Telecommunications: Direct Broadcast Satellite Subscribership Has
Grown Rapidly, but Varies across Different Types of Markets	 
(06-APR-05, GAO-05-257).					 
                                                                 
Since its introduction in 1994, direct broadcast satellite (DBS) 
service has grown dramatically, and this service is now the	 
principal competitor to cable television service. Although DBS	 
service has traditionally been a rural service, passage of the	 
Satellite Home Viewer Improvement Act of 1999 enhanced the	 
competitiveness of DBS service in suburban and urban markets. GAO
agreed to examine (1) how DBS subscribership changed since 2001; 
(2) how DBS penetration rates differ across urban, suburban, and 
rural areas; (3) how DBS penetration rates differ across markets 
based on the degree and type of competition provided by cable	 
operators; and (4) the factors that appear to influence DBS	 
penetration rates across cable franchise areas. To complete this 
report, GAO prepared descriptive statistics and an econometric	 
model using data from the Federal Communications Commission's	 
annual Cable Price Survey and the Satellite Broadcasting and	 
Communications Association's subscriber count database. 	 
-------------------------Indexing Terms------------------------- 
REPORTNUM:   GAO-05-257 					        
    ACCNO:   A20877						        
  TITLE:     Telecommunications: Direct Broadcast Satellite	      
Subscribership Has Grown Rapidly, but Varies across Different	 
Types of Markets						 
     DATE:   04/06/2005 
  SUBJECT:   Cable television					 
	     Communication satellites				 
	     Competition					 
	     Data collection					 
	     Econometric modeling				 
	     Economic analysis					 
	     Economic growth					 
	     Surveys						 
	     Telecommunication industry 			 
	     Comparative analysis				 
	     Television broadcasting				 

******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO Product.                                                 **
**                                                              **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced.  Tables are included, but    **
** may not resemble those in the printed version.               **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
******************************************************************
GAO-05-257

United States Government Accountability Office

GAO	Report to the Subcommittee on Antitrust, Competition Policy and
Consumer

                Rights, Committee on the Judiciary, U.S. Senate

April 2005

TELECOMMUNICATIONS

 Direct Broadcast Satellite Subscribership Has Grown Rapidly, but Varies across
                           Different Types of Markets

                                       a

GAO-05-257

[IMG]

April 2005

TELECOMMUNICATIONS

Direct Broadcast Satellite Subscribership Has Grown Rapidly, but Varies across
Different Types of Markets

                                 What GAO Found

Since 2001, the number of households subscribing to DBS service has grown
rapidly; thus the percentage of households subscribing to DBS service, the
DBS penetration rate, has grown to over 17 percent of American households.

Aggregate DBS Subscription and DBS Penetration Rates, 2001-2004

25 DBS subscribers DBS penetration rate 25

(in millions) (percent) Subscribers

20 20

Penetration rate

15 15

10 10

55

00 2001 2002 2003 2004 Year

Source: GAO.

The DBS penetration rate is highest in rural areas, but growing most
rapidly in suburban and urban areas. Between 2001 and 2004, the DBS
penetration rate grew 15 percent in rural areas to 29 percent of rural
households, 32 percent in suburban areas to 18 percent of suburban
households, and 50 percent in urban areas to 13 percent of urban
households.

The degree and type of competition influences the DBS penetration rate. In
areas with no cable service, the DBS penetration rate is about 53
percentage points greater than in areas where cable service is available.
Where cable service is available, cable operators increasingly offer
advanced services. The DBS penetration rate is approximately 20 percentage
points greater in areas where cable operators are not providing advanced
services, compared with areas where these services are available. While
relatively few areas have more than one wire-based cable operator, in
these areas the DBS penetration rate is 8 percentage points lower than in
areas with only one cable operator.

In addition to the differences in DBS penetration rates across rural,
suburban, and urban areas, and differences associated with the degree and
type of cable competition, additional geographic and competitive factors
also influence the DBS penetration rate. For example, the DBS penetration
rate is lower in areas with a high prevalence of multiple-dwelling units,
such as apartments. Additionally, the DBS penetration rate is higher in
areas where DBS providers offer local broadcast stations (such as ABC and
NBC affiliates) directly to their subscribers.

The Federal Communications Commission provided technical comments on

a draft of this report that we incorporated where appropriate.

                 United States Government Accountability Office

Contents

  Letter

Results in Brief
Background
Subscription to DBS Has Grown Rapidly Since 2001
DBS Has the Greatest Penetration in Rural Areas, but Subscriber

Growth Has Been Greater in Urban and Suburban Areas Since 2001

DBS Penetration Is Higher Where Cable Service Is Not Available, Where
Cable Providers Do Not Offer Advanced Services, and Where Wire-Based
Competitors Are Not Present

Several Geographic and Competitive Factors Are Associated with Different
Levels of DBS Penetration across Cable Franchise Areas

Agency Comments and Our Evaluation Industry Participants' Comments

1 3 5 6

7

9

13 16 16

Appendixes

        Appendix I: Scope and Methodology 18 Appendix II: Data Reliability 20
                           Appendix III: Cable-Satellite Econometric Model 22

Tables Table 1:   Explanatory Variables Used in Cable-Satellite Model   20 
          Table 2:      Descriptive Statistics for Variables Used in       
                                       Cable-Satellite                     
                                            Model                          25 
          Table 3:         Three-Stage Least Squares Model Results         27 
          Table 4:            Alternative Specification Results            30 

Figures	Figure 1: Figure 2: Figure 3: Figure 4: Figure 5:

Aggregate DBS Subscription and DBS Penetration Rates,
2001-2004 6
DBS Penetration Rates in Urban, Suburban, and Rural
Areas, 2001-2004 7
Growth in DBS Subscribers and DBS Penetration Rates,
2001-2004 8
Percentage of Households Using Different Modes of
Television Reception, 2004 10
Percentage of Cable Operators Providing Advanced
Services (Digital Cable, Cable Modem, and Telephone),
2001-2004 11

Contents

Figure 6:	DBS Penetration Rate and Cable Operators' Provision of Advanced
Services (Digital Cable, Cable Modem, and Telephone), 2001-2004 12

Figure 7:	DBS Penetration Rate in Cable Franchise Areas with and without
Wire-Based Cable Competition, 2004 13

Abbreviations

2SLS two stage least squares
3SLS three stage least squares
BLS Bureau of Labor Statistics
CUID Community Unit Identification
DBS direct broadcast satellite
DMA designed market area
FCC Federal Communications Commission
MSA metropolitan statistical area
MSO multiple system operator
SBCA Satellite Broadcasting and Communications Association

This is a work of the U.S. government and is not subject to copyright
protection in the United States. It may be reproduced and distributed in
its entirety without further permission from GAO. However, because this
work may contain copyrighted images or other material, permission from the
copyright holder may be necessary if you wish to reproduce this material
separately.

A

United States Government Accountability Office Washington, D.C. 20548

April 6, 2005

The Honorable Mike DeWine

Chairman

The Honorable Herb Kohl

Ranking Minority Member

Subcommittee on Antitrust, Competition Policy and Consumer Rights
Committee on the Judiciary United States Senate

For many years, cable television operators faced little competition in the
market for subscription video service. However, in 1994, a new type of
competitor emerged: direct broadcast satellite (DBS). Subscribers to DBS
service use small reception dishes to receive television programming
beamed down from satellites that orbit over the equator. DBS was
originally most popular in rural areas, where cable service was often
limited or did not exist. In recent years, DBS has also become popular in
suburban and urban markets. In particular, the competitiveness of DBS was
bolstered when the Congress enacted the Satellite Home Viewer Improvement
Act of 1999, which allows DBS carriers to provide local broadcast signals,
such as the affiliates of ABC and NBC, directly to subscribers.1 Today,
the two primary DBS providers-DIRECTV(R) and EchoStar-provide local
broadcast signals in 156 of 210 television markets.

