,” which significantly impedes
telephone companies from gaining a customer base, thereby heightening
barriers to entry and market expansion. Cross-marketing increasingly becomes
critical for telecommunication service providers when they enter new markets. It
is imperative for the new entrant to gain critical mass of subscriber base as
quickly as possible because it will determine how long the entrant must endure
financial deficits. This is particularly true, considering the most likely means of
funding the development of the advanced telecommunications systems is by
cross-subsidization (Baldwin, 1997). To limit telcos’ ability to allocate resources
may have a harmful effect on entry rate.

My suggestion is, therefore, that telephone companies he allowed to raise
the capital necessary for establishing their video division or integrated

telecommunications systems through cross-subsidization. But it is necessary to

255

address how to resolve the major risks that have historically attached to that
practice. There are two fundamental concerns with cross-subsidization by
telephone companies that enter the cable business: potential anticompetitive
practices and the protection of captive rate payers. These concerns are based
on the assumption that telephone companies would have the incentive and
ability to shift the costs and revenues of its unregulated activities to the detriment
of consumers and competitors alike. Cost shifting on the part of the telephone
company would have adverse consequences for both regulated ratepayers and
cable competitors. Telephone service consumers would face higher rates due to
the burden created by additional costs inappropriately shifted from cable services
to regulated operations. Cable operators, on the other hand, could be placed at a
competitive disadvantage due to the ability of telephone companies to reduce
prices in the cable service market (NTIA, 1988).

Recent developments in telecommunications, however, have already
alleviated these concerns associated with cross-subsidization by telephone
companies. First, it is highly unlikely for LECs to engage in such anticompetitive
abuse as predatory pricing because the cable industry has grown big enough to
defend itself now. The trend of consolidation will continue in the future and some
analysts anticipate that there will be only a handful of cable operators nationwide
eventually. Furthermore, the entry of AT&T into the cable industry will
significantly enhance their competitive position vis-a-vis LECs.

Second, the widespread implementation of incentive regulations for the

local telephone operation allows for the protection of captive rate payers. Under

256

the incentive regulation plans, most notably price caps, telephone service rates
are regulated based on the consumer price, not on the costs incurred by the
LEC’s new operations (For details about incentive regulations, see Loube, 1995;
Sappington & Weisman, 1996). In other words, telephone rates for “captive” rate
payers is largely irrespective of capital investment by LECs under the price cap
regulations. Thus, cross-subsidy would not cause harmful effects on service
rates of telephone senrices charged by telephone companies.

Moreover, the current regulatory regime for the telecommunications
industry do not seem to recognize the most critical feature of today's
telecommunications technology and economics. Models of industrial
organizations focus almost entirely on economies of scale as the traditional
source of efficiencies in most industries, yet there are other related factors that
may be important sources of cost savings as well for the multi-product firm.
Convergence in telecommunications may be translated to economies of scope,
which refers to cost savings that arise when a firm produces joint products as
part of the normal production process. Economies of scope imply that there may
be special cost savings that accrue to a multi-product firm that are unavailable to
the firms producing only one of the products but not both. In fact, the
requirements of a separate subsidiary for video service may block LECs ability to
achieve efficiencies resulting from economies of scope. Likewise, the ban of
cross-subsidization may have led to the foreclosure of potential entry into the

multichannel television industry by LECs.

257

List of References

References

Ahn, H. (1997 ) Vertical integration and consumer welfare' in the cable industry.
Journal/of Broadcasting & Electronic Media, 4 1(4), 453-477.

Ameritech Corp. v. United States. (1994). 867 F. Supp. 721.

Ameritech New Media. (1999). Comments of Ameritech New Media, Inc. In the
Matter of Annual Assessment of the Status of Competition in the
Marketplace for the Delivery of Video Programming (CS Docket No. 99-
230).

Applebaum, S. (1997, October 6). Cox's telephony push. Cablevision, 32, 34.

Applebaum, S. (1999, July 19). Chicago: Hotbed of competitive marketing.
Multichannel News.