You asked us to provide information on the extent to which DBS is
competitive with cable under varied market circumstances. Specifically,
this report provides information on (1) how DBS subscribership has changed
since 2001; (2) how DBS penetration rates (that is, the percentage of
households subscribing to DBS) differ across urban, suburban, and rural
areas; (3) how DBS penetration rates differ across markets based on the
degree and type of competition provided by cable operators; and (4) the

1In late 2004, the Congress passed and the President signed the Satellite
Home Viewer Extension and Reauthorization Act of 2004. This act may
further enhance the competitiveness of DBS providers by, for example,
permitting DBS providers to deliver broadcast signals that are
"significantly viewed" within a local market but originate from outside
that local market.

factors that appear to influence DBS penetration rates across cable
franchise areas.2

To respond to the objectives of this report, we gathered data on DBS
subscribers from the Satellite Broadcasting and Communications Association
(SBCA)3 and cable rates and services from the Federal Communications
Commission (FCC). For the first and second objectives, we calculated the
DBS penetration rate at the county level for the years 2001 to 2004. This
allowed us to examine the trend in the DBS penetration rate for that
period of time. We also classified counties as urban, suburban, and rural,
based on the location of central cities and designations of metropolitan
statistical areas (MSA), and calculated the DBS penetration rate for each
of these geographic categories. For the third objective, we used data from
Knowledge Network's 2004 The Home Technology Monitor4 survey to identify
responding households that said cable service was not available to them,
and to then examine the penetration of DBS service among these responding
households, compared with those who said that cable service was available
in their area. We also calculated the DBS penetration rate for cable
franchise areas included in FCC's 2002 and 2004 Cable Price surveys.5
Using data from FCC's surveys, we classified each cable franchise on the
basis of (1) whether "advanced services"-such as cable modem and digital
cable tiers of programming-were provided by cable operator and (2) whether
there was a second wire-based provider in the market. We then calculated
the DBS penetration rate for these different categories of cable franchise
areas. For the fourth objective, we used an econometric model we
previously developed that examines the competitive interaction of cable
and DBS providers. Using data from 2004, the model considers the effect of
various factors (such as the number of channels provided by the cable
company) on cable rates, the number of cable subscribers, the number of
channels that cable operators provide to

2At the community level, cable operators obtain a franchise agreement
under agreed-upon terms and conditions from a franchising authority, such
as a city, township, or county. The franchise agreement permits the cable
operator to provide service in the jurisdiction.

3SBCA is a national trade association representing the satellite industry.
Its members include DBS, C-band, satellite radio, and other satellite
service providers, among others.

4Knowledge Networks is a survey research firm that had conducted a
consumer survey on household television characteristics. The survey
provided the responses of 2,471 randomly selected American households.

5FCC conducts an annual survey of a random sample of cable franchise areas
to gather information about cable pricing and other related issues.

subscribers, and the DBS penetration rate in areas throughout the United
States.6 See appendix I for additional information on our scope and
methodology and appendix II for the steps we took to ensure the
reliability of the data we used to prepare this report. We found these
data to be sufficiently reliable for the purposes of this report.

We conducted our review from March 2004 to February 2005 in accordance
with generally accepted government auditing standards.

Results in Brief	Subscriptions to DBS service have grown rapidly since
2001. In July 2001, about 15.5 million American households-about 13
percent of households-subscribed to a DBS service. By January 2004, about
21.3 million households subscribed to a DBS service, or 17.4 percent of
households. Thus, the number of DBS subscribers increased by 37.8 percent
during this 2-1/2 year time frame.7

DBS penetration rates have been and remain highest in rural areas, but
since 2001, DBS penetration has grown most rapidly in urban and suburban
areas, where the penetration rates were originally low. In 2001, DBS
penetration rates were nearly 26 percent in rural areas, 14 percent in
suburban areas, and about 9 percent in urban areas. By 2004, DBS
penetration rates had increased to about 29 percent in rural areas, 18
percent in suburban areas, and 13 percent in urban areas, indicating a
consistent pattern of higher DBS penetration rates in rural areas. In
short, over the 2001 to 2004 time frame, the DBS penetration rate grew
about 50 percent and 32 percent in urban and suburban areas, respectively,
compared with a growth rate of 15 percent in rural areas.

DBS penetration rates are affected by the degree and type of competition
in a local market. Based on survey data, we found that relatively few

6This four equation model is an adaptation of a model that GAO developed
and discussed in three previous reports. See GAO, Telecommunications: The
Effect of Competition from Satellite Providers on Cable Rates, RCED-00-164
(Washington, D.C.: July 18, 2000); Telecommunications: Issues in Providing
Cable and Satellite Television Services, GAO-03130 (Washington, D.C.: Oct.
15, 2002); and Telecommunications: Issues Related to Competition and
Subscriber Rates in the Cable Television Industry, GAO-04-8 (Washington,
D.C.: Oct. 24, 2003).

7The data cover 2-1/2 years, rather than 3 years, because FCC changed the
time frame of its survey from July to January in 2004.

American households-less than 9 percent-do not have the opportunity to
purchase cable television service because it is not available where they
live.8 However, in these areas, the DBS penetration rate is about 53
percentage points greater than it is in areas where cable television
service is available. In areas where cable television service is
available, cable operators are increasingly providing advanced services,
such as digital cable, cable modem, and telephone service. In 2004, the
DBS penetration rate was approximately 20 percentage points greater in
areas where cable operators were not providing advanced services, compared
with areas where these services were available. Finally, in most areas,
cable companies do not compete with other wire-based competitors, but in
limited areas there is more than one wire-based provider of cable
television service. Where more than one cable provider exists, the DBS
penetration rate is 8 percentage points lower than in areas with only one
cable provider.

Using an econometric model to control for the many factors that influence
the DBS penetration rate, we identified three key geographic factors and
three key competitive factors that influence the DBS penetration rate.
Some of these findings confirm the findings discussed above based on
descriptive statistics, and other findings could only be examined within
the model. Regarding the geographic factors, we found that (1) the DBS
penetration rate is lower in markets with a high prevalence of multiple
dwelling units, such as apartments and condominiums; (2) the DBS
penetration rate is lower in areas where, in order to face the
transmitting satellite, the satellite dish must be installed at a
relatively low angle, facing the horizon more than the sky; and (3) the
DBS penetration rate is higher in nonmetropolitan areas. Regarding the
competitive factors, we found that (1) the DBS penetration rate is lower
in areas where the cable operator's system has greater system capacity;9
(2) the DBS penetration rate is lower in areas where there is more than
one wire-based cable provider; and (3) the DBS penetration rate is higher
in areas where DBS providers carry local broadcast stations, such as an
ABC affiliate.

8The percentage of households with cable service available has been a
subject of controversy. Different industry participants, using different
data sources and different bases of comparison, arrive at different
figures. Using industry data, FCC noted that the percentage of homes with
a television passed by cable must be less than 97.8 percent.
Alternatively, the National Rural Telecommunications Cooperative reports
that 22.4 million households (or about 20 percent) lack access to cable
service. We used a survey of a random sample of households to arrive at
the 9 percent figure.

9System capacity is measured in terms of the system megahertz. Systems
with larger capacity are able to provide more channels and other services
to their subscribers.

We provided a draft of this report to the Federal Communications
Commission (FCC) for their review and comment. FCC staff provided
technical comments that we incorporated where appropriate.

Background	Cable television service emerged in the late 1940s to fill a
need for television service in areas with poor over-the-air reception,
such as mountainous or remote areas. At that time, cable operators simply
retransmitted the signals of local broadcast stations. By the late 1970s,
cable operators began to provide new cable networks,10 such as HBO,
Showtime, and ESPN, and the number of cable subscribers increased rapidly.
Two significant changes occurred in the 1990s and early 2000s. First, the
Congress passed the Cable Television Consumer Protection and Competition
Act of 1992 that, among other things, prohibited local franchising
authorities from awarding exclusive (or monopoly) franchises to cable
operators, thereby opening the door to wire-based competition. Second,
cable operators began offering new services, such as digital cable, cable
modem Internet access, and telephone, in addition to their basic video
service. Today, many cable operators offer these advanced services in
bundles with their basic video service.

Since its introduction in 1994, direct broadcast satellite (DBS) service
has grown dramatically and is now the primary competitor to cable
operators. Subscribers to DBS service use a small reception dish to
receive signals beamed down from satellites. Because DBS satellites orbit
above the equator, a reception dish must point toward the southern sky,
and households located in the northern part of the United States need to
angle the dish more toward the horizon than households in the southern
part of the United States. Unlike cable, which upgraded to digital service
in recent years, DBS service has been a digital-based service since its
inception. DBS providers generally offer most of the same cable networks
as cable operators. However, for many years DBS providers did not offer
local broadcast stations to their subscribers in most instances because of
copyright obstacles, obstacles that cable operators did not face. After
the Congress passed the Satellite Home Viewer Improvement Act of 1999,
which altered the copyright rules that applied to DBS providers, cable and
DBS companies were placed on a more equal competitive footing.

10Today, a variety of subscription video providers, in addition to cable
operators, deliver these networks to their subscribers. For consistency,
we refer to these networks as cable networks.

  Subscription to DBS Has Grown Rapidly Since 2001

From 2001 to 2004, the aggregate number of U.S. households that subscribe
to DBS television service grew rapidly. Figure 1 illustrates the growth in
total DBS subscription and penetration rates for 2001 through 2004. In
July 2001, about 15.5 million households were served by DBS. By January
2004, about 21.3 million households were served by DBS-an increase of 37.8
percent in 2-1/2 years. Similarly, over the same period of time, the
overall penetration rate of DBS rose from 13 percent in 2001 to 17.4
percent in 2004-a 33.5 percent increase.

Figure 1: Aggregate DBS Subscription and DBS Penetration Rates, 2001-2004
DBS subscribers (in millions) DBS penetration rate (percent)

25

20

15

10

5

0 2001 2002 2003 2004 Year

Subscribers Penetration rate Source: GAO.

                                       25

                                       20

                                       15

                                       10

                                       5

                                       0

  DBS Has the Greatest Penetration in Rural Areas, but Subscriber Growth Has
  Been Greater in Urban and Suburban Areas Since 2001

DBS penetration rates have been higher in rural areas than in suburban and
urban areas throughout the last several years, as shown in figure 2. From
July 2001 to January 2004, DBS penetration has grown steadily in all three
types of geographic areas. In 2001, penetration rates were highest in
rural areas at 25.6 percent, followed by 13.9 percent in suburban areas
and 8.6 percent in urban areas. As of January 2004, DBS penetration
remained the highest in rural areas, growing to about 29 percent, while it
grew to 18 percent of suburban households and 13 percent of urban
households.

Figure 2: DBS Penetration Rates in Urban, Suburban, and Rural Areas,
2001-2004

DBS penetration rate (percent) 30

25

20

15

10

5

0

2001 Year

Urban Suburban Rural Source: GAO.

2002 2003 2004

Although the DBS penetration rate in rural areas has been and remains
higher than it is in other geographic areas, subscribership has grown more
rapidly in suburban and urban areas than in rural areas from 2001 to 2004.
In fact, urban areas have experienced the highest growth in overall DBS
subscribership. Figure 3 displays the percentage growth in total DBS
subscribers and the percentage growth in DBS penetration rates in urban,

suburban, and rural areas. From 2001 to 2004, DBS subscribership grew 55
percent in urban areas, 37 percent in suburban areas, and 17 percent in
rural areas. In the same time period, the growth in penetration rates was
also highest in urban areas, at 50.4 percent, followed by suburban
penetration growth at 32 percent, and rural penetration growth of 15
percent.

Figure 3: Growth in DBS Subscribers and DBS Penetration Rates, 2001-2004

Growth percentages 60

50

40

30

20

10

0 Urban Suburban Rural Geographic area

Subscriber growth
DBS penetration growth
Source: GAO.

  DBS Penetration Is Higher Where Cable Service Is Not Available, Where Cable
  Providers Do Not Offer Advanced Services, and Where Wire-Based Competitors Are
  Not Present

Less than 9 percent of American households do not have the opportunity to
purchase cable television service because it is not available where they
live. However, in these areas, the DBS penetration rate is about 53
percentage points greater than in areas where cable television service is
available. Where cable television service is available, cable operators
are increasingly providing advanced services, such as digital cable, cable
modem, and telephone service. In 2004, the DBS penetration rate was over
20 percentage points greater in areas where cable operators did not
provide advanced services, compared with areas where these services were
available. Finally, in some limited areas, cable companies compete with
other wire-based competitors, and where there is more than one wirebased
cable competitor, the DBS penetration rate was 8 percentage points lower
than in areas without such an additional competitor.

    DBS Penetration Is Much Higher in Areas without Cable Service

Most households in the United States have access to cable television
service. Using Knowledge Network's 2004 survey, we found that less than 9
percent of responding households reported that cable television service
was not available. According to FCC, households without access to cable
television service generally reside in smaller and rural markets.11

Where cable television service is not available, households are far more
likely to purchase DBS service. In figure 4, we illustrate the percentage
of households receiving television service through four different modes
(overthe-air, cable, DBS, and other) for areas where households report
that cable television service is available and where it is not available.
In areas where cable television service is available, 65 percent purchase
cable service, 16 percent use free over-the-air television, and about 15
percent purchase DBS service. When cable television service is not
available, a significant percentage of households-nearly 68
percent-purchase DBS service, while nearly all of the remainder-31
percent-rely on over-the-air television.

11Federal Communications Commission, Tenth Annual Report, In the Matter
of: Annual Assessment of the Status of Competition in the Market for the
Delivery of Video Programming, FCC-04-5 (Washington, D.C.; Jan. 28, 2004),
para. 21.

Figure 4: Percentage of Households Using Different Modes of Television
Reception, 2004

                          Percentage of households 100

                                       80

                                       60

                                       40

                                      20 0

                       No cable available Cable available

Availability of cable service

Other

Cable

Over-the-air DBS Source: GAO analysis of Knowledge Network's 2004 The Home
Technology Monitor survey.

    DBS Penetration Is Much Higher in Areas Where Cable Operators Do Not Offer
    Advanced Services

Since 2001, the percentage of cable operators providing advanced services
(digital cable, cable modem, and telephone services) has increased. In
figure 5, we illustrate the percentage of cable operators providing no
advanced services; one or more, but not all, advanced services; and all
three advanced services based on FCC's annual survey of cable
franchises.12 In 2001, over 18 percent of cable operators did not provide
advanced services, while less than 3 percent did not provide advanced
services by 2004. At the same time, the percentage of cable operators

12Federal Communications Commission, Order, In the Matter of: Statistical
Report on Average Rates for Basic Service, Cable Programming Services and
Equipment, DA 04-35 (Washington, D.C.; Jan. 14, 2004), and DA 02-1285
(Washington, D.C.; May 31, 2002).

providing all three advanced services increased from 16 percent in 2001 to
26 percent in 2004. In 2004, most cable operators (about 66 percent)
provided both digital cable and cable modem services, but not telephone
service.