Areeda, P., 8 Turner, D. F. (1975). Predatory pricing and related practices under
section 2 of the Sherman act. Harvard Law Review, 88, 697-733.

Aron, D. J., 8 Wildman, S. S. (1999, June). Efl'ecting a price squeeze through
bundled pricing. Paper presented at the NCTA annual conference,
Chicago, IL.

Bailey, E. E., 8 Friedlaender, A. F. (1982). Market structure and multiproduct
industries. Journal of Economic Literature, 20(3), 1024-1048.

Bain, J. (1956). Barriers to new competition. Cambridge: Harvard University
Press.

Baldwin, T. F. (1997, June 17). Convergence: Its meaning, scenarios, efi'ect on
industry. Paper presented at the 9th KISDI lntemational Conference:
Convergence of Telecommunications and Broadcasting, Seoul, Korea.

Baldwin, T. F., McVoy, D. S., 8 Steinfield, C. (1996). Convergence: Integrating
media, information and communication. Thousand Oaks: Sage.

Bar, F., Cohen, S., Cowhey, P., DeLong, B., Kleeman, M., 8 Zysman, J. (1999).
Defending the lntemet revolution in the broadband era: When doing
nothing is doing harm (E-conomy working paper 12): Berkeley Roundtable
on the lntemational Economy.

Aarrett, MGQ/V‘ol. Strategic behavior and competition in cable television. Journal
of Media Economics, 9(2), 43-62.

258

Barthold, J. (1999, September 20). Comcast, BA vie for high-speed biz clients.
Cable Worid.

Barthold, J. (1999b, May 3). Telcos use video to battle cable telephony. Cable
Worid.

Baumol, W., Panzar, J., 8 Willig, R. (1982). Contestable markets and the theory
of industry structure. New York: Harcourt Brace Jovanovich.

p... Beil, R. O., Dazzio, P. T., Ekelund, R. B., 8 Jackson, J. D. (1993). Competition
and the price of municipal cable television services: an empirical study.
Joumal of Regulatory Economics, 6, 401-415.

BellSouth Corp. (1999). Comments of BellSouth Corporation, Bel/South
Interactive Media Services, Inc., and BellSouth IMreless Cable Inc. In the
Matter of Annual Assessment of the Status of Competition in the
Marketplace for the Delivery of Video Programming (CS Docket No. 99-
230)

Bel/south Corp. v. United States. (1994). No. CV 93-B-2661-S (N.D. Ala. Sep.
23) .

Bork, R. H. (1978). The antitrust paradox: A policy at war with itself. New York:
Basic Books.

Breznick, A. (1999a). Cable Modem Growth Rolls Toward 1M Mark. Cable
Worid. April 19

Breznick, A. (1999b). Ameritech Keeps Plugging into Video. Cable World. June
7.

Breznick, A. (1999c). B.A. Video Rings Up Two More Markets. Cable World. May
31.

Breznick, A. (1999d). DBS inches toward long-awaited 10M mark. Cable Worid.
April 12.

Breznick, A. (1999c). DBS providers unveil plans to challenge cable. Cable
Worid. July 26.

Brown, K. (1999b, May 17). BellSouth Chooses Service Over Technology. Cable
World.

259

\/

/‘

/ \‘
/'_“\
Chan-Olmsted, S. M. (1?; Market competition for cable television:
Reexamining its orizontal mergers and industry concentration. Journal of
Media Economics, 9(2), 2542.

Chesapeake 8 Potomac Telephone Co. of Virginia v. United States. (1994). . 42
F. 3d. 181.

Chesapeake and Potomac Telephone Co. v. U. S. (1993). . 830 F. Supp. 909.

Chipty, T. (1995). Horizontal integration for bargaining power: evidence from the
cable television industry. Joumal of Economics 8 Management Strategy,
4(2), 375-397.

Cho, J. (1999c, March 8). Study says voice service cable's cash cow. Cable
World.

Cho, J. (1999d, April 26). Bells See Huge Revenue Gains From Data. Cable
Worid.

Ciciora, W., Farmer, J., 8 Large, D. (1999). Modern cable television technology:
Video, voice, and data communications. San Francisco: Morgan
Kaufmann.

Comanor, W. 6., 8 Mitchell, B. M. (1971). Cable television and the impact of
regulation. Bell Journal of Economics and Manpower Science, 154-212.

Comcast. (1999b). Comcast Corporation and Jones lntercable announce cable
system swaps with Adelphia communications (Vol. 1999, ).

CoreComm Ltd. (1999). Reply Comments of CoreComm Ltd. In the Matter of
Annual Assessment of the Status of Competition in the Marketplace for
the Delivery of Video Programming (CS Docket No. 99-230).

Dawson, F. (1998c, March 23). GTE plots huge ADSL rollout. Multichannel
News.

Dawson, F. (1999a, June 21). U S West's Broadband TV, Data Launches.
Multichannel News.

Dawson, F. (1999b, March 22). Cable can exploit e-commerce apps.
Multichannel News.

Demsetz, H. (1968). Why regulate utilities? Journal of Law and Economics, 11,
55-65.

260

 

\/

\/'

' ,z/

Dertouzos, J. N., 8 Wildman, S. S. (1998). Programming access and effective
competition in cable television (Attachment 2 to Comments of Ameritech
New Media, Inc. (In the matter of Implementation of Section 11(c) of the
Cable Television Consumer Protection and Competition Act of 1992) MM
Docket No. 92-264).

DIRECTV, I. (1999). Comments of DIRECTV, Inc. In the Matter of Annual
Assessment of the Status of Competition in the Marketplace for the
Delivery of Video Programming (CS Docket No. 99-230).

Dixit, A. (1979). A model of duopoly suggesting a theory of entry barriers. Bell
Journal of Economics, 10, 20-32.

Dixit, A. (1980). The role of investment in entry deterrence. Economic Journal,
90, 95-106. ,

Easley, D., Masson, R. T., 8 Reynolds, R. J. (1985). Preying for time. Joumal of
Industrial Economics, 33(4), 445-460.

Easterbrook, F. H. (1981 ). Predatory strategies and counterstrategies. University
of Chicago Law Review, 48(2), 263-337.

Egan, B. L. 696)? Information superhighways revisited: The economics of
multimedia. Boston: Artech House.

Ellis, L., 8 Dawson, F. (1998, March 2). After the hype, modems slow to roll.
Multichannel News.

Emmons, W. M., 8 Prager, R. A. (i992). The effects of market structure and
ownership on prices and service offerings in the US. cable television
industry. RAND Journal of Economics, 28(4), 732-750.

Estrella, J. (1998a). LyncStar applies for cable franchise. Multichannel News.
November 30.

Estrella, J. (1998b). MediaOne gets two Time Warner Mich. systems.
Multichannel News. June 15.

Estrella, J. (1998c). Mich. Outlaws AmeriChecks Promotion. Multichannel News.
January 5.

Estrella, J. (1998d). Mich. Ops beat telco on asset case. Multichannel News.
March 30.

Estrella, J. (1998e). Ameritech nabs Chicago area franchise. Multichannel News.
October 19.

261

 

Estrella, J., 8 Haugsted, L. (1999d). Competition thrives in Vlfindy City.
Multichannel News. June 14.

Estrella, J., 8 Heam, T. (1998). Robbins hopes for better from @Home.
Multichannel News. February 9.

Farrell, M. (1998a). BellSouth rolls out MMDS in Orlando. Multichannel News.
October 19.

Farrell, M. (1998b). BellSouth, GTE still in MMDS game. Multichannel News.
November 30.

Farrell, M. (1999a). New services power strong Cox results. Multichannel News.
May 3.

 

Farrell, M. (1999b). Telcos tout ADSL price plunge. Multichannel News. January
18.

Farrell, M. (1999c). AT&T wins battle for MediaOne. Multichannel News. May 10.