Figure 5: Percentage of Cable Operators Providing Advanced Services
(Digital Cable, Cable Modem, and Telephone), 2001-2004 Percentage of cable
operators 100

80

60

40

20

0 2001 2002 2004 Year

All advanced services

One or more, but not all, advanced services No advanced services Sources:
FCC 2002 and 2004 surveys.

In areas where cable operators do not provide advanced services, the DBS
penetration rate is significantly greater than in areas where cable
operators provide advanced services. In figure 6, we illustrate the DBS
penetration rate for 2001, 2002, and 2004 based on the availability of
advanced services from cable operators. In 2004, the DBS penetration rate
was over 36 percent in areas where cable operators did not provide
advanced services, compared with approximately 16 percent in areas where
cable operators provided one or more, but not all, advanced services, and
only 14 percent in areas where cable operators provided all three advanced
services. In fact, the DBS penetration rate increased modestly since 2001
in areas where

cable operators provide one or more advanced services. However, the DBS
penetration rate increased 12 percentage points since 2001 in areas where
cable operators do not provide advanced services.

Figure 6: DBS Penetration Rate and Cable Operators' Provision of Advanced
Services (Digital Cable, Cable Modem, and Telephone), 2001-2004

                       DBS penetration rate (percent) 40

                                       30

                                       20

                                      10 0

2001 2002 2004 Year

All advanced services

        One or more, but not all, advanced services No advanced services

                             Sources: FCC and SBCA.

DBS Penetration Is Lower in Although the Telecommunications Act of 1996
sought to increase wire-Areas Where There Is Wire-based competition, few
American households have a choice among Based Cable Competition companies
providing television service via wire-based facilities. In a 2005

report, FCC noted that few franchise areas-about 1 percent-have

effective competition based on the presence of a wire-based competitor.13
These competitors include telephone companies, electric and gas utilities,
and broadband service providers.

In areas with more than one wire-based cable provider, the DBS penetration
rate is lower compared with areas with only one wire-based provider. In
figure 7, we illustrate the DBS penetration rate for 2004 in cable
franchise areas with and without wire-based cable competition. The DBS
penetration rate is 18 percent in areas without wire-based competition and
10 percent in areas with wire-based competition.

Figure 7: DBS Penetration Rate in Cable Franchise Areas with and without
Wire-Based Cable Competition, 2004

Status of wire-based cable competition

  Several Geographic and Competitive Factors Are Associated with Different
  Levels of DBS Penetration across Cable Franchise Areas

We found that three key geographic factors and three key competitive
factors influence DBS penetration rates in cable franchise areas
throughout the United States. Regarding geographic factors, we found that
(1) the DBS penetration rate is lower in areas with a high prevalence of
multiple dwelling units, such as apartments and condominiums; (2) the DBS
penetration rate is lower in areas where the angle at which the satellite
dish must be installed is relatively low, such that the satellite points
more toward the horizon than toward the sky; and (3) the DBS penetration
rate is higher in nonmetropolitan areas. In terms of competitive factors,
we found that (1) the DBS penetration rate is lower in areas where the
cable operator's system has greater system capacity; (2) the DBS
penetration rate is lower in areas where there is more than one wire-based
cable provider;

13Federal Communications Commission, Eleventh Annual Report, In the Matter
of: Annual Assessment of the Status of Competition in the Market for the
Delivery of Video Programming, FCC-05-13 (Washington, D.C.; Feb. 4, 2005),
para. 136.

Wire-based competition present

No wire-based competition

0 51015

                         DBS penetration rate (percent)

                             Sources: FCC and SBCA.

and (3) the DBS penetration rate is higher in areas where DBS providers
carry local broadcast stations, such as an ABC affiliate.

    Key Geographic Factors Influence DBS Penetration Rates

Using an econometric model to control for the many factors that influence
the DBS penetration rate, we identified three geographic factors that
influenced the DBS penetration rate in cable franchise areas in 2004; see
appendix III for a full explanation of, and results from, our econometric
model.

o 	The DBS penetration rate is lower in areas with a relatively large
number of housing units represented by multiple dwelling units (such as
apartments and condominiums). A 10 percent increase in the percentage of
housing units represented by multiple dwelling units is associated with a
2.5 percent decrease in the DBS penetration rate. One possible explanation
for this result is that residents of multiple dwelling units are more
likely to encounter greater difficulty installing a DBS satellite dish,
since the dish requires a clear line of sight to the southern sky.14

o 	The DBS penetration rate is lower in areas where, to see the southern
sky, the satellite dish must be pointed more toward the horizon than up at
the sky. In general, the farther north one is within the United States,
the more the dish must be angled toward the horizon to see the satellite
over the equator. We found that a 1 percent decrease in the angle at which
the DBS satellite dish must be set at is associated with a 1 percent
decrease in the DBS penetration rate. A possible explanation for this
result is that a satellite dish facing the horizon is less likely to have
a clear line of sight to the southern sky because of interference from
surrounding buildings or trees.

o 	The DBS penetration rate is generally higher in nonmetropolitan areas.
The DBS penetration rate is about 41 percent greater in cable franchise
areas outside metropolitan areas compared with cable franchise areas
within metropolitan areas. This result is consistent with the results
discussed above for 2001 to 2004 and may be attributed to the early
popularity of satellite service in rural areas.

14To mitigate this problem, DBS providers could negotiate with MDU owners
to install a single satellite dish on the roof of MDU buildings and
subsequently relay the signal to residents via internal wiring.

    Key Competitive Factors Influence DBS Penetration Rates

Using the same econometric model, we also identified three competitive
factors that influence the DBS penetration rate in cable franchise areas
in 2004.

o 	The DBS penetration rate is lower in areas where the cable operator's
system has greater capacity. A 10 percent increase in the cable operator's
system capacity is associated with a 2.4 percent decrease in the DBS
penetration rate. With greater system capacity, a cable operator can
provide more channels and advanced services, such as digital cable, cable
modem, and telephone services. Thus, greater system capacity allows the
cable operator to provide a compelling alternative to DBS service that can
contribute to lower DBS penetration rates. This result is consistent with
the lower DBS penetration rate in areas where cable operators provided
advanced cable services for 2001 to 2004 that we discussed above.

o 	The DBS penetration rate is lower in areas with wire-based cable
competition, compared with areas without wire-based competition. In
particular, we found that DBS penetration rates are about 37 percent lower
in areas with wire-based cable competition compared with areas without
wire-based competition. Again, this result is consistent with the results
discussed above. With wire-based competition, additional companies are
competing for customers. The addition of a second cable operator can
attract some customers who might otherwise have purchased DBS service,
thereby reducing the DBS penetration rate.

o 	The DBS penetration rate is higher in areas where DBS customers can
receive local-into-local service. Local-into-local service allows DBS
subscribers to receive the local broadcast stations in their area (e.g.,
the ABC, CBS, Fox, and NBC affiliates) from the DBS provider, just as
cable subscribers receive local broadcast stations from their cable
operator. Since individual programming appearing on broadcast stations
generally has higher ratings than individual programming appearing on
cable channels, the ability of DBS providers to offer local broadcast
stations to their customers remains an important competitive factor. We
found that where local-into-local service is available, the DBS
penetration rate is about 12 percent higher than areas where
local-into-local is not available.

Agency Comments and 	We provided a draft of this report to the Federal
Communications Commission (FCC) for its review and comment. FCC staff
provided

Our Evaluation technical comments that we incorporated, where appropriate.

  Industry Participants' Comments

We provided a draft of this report to the National Cable and
Telecommunications Association (NCTA) and the Satellite Broadcasting and
Communications Association (SBCA) for their review and comment. NCTA
provided no comments. SBCA officials noted that, in addition to the
factors we discuss in the report, the inability of DBS providers to carry
certain programming developed by cable operators also influences the DBS
penetration rate in certain markets. In particular, SBCA noted that FCC's
program access rules require that vertically integrated cable operators
make satellite-delivered programming available to competing subscription
video providers, such as DBS providers, but that the program access rules
do not apply to terrestrially delivered programming. SBCA officials note
that the ability of cable operators to deliver programming terrestrially,
especially popular programming such as regional sports networks, and
therefore deny DBS providers access to this programming, negatively
affects the DBS penetration rate in certain markets.

As agreed with your office, unless you publicly announce its contents
earlier, we plan no further distribution of this report until 15 days
after the date of this letter. At that time, we will send copies to
interested congressional committees; the Chairman, FCC; and other
interested parties. We will also make copies available to others upon
request. In addition, this report will be available at no cost on the GAO
Web site at http://www.gao.gov. If you or your staff have any questions
concerning this report, please contact me at (202) 512-2834 or at
[email protected].

Major contributors to this report include Amy Abramowitz, Stephen Brown,
Michael Clements, Simon Galed, and Bert Japikse.

Mark L. Goldstein Director, Physical Infrastructure Issues

Appendix I

Scope and Methodology

To respond to the first and second objectives-to provide information on
how direct broadcast satellite (DBS) subscribership has changed since 2001
and how the DBS penetration rate differs across urban, suburban, and rural
areas-we gathered data on DBS subscribers from the Satellite Broadcasting
and Communications Association (SBCA). SBCA provided us with the number of
DBS subscribers by ZIP Code(TM) 1 for the two DBS providers, DIRECTV(R)
and EchoStar. Using information from the Census Bureau and a private
vendor, we matched the zip codes to counties and calculated the number of
DBS subscribers in each county throughout the United States. We also
gathered data on housing unit projections from the Census Bureau, which,
when combined with the number of DBS subscribers, allowed us to calculate
the DBS penetration rate by county for July 2001 to January 2004. This
allowed us to examine changes in the DBS penetration rate for that period
of time. Further, using data from the Office of Management and Budget, we
classified counties as urban, suburban, and rural, based on the location
of central cities and designations of metropolitan statistical areas
(MSA). This allowed us to calculate the DBS penetration rate for each of
these geographic categories.

To respond to the third objective-to provide information on the DBS
penetration rate based on the degree and type of competition-we used two
different methodologies. First, to examine the DBS penetration rate in
areas with and without cable service, we used survey data from Knowledge
Network's The Home Technology Monitor: Spring 2004 Ownership and Trend
Report. Knowledge Networks is a survey research firm that conducted a
consumer survey on household television characteristics. Knowledge
Networks interviewed 2,471 randomly sampled telephone households, asking
questions regarding the household's ownership of television equipment and
use of television service; see appendix II for a discussion of the steps
we took to evaluate the reliability of Knowledge Network's data. The
survey included questions regarding whether the household had cable
service available and which method the household used to receive
television (e.g., over-the-air, cable, or DBS). We used these data to
identify the percentage of households receiving DBS service in areas with
and without cable service. Second, to examine the DBS penetration rate in
areas with advanced cable services and wire-based cable competition, we
used data from FCC's annual Cable Price Survey. We used data from FCC's
2002 and 2004 surveys, which included questions

1ZIP Code(TM) is a registered trademark of the United States Postal
Service. For simplicity, we refer these as zip codes.

Appendix I Scope and Methodology

regarding the availability of digital cable, cable modem, and telephone
service and the presence of wire-based competition. We matched individual
zip codes to the cable franchise areas that formed the unit of analysis in
FCC's survey. When combined with the count of DBS subscribers by zip code
from SBCA, we calculated the DBS penetration rate for each cable franchise
area in FCC's survey. We used these data, combined with cable operators'
responses to FCC's survey regarding advanced services and wire-based
competition, to calculate the DBS penetration rate under these various
scenarios.

To respond to the fourth objective-to provide information on the factors
that appear to influence the DBS penetration rate in cable franchise
areas-we used an econometric model we previously developed that examines
the effect of competition on cable rates and service and the DBS
penetration rate.2 Using data from FCC's 2004 Cable Price Survey, the
model considered the effect of various factors on cable rates, the number
of cable subscribers, the number of channels that cable operators provide
to subscribers, and the DBS penetration rate for areas throughout the
United States. See appendix III for a more detailed explanation of, and
results from, our econometric model.

2See, GAO, Telecommunications: Issues Related to Competition and
Subscriber Rates in the Cable Television Industry, GAO-04-8 (Washington,
D.C.: Oct. 24, 2004); Telecommunications: Issues in Providing Cable and
Satellite Television Services, GAO-03130 (Washington, D.C.: Oct. 15,
2002); and Telecommunications: The Effect of Competition from Satellite
Providers on Cable Rates, GAO/RCED-00-164 (Washington, D.C.: July 18,
2000).

Appendix II

Data Reliability

To respond to the objectives of this report, we relied extensively on
three data sets and took steps to ensure the reliability of these data.
The data sets we relied on include the Federal Communications Commission's
(FCC) 2002 and 2004 Cable Price surveys, direct broadcast satellite (DBS)
subscriber counts by zip code from the Satellite Broadcasting and
Communications Association (SBCA), and Knowledge Network's 2004 The Home
Technology Monitor survey. In this appendix, we explain the steps we took
to ensure that these data were sufficiently reliable for the purposes of
our work.

FCC's Cable Price surveys	FCC annually surveys approximately 700 cable
franchises to fulfill a congressional mandate to report on average cable
rates for cable operators found to be subject to "effective competition"-a
legally defined term- compared with operators not subject to effective
competition. In previous testimonies and a report, we have noted
weaknesses with FCC's survey, including insufficient instructions and
inaccuracies in the classification of the competitive status of cable
operators.1 In response to our recommendations, FCC has taken several
steps to improve the reliability of its survey, including editing the
survey document and correcting inaccurate classifications of the
competitive status of cable franchises. Additionally, FCC conducts
follow-ups with survey respondents and edits survey data when inaccuracies
are apparent.

We used FCC's 2002 and 2004 Cable Price surveys to identify areas where
cable operators provided advanced services and also for information on
price, number of channels, and other operating data necessary for our
cable-satellite econometric model. Because our use of data from FCC's
surveys was important in a comparative manner, rather than an absolute
sense-that is, our primary concern with cable rates was the relative level
of rates between cable franchises, rather than the absolute rate in a
particular cable franchise-it is not important for our use that the data
be precise. We conducted logic tests to identify any observations with
apparent inaccuracies in the variables of interest for our work. We
determined that the data were sufficiently reliable for our analysis.

1See, GAO, Telecommunications: Data Gathering Weaknesses in FCC's Survey
of Information on Factors Underlying Cable Rate Changes, GAO-03-742T
(Washington, D.C.: May 6, 2003); Telecommunications: Issues Related to
Competition and Subscriber Rates in the Cable Television Industry,
GAO-04-8 (Washington, D.C.: Oct. 24, 2003); and Telecommunications:
Subscriber Rates and Competition in the Cable Television Industry,
GAO-04-262T (Washington, D.C.: Mar. 25, 2004).

                          Appendix II Data Reliability

  SBCA's DBS Subscriber Counts

SBCA possesses data on the number of DBS subscribers by zip code. To
respond to the objectives of this report, we sent SBCA a letter
identifying the specific data elements we required. SBCA officials
prepared a set of data sets consistent with our needs. We conducted logic
tests on SBCA's data and identified some inconsistencies, which we
discussed with SBCA officials. SBCA officials subsequently took steps to
resolve these inconsistencies. Based on the revised data we received from
SBCA and our subsequent tests, we determined that the data were
sufficiently reliable for our analysis.