FCC. (1990). In the Matter of Competition, Rate Deregulation and the
Commission '3 Policies Relating to the Provision of Cable Television
Service (FCC 90-276 MM Dkt. No. 89-600).

FCC. (1993). In the Matter of Implementation of Sections of the Cable Television
Consumer Protection and Competition Act-Rate Regulation, Report and
Order and Further Notice of Proposed Rulemaking, (MM Docket No. 92-
266).

FCC. (1994a). In the Matter of Implementation of Sections of the Cable
Television Consumer Protection and Competition Act-Rate Regulation,
Buy- Through Prohibition, Third Report and Order (MM Dkt. No. 92-266
and MM Dkt. No. 92-262). Washington, DC.

FCC. (1994b). Second Order on Reconsideration, Fourth Report and Order, and
Fifth Notice of Proposed Rulemaking in the Matter of Implementation of
Sections of the Cable Television Consumer Protection and Competition
Act of 1992: Rate Regulation. FCC Record, 9, 4119-4310.

FCC. (1994c). Annual assessment of the status of competition in markets for the

delivery of video programming (CS Docket No. 94-48 First Report).
Washington, DC.

262

FCC. (1997). Annual assessment of the status of competition in markets for the
delivery of video programming (CS Docket No. 97-141 Fourth Annual
Report). Washington, DC.

FCC. (1998). Annual assessment of the status of competition in markets for the
delivery of video programming (CS Docket No. 98-102 Fifth Annual
Report). Washington, DC.

FCC. (1999a). In the matter of Inquiry Concerning the 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 of 1996 (CC
Docket No. 98-146 FCC 99-5).

FCC. (1999b). In the matter of section 3 of the Cable Television Consumer
Protection and Competition Act of 1992: Statistical report on average
rates for basic service, cable programming services, and equipment
(Report on cable industry prices MM Docket No. 92-266). Washington,
DC.

FCC. 99c In the matter of applications for consent to the transfer of control
of licenses and section 214 authorizations from TeIe-Communications,
Inc. to AT&T Corp. (CC Docket No. 98-178 FCC 99-24).

FCC. 2000 . Annual assessment of the status of competition in markets for the
delivery of video programming (CS Docket No. 99-230 Sixth Annual
Report). Washington, DC.

I Ford, G. S., 8 Jackson, J. D. ((1997). Horizontal concentration and vertical
integration in the cable vision industry. Review of Industrial
Organization, 12, 501-518.

Gabel, D., 8 Rosenbaum, D. I. (1995). Prices, costs, extemalities and
entrepreneurial capital: lessons from Wisconsin. Antitnrst Bulletin, 40(3),
581-608.

General Accounting Office. (1999a). Telecommunications: Impact of Sports
Programming Costs on Cable Television Rates (GAO/RCED-99-136).
Washington, DC: US. General Accounting Office.

General Accounting Office. (1999b). Telecommunications: The Changing Status
of Competition to Cable Television (GAO/RCED-99-158). Washington,
DC: US. General Accounting Office.

Gibbons, K. (1997). BellSouth hits big easy. Multichannel News. November 24.
1, 70.

263

Gibbons, K. (1998a). Fast-growing Knology 'scares' cable. Multichannel News.
June 22.

Gibbons, K. (1998b). Insight buys competitive Columbus system. Multichannel
News. June 29.

Gibbons, K. (1998c). AT&T makes local phone play with $48B merger.
Multichannel News. June 29.

Glick, E. (1999a). Satellite bill passes. Cable World. November 22.

Glick, E. (1999b). FCC seeks balance on ownership rules cap. Cable World.
October 11.

Gooley, K. J. (1996). Telecommunications: wireline. Standard 8 Poors Industry
Surveys, September 1. 21-34.

Grebb, M. (1999). Knology knows Southeast overbuilding. Multichannel News.
July 19.

Haugsted, L. (1998). Cable fails to stop BellSouth in Miami. Multichannel News.
February 23.

Haugsted, L. (1999a). PrimeOne’s Tele-TV challenges L.A. ops. Multichannel
News, June 21. 46.