  Knowledge Network's The Home Technology Monitor: Spring 2004 Ownership and
  Trend Report

To obtain information on the availability of cable service and types of
television service used by U.S. households, we purchased existing survey
data from Knowledge Networks Statistical Research. This survey was
completed with 2,375 of the estimated 5,075 eligible sampled individuals
for a response rate of 47 percent; partial interviews were conducted with
an additional 96 people, for a total of 2,471 individuals completing some
of the survey questions. The survey was conducted between February 23 and
April 25, 2004. Because we did not have information on those contacted who
chose not to participate in the survey, we could not estimate the impact
of the nonresponse. Our findings will be biased to the extent that the
people at the 53 percent of the telephone numbers that did not yield an
interview have experiences with television service or equipment that are
different from the 47 percent of our sample who responded. However,
distributions of selected household characteristics (including presence of
children, race, and household income) for the sample and the U.S. Census
estimate of households show a similar pattern.

To assess the reliability of these survey data, we reviewed documentation
of survey procedures provided by Knowledge Networks and questioned
knowledgeable officials about the survey process and resulting data. We
determined that the data were sufficiently reliable for the purposes of
this report.

Appendix III

                       Cable-Satellite Econometric Model

This appendix describes our econometric model of cable-satellite
competition. In particular, we discuss (1) the specification of the model,
(2) the data sources used for the model, (3) the merger of various data
sources into a single data set, (4) the descriptive statistics for
variables included in the model, (5) the estimation methodology and
results, and (6) alternative specifications.

  Specification of Econometric Model of Cable-Satellite Competition

We developed an econometric model to examine the influence of various
factors, including those describing aspects of cable competition at the
local level, on local DBS penetration rates. Estimating the importance of
various factors on the DBS penetration rate is complicated by the
possibility that the DBS penetration rate in an area may help determine,
but also be determined by, in part, the local cable price in that area.
One statistical method applicable in this situation is to estimate a
system of structural equations in which certain variables that may be
simultaneously determined are estimated jointly. In our previous reports,
we estimated a four-equation structural model in which cable prices, the
number of cable subscribers, the number of cable channels, and the DBS
penetration rate were jointly determined.1 We use this same general
structure again, this time using the most recent information available
from FCC's 2004 Cable Price Survey and contemporaneous satellite
subscriber information provided by the Satellite Broadcasting and
Communications Association. We made some minor modifications because of,
for example, changes in the subscription video market.

We estimated the following four-equation structural model of the
subscription video market:

o 	DBS penetration rate in a local market is hypothesized to be related to
(1) cable prices per channel; (2) the DBS companies' provision of local
stations in the franchise area; (3) the size of the television market as
measured by the number of television households; (4) the age of the cable
franchise; (5) the median household income of the local area; (6) cable
system capacity in terms of megahertz; (7) a dummy variable for

1See GAO, Telecommunications: Issues Related to Competition and Subscriber
Rates in the Cable Television Industry, GAO-04-8 (Washington, D.C.: Oct.
24, 2003); GAO-03-130 (Washington, D.C.: Oct. 15, 2002); and
Telecommunications: The Effect of Competition from Satellite Providers on
Cable Rates, GAO/RCED-00-164 (Washington, D.C.: July 18, 2000).

Appendix III
Cable-Satellite Econometric Model

areas outside metropolitan areas; (8) the percentage of multiple dwelling
units; (9) the angle, or elevation, at which a satellite dish must be
fixed to receive a satellite signal in that area; and (10) the presence of
a nonsatellite competitor. The DBS penetration rate variable is defined as
the number of DBS subscribers in a franchise area expressed as a
proportion of the total number of housing units in the area. As
hypothesized, the DBS penetration rate is expected to depend on the prices
set by the cable provider as well as on the demand, cost, and regulatory
conditions in the subscription video market that directly affect DBS.

o 	Cable prices are hypothesized to be related to (1) the number of
channels, (2) the number of cable subscribers, (3) the DBS penetration
rate, (4) the DBS companies' provision of local stations in the franchise
area, (5) the size of the television market as measured by the number of
television households, (6) horizontal concentration, (7) vertical
relationships, (8) the presence of a nonsatellite competitor, (9)
regulation, (10) average wages, and (11) population density. The cable
price variable used in the model is intended to reflect the total monthly
rate charged by a cable franchise to the typical subscriber. The
explanatory variables in the cable price relationship are essentially cost
and market structure variables.

o 	Number of cable subscribers is hypothesized to be related to (1) cable
prices per channel, (2) the DBS penetration rate, (3) the number of
broadcast stations, (4) urbanization, (5) the age of the cable franchise,
(6) the number of homes passed by the cable system, (7) the median
household income of the local area, and (8) the presence of a nonsatellite
competitor. The number of cable subscribers is defined as the number of
households in a franchise area that subscribe to the most commonly
purchased programming tier. This represents the demand equation for cable
services, which depends on rates and other demandrelated factors.

o 	Number of channels is hypothesized to be related to (1) the number of
cable subscribers, (2) the DBS penetration rate, (3) the size of the
television market as measured by the number of television households, (4)
the median household income of the local area, (5) cable system capacity
in terms of megahertz, (6) the percentage of multiple dwelling units, (7)
vertical relationships, and (8) the presence of a nonsatellite competitor.
The number of channels is defined as the number of channels included in
the most commonly purchased programming tier.

                                  Appendix III
                       Cable-Satellite Econometric Model

The number of channels can be thought of as a measure of cable programming
quality and is explained by a number of factors that influence the
willingness and ability of cable operators to provide highquality service
and consumers' preference for quality.

Table 1 presents the explanatory variables in the structural model on
cable prices and DBS penetration rates.

          Table 1: Explanatory Variables Used in Cable-Satellite Model

Explanatory variable Definition of variable

Cable price	The monthly rate charged for the Basic Service Tier and Cable
Programming Service Tier (the most commonly purchased tier).

Number of cable subscribers	The number of subscribers to the Basic Service
Tier and Cable Programming Service Tier (the most commonly purchased
tier).

Number of channels	The number of channels provided with the Basic Service
Tier and Cable Programming Service Tier (the most commonly purchased
tier).

DBS penetration rate The fraction of housing units in a cable franchise
area that have satellite service.

DBS provision of local stations	A binary variable that equals 1 if one or
both DBS providers offer local broadcast stations in the cable franchise
area.

Television market size The number of television households in the market.

Horizontal concentration	A binary variable that equals 1 if the cable
operator providing service in the franchise area is affiliated with a
multiple system operator (MSO) that serves over 1 million subscribers
nationally.

Vertical relationship	A binary variable that equals 1 if the cable
operator is affiliated with an MSO that has an ownership interest in a
national or regional video programming service.

Presence of nonsatellite competitor

A binary variable that equals 1 if a second wireline company provides
cable service (including, for example, a broadband service provider) in
the franchise area.

Average wage	The average weekly wage for telecommunications equipment
installers and repairers in the metropolitan area, or state, in the case
of nonmetropolitan areas, where the cable franchise is located.

Population density The ratio of population to square miles in the
franchise area.

Number of broadcast stations The number of over-the-air broadcast stations
in the television market.

Urbanization The percentage of the county's population that is classified
as urban by the Census Bureau.

Age of cable franchise The number of years between when the cable
franchise began operation and 2004.

Homes passed by cable system	The number of homes passed by the cable
system that serves the franchise area, including homes outside the
franchise area.

Median household income The median household income in the franchise area.

Cable system megahertz The capacity, measured in megahertz, of the cable
system that serves the franchise area.

Percentage of multiple dwelling The percentage of housing units accounted
for by structures with five or more housing units. units

 Nonmetropolitan areas A binary variable that equals 1 if the franchise area is
                 outside a metropolitan statistical area (MSA).