Haugsted, L. (1999b). Billing advances aid in bundling. Multichannel News. July
12.

Haugsted, L. (1999c). State raps Boston Edison on cable deal. Multichannel
News. June 7.

Hazlett, T. W. (1986). Competition vs. franchise monopoly in cable television.
Contemporary Policy Issues, 80-97.

Hazlett, T. W. (1987). The policy of exclusive franchising in cable television.
Journal of Broadcasting 8 Electronic Media, 31(1), 1-20.

Hazlett, T. W. (1990). Duopolistic competition in cable television: Implications for
public policy. Yale Joumal on Regulation, 65(7), 65-119.

Hazlett, T. W. (1991). The demand to regulate franchise monopoly: evidence

from CATV rate deregulation in California. Economic Inquiry, 29(2), 275-
296.

264

Hazlett, T. W. (1995a). Predation in local cable TV markets. Antitrust Bulletin,
40(3), 609-644.

Hazlett, T. W. (1995b). Telco entry into video. Annual Review of
Communications 1994-1995, 48, 212-228.

Hazlett, T. W. (1996). Cable television rate deregulation. Intemation Journal of
the Economics of Business, 3(2), 145-163.

Hazlett, T. W. (1997). Prices and outputs under cable TV regulation. Journal of
Regulatory Economics, 12, 173-195.

Hazlett, T. W., 8 Spitzer, M. L. (1997). Public policy toward cable television: The
economics of rate controls. Cambridge, MA: The MIT Press.

Heam, T. (1998a). FCC OKs RCN in Philly, S.F. Multichannel News. June 22.

Heam, T. (1998b). Ameritech lobbies on program-access rules. Multichannel
News. February 9.

Heam, T. (1998c). Cable presses FCC on ownership rules. Multichannel News.
August 17.

Heam, T. (1999). B'casters want no limits on DBS retrans. Multichannel News.
July 5.

Hogan, M. (1997, December 8). TCI, Ameritech battle in Arlington Heights. ‘
Multichannel News, 69.

Hogan, M. (1998b). GTE steps up marketing efforts in Hawaii. Multichannel
News. July 20.

Hogan, M. (1998c). BellSouth brings digital cable to Atlanta. Multichannel News.
June 8.

Hogan, M. (1998d). RCN makes advances on Queens. Multichannel News. May
4.

Hogan, M. (1998s). Cox Rings Up Two More Phone Markets. Multichannel
News. September 14.

Hogan, M. (1998f). Bundle of opportunity: Cable operators step up pace of
telephony. Multichannel News. March 23.

Hogan, M. (19989). PacBell faces video birthday with uncertainty. Multichannel
News. April 27

265

Hogan, M. (1998h). PrimeOne buy 3 PacBell Video. Multichannel News. October

Hogan, M. (1999). MediaOne rings up motor city. Multichannel News, April 26.
10.

Jayaratne, J. (1996). A note on the implementation of cable TV rate caps.
Review of Industrial Organization, 1 1, 823-840.

‘/Johnson, L. L. (1994). Toward competition in cable television. Cambridge, MA:
MIT Press.

Johnson, L. L., 8 Reed, D. P. (1992). Telephone company entry into cable
television: an evaluation. Telecommunications Policy, 16, 122-134.

Johnson, R. C., 8 Blau, R. T. (1974). Single versus multiple-system cable
television. Joumal of Broadcasting, 18, 323-346.

Jones, T. (1999, November 16). SBC halts Ameritech cable growth plans.
Chicago Tribune.

Kagan Media Inc. (1997). TV programming costs: An analysis of the market
forces driving entertainment and sports rights fees : Kagan Media
Appraisals, Inc.

Katz, M. L. (1996). Remarks on the economic implications of convergence.
Industrial and Corporate Change, 5(4), 1079-1095.

Kinetic Strategies. (1999). Overview of cable modern technology and services.
Cable Datacom News.