                                  Appendix III
                       Cable-Satellite Econometric Model

                         (Continued From Previous Page)

                  Explanatory variable Definition of variable

Angle (or "elevation") of satellite The angle relative to the ground that
a DBS subscriber must mount the satellite dish to "see" the dish
satellite.

Regulation	A binary variable that equals 1 if the cable franchise is
subject to regulation of the rate charged for the Basic Service Tier.

                                  Source: GAO.

  Data Sources Used for the Econometric Model

We required several data elements to build the data set used to estimate
this model. The following is a list of our primary data sources.

o 	We obtained data on cable prices and service characteristics from the
2004 Cable Price Survey that FCC conducted as part of its mandate to
report annually on cable prices. FCC's survey asked a sample of cable
franchises to provide information, as of January 1, 2004, about a variety
of items pertaining to cable prices, service offerings, subscribership,
franchise area reach, franchise ownership, and system capacity. We used
the survey to define measures of each franchise area's cable prices,
number of subscribers, and number of channels as described above. In
addition, we used the survey to define variables measuring (1) system
megahertz (the capacity of the cable system in megahertz), (2) homes
passed by the cable system serving the franchise area and perhaps other
franchises in the same area, (3) regulation-a dummy variable equal to 1 if
the franchise is subject to rate regulation of its Basic Service Tier, (4)
horizontal concentration-a dummy variable equal to 1 if the franchise area
is affiliated with one of the largest MSOs with at least 1 million
subscribers nationally, and (5) the status of nonsatellite competition-a
dummy variable equal to 1 if the franchise faced competition from a second
wireline company that provides cable service.

o 	From the Satellite Broadcasting and Communications Association, we
obtained DBS subscriber counts as of January 2004 for each zip code in the
United States. We used this information to calculate the number of DBS
subscribers in a cable franchise area, which, when divided by the number
of housing units, was used to define the DBS penetration rate.

o 	We used the most recent data from the Census Bureau to obtain the
following demographic information for each franchise area: housing units,
median household income, proportions of urban and rural populations,
housing units accounted for by structures with more than

Appendix III
Cable-Satellite Econometric Model

five units (multiple dwelling units), population density, and
nonmetropolitan statistical areas.

o 	For average wage, we used May 2003 estimates for Installation,
Maintenance, and Repair Occupations from the Bureau of Labor Statistics'
(BLS) National Occupational Employment and Wage Estimates. We used
metropolitan area data for most franchise areas, and state-level data for
those franchise areas located outside of metropolitan areas.

o 	We used data from BIA MEDIA AccessPro(TM) to determine the number of
broadcast television stations in each television market.

o 	To define the dummy variable indicator of vertical integration, we used
information on the corporate affiliations of the franchise operators
provided in FCC's survey. We used this information in conjunction with
industrywide information on vertical relationships between cable operators
and suppliers of program content gathered by FCC in its Tenth Annual
Report on the status of competition in the market for delivery of video
programming.

o 	From Nielsen Media Research, we acquired information to determine the
number of television households in each designed market area (DMA), or
television market, and the DMA in which each cable franchise was located.

o 	We used information from the two DBS companies (DIRECTV(R) and
EchoStar) to identify DMAs in which these companies provide local stations
and, if local stations are available, when the companies initiated this
service. We used this to construct a measure of local station
availability, as well as alternative specifications presented in the final
section.

o 	Based on a zip code associated with each cable franchise area, we
determined the necessary satellite dish elevation for each area based on
information available from the Web pages of the two DBS companies.

                                  Appendix III
                       Cable-Satellite Econometric Model

  Merging Various Data Sources into a Single Data Set

The level of observation in our model is the local cable franchise.2 Many
of the variables we used to estimate our model, such as each cable
franchise's price, come directly from FCC's Cable Price Survey. However,
we also created variables describing competitive, geographic, and economic
conditions in each franchise area. For these variables, we used
information from other sources. For example, we obtained median household
income and the extent of multiple dwelling units from Census Bureau data,
and derived the DBS penetration rate from information provided by the
Satellite Broadcasting and Communications Association. Generally, these
data are reported at other geographic levels, and we describe briefly the
process by which we merged these different data sources.

Cable franchise areas take a variety of jurisdictional forms, such as city
or town, or unnamed, unincorporated area. As a consequence, they do not
correspond in many cases to well-recognized geographical units, such as
Census places, for which other data are readily available. Our approach to
identifying the geographic extent of each franchise area and relating
information processed at different geographic levels to each franchise
area is similar to that we have used and described in detail in our
previous reports. In general, we used information in FCC's survey
identifying franchise community name and type (such as city or town) to
match to Census geographic identification codes for particular places or
county subdivisions that do correspond to Census geography. In particular,
we used 2000 Census information on the number of housing units in these
jurisdictions as the basis for our measure of DBS penetration. For other
franchises, however, the link to Census records was not as direct. For
franchises in unincorporated unnamed areas, for example, and those whose
franchise areas represent a section of the associated community (which
occurs in some large cities3), we acquired additional information on the
geographic boundaries of the franchise areas.4

2We define a cable franchise in terms of its FCC assigned Community Unit
Identification (CUID) number.

3Many large cities, such as New York City, Los Angeles, and Chicago, have
multiple cable franchise areas.

4For those jurisdictions for which there were multiple franchises,
including counties with franchises in unincorporated unnamed areas, we
attempted to define more precise geographical boundaries for each
franchise. Specifically, we contacted local government offices responsible
for cable franchise oversight and received maps or other descriptive
information linking the specific franchise areas to zip codes, census
tracts, local government districts, or some other boundary information.

Appendix III
Cable-Satellite Econometric Model

The satellite subscriber information we obtained was organized by zip
code. In order to link these subscriber counts to franchise area
geographies, we determined the zip code or zip codes associated with each
franchise. Because zip codes often do not share boundaries with other
geographies, one zip code can be associated with more than one cable
franchise area. Also, many franchises, particularly larger ones, span many
zip codes. Therefore, we needed to identify the zip code or codes in each
franchise area as well as the degree to which each of those zip codes is
contained in each franchise area to calculate the degree of satellite
penetration for each franchise area. We accomplished this by using
software designed to relate various levels of census geography to one
another.5 For most franchise areas-that is, those that correspond to
census places, county subdivisions, or entire counties-we were able to use
this software to relate census places, county subdivisions, or other
census geographies directly to the zip codes that corresponded to those
areas and to calculate the share of each zip code's population according
to the 2000 Census that was contained in that area. We used these
population shares to allocate shares of each zip code's total DBS
subscribers to the relevant franchise area, and then summed the resulting
subscribers across all zip codes in that franchise area.6 We defined the
penetration by dividing this subscriber total by an estimate of the
housing units in that franchise area in January 2004.7

5In particular, we used the MABLE/Geocorr correspondence engine
(http://mcdc2.missouri.edu/websas/geocorr2k.html). MABLE is an acronym for
Master Area Block Level Equivalency file.

6As an illustration, assume we have a cable franchise area in the town of
Anytown, which the MABLE software identifies as served by zip codes 12345
and 12346. Assume further that zip code 12345 had a population of 10,000
people in 2000, of which 8,000 were in Anytown proper and 2,000 were in
the surrounding unincorporated area, and zip code 12346 had a population
of 12,000 people, of which 6,000 were in Anytown. In this case, 80 percent
of the 12345 zip code and 50 percent of the 12346 populations are
associated with Anytown, so that our approach would assign 80 percent of
the satellite subscribers in zip code 12345 and 50 percent of those in
12346 to the cable franchise in the town of Anytown. Because we defined
the DBS penetration rate as the number of subscribers divided by the
number of housing units, our approach would divide this estimate of the
number of DBS subscribers in Anytown by the number of housing units
reported in the 2000 Census for the town of Anytown.