Kinetic Strategies. (1999b). Cable IP telephony primer. Cable Datacom News.
Kreps, D. M., Milgrom, P., Roberts, J., 8 Wilson, R. (1982). Rational cooperation

in the finitely repeated prisoners' dilemma. Journal of Economic Theory,
27, 245-252.

Kreps, D. M., 8 Vlfilson, R. (1982). Reputation and imperfect information. Journal
of Economic Theory, 27, 253-279.

Kuhl, C. A. (1998, November 30). Telcos warily eye IP Fremont trial.
Multichannel News.

Lathen, D. A. (1999). Broadband Today (A staff report): FCC Cable Services
Bureau.

266

 

Lee, W. (1983). Cable franchising and the first amendment. Vanderbilt Law
Review, 36, 867-928.

Levin, S. L., 8 Meisel, J. B. (1991). Cable television and competition: theory,
evidence, and policy. Telecommunications Policy, 15, 519-528.

Levine, S. (1997a, May 26). Anti-competitive or competitive advantage? Policy 8
Regulation.

Levine, S. (1997b, April 14). Fiber to the mansion?:Scott3dale gets prickly over U
S West‘s FTTC plans. Telephony.

Lieberman, M. B. (1987). Excess capacity as a barrier to entry: An empirical
appraisal. Journal of Industrial Economics, 35(4), 607-627.

Lindenberg, E., 8 Ross, S. (1981). Tobin's q ratio and industrial organization.
Joumal of Business, 54, 1-32.

Litman, B. R. (1998). The motion picture mega-industry. Boston: Allyn and
Bacon.

Loube, R. (1995). Price Cap Regulation: Problems and Solutions. Land
Economics, 71(3), 286-98.

Maggi, L. (1998, August 10). Cable competitors grouse about state of
competition. States News Service.

Markus, M. L. (1990). Toward a "Critical Mass" theory of interactive media. In J.
F ulk 8 C. Steinfield (Eds), Organizations and communication technology
(pp. 194-218). Newbury Park: Sage Publications.

Lev/Mayo, J. W., 8 Otsuka, Y. (1991). Demand, pricing, and regulation: evidence

from the cable TV industry. RAND Journal of Economics, 22(3), 396-410.

McFarland, H. (1988). Evaluating q as an alternative to the rate of return in
measuring profitability. Review of economics and statistics, 70(4), 614-
622.

McGee, J. S. (1958). Predatory price cutting: the Standard Oil (N.J.) case.
Joumal of Law and Economics, 1, 137-169.

McGee, J. S. (1980). Predatory pricing revisited. Journal of Law and Economics,
23(2), 289-330.

Menezes, B. (1999). BellSouth offering Atlanta video over fiber. Multichannel
News. June 7.

267

JMerline, J. W. (1990). How to get better cable TV at lower prices. Consumers'
Research, 73(5),10-16.

Milgrom, P., 8 Roberts, J. (1982a). Predation, reputation, and entry deterrence.
Joumal of Economic Theory, 27, 280-312.

\/ Miller, I. R. (1997). Models for determining the economic value of cable television
systems. Joumal of Media Economics, 10(2), 21-33.

Mueller, D. C. (1997). First-mover advantages and path dependence.
lntemational Journal of Industrial Organization, 15, 827-850.

Multichannel News. (1989, August 22). Telesat sells out to Palmer in two Florida
counties. Multichannel News.

Multichannel News. (1997, November 24). SBC, Americast spar over pullout.
Multichannel News. 70.

Multichannel News. (1998, April 17). TCI Antes Up $500M for Jones Subs.
Multichannel News.

Multichannel News. (1998, August 17). MediaOne Seeks Looser Ownership
Rules. Multichannel News.

Multichannel News. (1999, June 7). RCN files complaint against Cablevision.
Multichannel News.

Multichannel News. (1999, July 26). RCN launches in DC. suburb. Multichannel
News.

\/ NCTA. (1999). Cable television development. Vol. 22.

Neel, K. C. (1996). Viacom Inc-TCI Mega-Deal Finally Wins an OK From IRS.
Cable World. June 24.

Neel, K. C. (1997). Cable Deal Market Rebounded in 1996. Cable World. April
21.

Neel, K. C. (1999a). Bell Atlantic Boxes Gather Dust at Adelphia. Cable World.
January 25.

Neel, K. C. (1999b). Shifting DBS Landscape Concerns Industry. Cable World.
May 17.

Neel, K. C. (1999c). Armstrong Cable's Master Dealmaker. Cable World. May 10.

268

\// Noam, E. M. (1985). Economies of scale in cable television: A multiproduct
analysis. In E. M. Noam (Ed.), Video media competition: Regulation,
Economics and Technology (pp. 93-120). New York: Columbia University
Press.

Noll, R. G., Peck, M. J., 8 McGowan, J. J. (1973). Economic aspects of television
regulation. Washington, DC: The Brookings Institute.

NTIA. (1988). Video program distribution and cable television: cunent policy
issues and recommendations (NTIA Report 88-233). Washington DC:
US. Department of Commerce.

NTIA. (1991). Telecommunications in the age of information (The NTIA
infrastructure report NTIA Special Publication 91-26). Washington, DC:
National Telecommunications and Information Administration.

NYNEX Corp. v. United States. (1994). Civil No. 93-323-P-C (D. Me. Dec. 8). .

OECD. (1992). Telecommunications and Broadcasting: Convergence or
Collision? (Vol. 29). Paris.

Ordover, J. A. (1998). Predatory pricing. In J. Eatwell, M. Milgate, 8 P. Newman
(Eds), The New Palgrave (Vol. 3, pp. 77-84). New York: Stockton Press.

Ordover, J. A., 8 Saloner, G. (1998). Predation, monopolization and antitrust. In
R. Schmalensee 8 R. D. Willg (Eds), Handbook of industrial organization
(Vol. 1, pp. 537-596). Amsterdam: Elsevier Science.

Otsuka, Y. (1997). A welfare analysis of local franchise and other types of
regulation: evidence from the cable TV industry. Journal of Regulatory
Economics, 11, 157-180.

Owen, B. M., 8 Gottlieb, P. D. (1986). The rise and fall and rise of cable
television regulation. In L. W. Weiss 8 M. W. Klass (Eds), Regulatory
refonn: What actually happened (pp. 78-104). Boston: Little, Brown, and
Company.

Owen, B. M., 8 Greenhalgh, P. R. (1986). Competitive considerations in cable
television franchising. Contemporary Policy Issues, 4(2), 69-79.

Owen, B. M., 8 Wildman, S. S. (1992). Video economics. Cambridge, MA:
Harvard University Press.

Paikert, C. (1998). MediaOne takes to phones in Atlanta. Multichannel News.
February 2.

269

Panzar, J. C., 8 Willig, R. D. (1981). Economies of scope. American Economic
Review, 71 (2), 268-72.

Park, R. E. (1972). Prospects for cable in the 100 largest television markets. Bell
Joumal of Economics, 130-150.

Peltzman, S. (1976). Towards a more general theory of regulation. Journal of
Law and Economics, 211-240.

Phlips, L. (1995). Competition policy: A game-theoretic perspective. Cambridge:
Cambridge University Press.

Posner, R. A. (1972). The appropriate scope of regulation in the cable television
industry. Bell Journal of Economics and Management Science, 3(1), 98-
129.

Posner, R. A. (1979). The Chicago school of antitrust analysis. University of
Pennsylvania Law Review, 127, 925-948.

Prager, R. A. (1990). Firm behavior in franchise monopoly markets. RAND
Journal of Economics, 21(2), 211-225.

Quinton, B. (1999a). Americhecks not in the mail, says Ohio high court.
Multichannel News. July 16.

Quinton, B. (1999b). Ameritech buys up data integration. Telephony. March 1.
10.

Rasmusen, E. (1989). Games and information: An introduction to game theory.
New York: Basil Blackwell.

RCN Corp. (1999). Comments of RCN Corp. In the Matter of Annual Assessment
of the Status of Competition in the Marketplace for the Delivery of Video
Programming (CS Docket No. 99-230).

Rosenthal, R. W. (1981). Games of perfect information, predatory pricing and
the Chain-Store Paradox. Journal of Economic Theory, 25, 92-100.

Rubinovitz, R. N. (1993). Market power and price increases for basic cable
service since deregulation. RAND Journal of Economics, 24(1), 1-18.

Sappington, D. E., 8 Weisman, D. L. (1996). Designing Incentive Regulation for
the Telecommunications Industry. Cambridge, MA: The MIT Press.

Selten, R. (1978). The chain store paradox. Theory and decision, 9, 127-159.

270

Sherer, F. M., 8 Ross, D. (1990). Industrial market structure and economic
performance. (3rd ed.). Boston: Houghton Miffin.

Shooshan, H. M. (1989). Cable television: promoting a competitive industry
structure. In P. R. Newberg (Ed.), New directions in telecommunications
policy (Vol. 1, pp. 222-246). Durham, NC: Duke University Press.

Smiley, A. K. (1986). Direct competition among cable television systems
(Economic Analysis Group #86-9). Washington, DC: US. Department of
Justice.

Smiley, A. K. (1990). Regulation and competition in cable television. Yale Journal
on Regulation, 7, 121-139.

Smiley, R. H., 8 Ravid, S. A. (1983). The importance of being first: Ieaming price
and strategy. Quarterly Journal of Economics, 98(2), 353-362.

Smirlock, M., Gilligan, T., 8 Marshal, W. (1984). Tobin's q and the structure-
perfonnance relationship. American Economic Review, 74(5), 1051-60.

SNET. (1997). Choice in cable TV comes to Connecticut .
Solomon, D. (1997). Cable deals bloom. Detroit Free Press, July 1. pp. 1A-2A.

Spence, A. M. (1977). Entry, capacity, investment and oligopolistic pricing. Bell
Journal of Economics, 8, 534-544.

Stigler, G. J. (1971 ). The theory of economic regulation. Bell Joumal of
Economics and Management Science, 3-21.

Tacoma Public Utilities. (1997). Telecommunications study. City of Tacoma,
WA.

Teece, D. J., Pisano, G., 8 Shuen, A. (1997). Dynamic capabilities and strategic
management. Strategic Management Joumal, 18(7), 509-533.

Telser, L. G. (1965). Abusive trade practices: An economic analysis. Law and
Contemporary Problems, 30, 488, 494-496.

Telser, L. G. (1966). Cutthroat competition and the long purse. Journal of Law
and Economics, 9, 259-265.

Thompson, R. W. (1989, July 24). Telesat pickets competitor over pricing.
Multichannel News, 47.

271

Thorpe, K. (1985). The impact of competing technologies on cable television. In
E. M. Noam (Ed.), Video media competition: Regulation, Economics and
Technology (pp. 135-168). New York: Columbia University Press.

U. S. West Inc. v. United States. (1994). No. 94-35775, DC. No. CV-93-01523-
BJR.

Umstead, R. T. (1998). Ops eye low-cost local heroes. Multichannel News. May
4.

Vittore, V. (1999). The price of bundling. Telephony. May 36.

Wenders, J. T. (1971). Excess capacity as a barrier to entry. Journal of Industrial
Economics, 20(1), 14-19.

Williamson, O. E. (1976). Franchise bidding for natural monopolies—in general
and with respect to CATV. Bell Journal of Economics, 7(1), 73-104.

Yamey, B. S. (1972). Predatory price cutting: Notes and comments. Journal of
Law and Economics, 15(1), 129-142.

Zupan, M. A. (1989a). Non-price concessions and the effect of franchise bidding
schemes on cable company costs. Applied Economics, 21, 305-323.

Zupan, M. A. (1989b). The efficacy of franchise bidding schemes in the case of

cable television: Some systematic evidence. Joumal of Law 8 Economics,
32, 401 -456.

272

MICHIGAN STRTE UNIV. LIBRARIES
III III II llllll III III III I III III III III Illlll
31293020486266