7We used county-level housing unit projections made by the Census Bureau
to adjust the 2000 housing unit counts to January 2004. This adjustment
process assumes that growth was uniform within the boundaries of each
county.

                                  Appendix III
                       Cable-Satellite Econometric Model

As part of the process of identifying the zip codes associated with each
franchise area, we identified a key zip code that we used for linking
other data items. We used Census data organized at the zip code level to
assign demographic data, such as income and the extent of multiple
dwelling units, to each franchise area. We also used this key zip code to
attach information concerning the proper satellite dish elevation.

We assigned other information to each franchise on the basis of the
franchise's county, state, or metropolitan area. We assigned wage data
from BLS at the metropolitan or state level and we assigned
nonmetropolitan status, percentage of urban population, and the Nielsen
television market of each franchise at the county level.8 Information on
the provision of local stations by DBS companies, which occurs at the
television market level, was then assigned to each franchise.

Descriptive Statistics for Table 2 provides basic statistical information
on all of the variables Variables Included in the included in the
cable-satellite competition model. We calculated these Econometric Model
statistics using 624 observations in our data set. We excluded those

franchises sampled by FCC that were municipally operated or that competed
directly with municipally operated franchises because we believe that
these cable franchises are likely to be operated differently from the
majority of other franchises.

  Table 2: Descriptive Statistics for Variables Used in Cable-Satellite Model

               Variable                  Mean Standard  Minimum Maximum value 
                                              deviation  value  
              Cable price               40.25      5.10   16.99         53.90 
      Number of cable subscribers    27,497.7  50,744.9    35.0     401,174.0 
          Number of channels             70.7      11.5    14.0         120.0 
       Number of local broadcast         15.6       6.6     3.0 
               stations                                         
         DBS penetration rate            16.2      10.6     1.1          77.2 
             Urbanization                78.6      25.2     0.0         100.0 
        Age of cable franchise           27.7      10.2     2.0          53.0 

8In the Nielsen data, some counties are split between different television
markets. In cases where a franchise's county was not uniquely placed in
one television market, we used additional information on zip codes to
assign the franchise to a television market.

                                  Appendix III
                       Cable-Satellite Econometric Model

                         (Continued From Previous Page)

              Variable                  Mean  Standard  Minimum Maximum value 
                                             deviation   value  
    Homes passed by cable system   233,493.8  265,476.7   190.0   1,368,050.0 
       Median household income          45.6       16.3    18.7         146.8 
            Average wage               704.1      66.22  475.19         852.7 
       Cable system megahertz          730.6      127.2   212.0         870.0 
             Regulation                 0.36       0.48    0.00 
      Horizontal concentration          0.83       0.38    0.00 
        Vertical relationship           0.64       0.48    0.00 
Percentage of multiple dwelling     16.04      13.96    0.00         98.12 
                units                                           
      Presence of nonsatellite          0.22       0.41    0.00 
             competitor                                         
       Television market size        1,685.4    1,842.5    57.0       7,301.0 
        Nonmetropolitan area            0.21       0.40    0.00 
DBS provision of local stations      0.85       0.36    0.00 
         Population density          2,949.6    4,891.9    16.8      66,940.0 
      Angle (or "elevation") of         40.1        6.5    27.6 
           satellite dish                                       

                                  Source: GAO.

Estimation Methodology We employed the Three-Stage Least Squares (3SLS)
method to

and Results	estimate our model.9 Table 3 includes the estimation results
for each of the four structural equations. All of the variables, except
dummy variables,10 are expressed in natural logarithmic form.11 This means
that coefficients can be interpreted as "elasticities"-the percentage
change in the value of the dependent variable associated with a 1 percent
change in the value of an independent, or explanatory,

9We preferred 3SLS to Two-Stage Least Squares (2SLS) because 3SLS accounts
for the contemporaneous relationships among cable rates, cable
subscribers, cable channels, and DBS penetration by using all available
information. Also, we assumed that price per channel in the subscriber
equation is exogenous because cable providers simultaneously decide how
many channels to provide and what to charge for a package of channels,
rather than deciding how much to charge for each channel.

10A dummy variable takes a value of 1 if a certain characteristic is
present and a value of 0 otherwise.

11The dummy variables in the model include the following: horizontal
concentration of cable systems, vertical relationship, regulation,
presence of nonsatellite competitor, DBS provision of local stations, and
nonmetropolitan area. Also, because the natural log of 0 is undefined, we
added 1 to the observed value of any continuous variable that can take the
value of 0.

                                  Appendix III
                       Cable-Satellite Econometric Model

variable. The coefficients on the dummy variables are elasticities in
decimal form.

                Table 3: Three-Stage Least Squares Model Results

                              Cable prices Cable       Cable              DBS 
                                           subscribers channels   penetration 
            Variable              equation    equation   equation    equation 
    Cable price per channel                    -2.6260                 0.4582 
                                             [0.0001]a              [0.0002]a 
       Number of channels           0.3955                        
                                 [0.0001]a                        
        Number of cable            -0.0131                 0.0340 
          subscribers                                             
                                  [0.1692]              [0.0001]a 
        DBS penetration            -0.0476     -1.4420     0.0419 
                                 [0.0152]b   [0.0001]a  [0.0586]c 
     DBS provision of local         0.0139                             0.1131 
            stations                                              
                                  [0.4317]                          [0.0770]c 
           Regulation               0.0157                        
                                  [0.2234]                        
      Number of broadcast                                         
            stations                            0.2838            
                                             [0.0366]b            
    Median household income                    -0.3974     0.0673      0.2006 
                                             [0.0358]b  [0.0007]a   [0.0026]a 
    Horizontal concentration        0.0133                        
                                  [0.4591]                        
     Vertical relationship         -0.0414                 0.0163 
                                 [0.0035]a               [0.2815] 
    Presence of nonsatellite       -0.1694     -1.4280     0.0808     -0.4607 
           competitor                                             
                                 [0.0001]a   [0.0001]a  [0.0001]a   [0.0001]a 
      Nonmetropolitan area                                             0.3460 
                                                                    [0.0001]a 
          Urbanization                          0.4624            
                                             [0.0001]a            
     Percentage of multiple                                0.0032     -0.2485 
         dwelling units                                           
                                                         [0.7428]   [0.0001]a 
     Age of cable franchise                     0.2738                -0.1332 
                                             [0.0236]b              [0.0011]a 
     Homes passed by cable                      0.2546            
             system                                               
                                             [0.0001]a            
     Cable system megahertz                                0.4654     -0.2406 
                                                        [0.0001]a   [0.0118]b 
     Television market size        -0.0067                 0.0267      0.0848 
                                  [0.3588]              [0.0001]a   [0.0003]a 

                                  Appendix III
                       Cable-Satellite Econometric Model

                         (Continued From Previous Page)

                               Cable prices Cable       Cable             DBS 
                                            subscribers channels  penetration 
            Variable               equation    equation  equation    equation 
       Population density            0.0015                       
                                   [0.8090]                       
          Average wage               0.0609                       
                                   [0.3791]                       
    Angle (or "elevation") of                                          1.0697 
         satellite dish                                           
                                                                    [0.0001]a 
            Intercept                1.9190      9.1412   -0.1648     -1.2538 
                                  [0.0001]a   [0.0001]a  [0.5435]    [0.2452] 
           Sample size                  624         624       624 

Source: GAO.

Notes: System-weighted R-square: 0.49. P-values are in square brackets.
aSignificant at the 1 percent level. bSignificant at the 5 percent level.
cSignificant at the 10 percent level.

We found that several factors related to the geographical conditions
influence the DBS penetration rate. Specifically, as shown in table 3, DBS
penetration rates are likely to be significantly higher in nonmetropolitan
areas. This could be associated with the historical development of
satellite service, which had been marketed for many years in smaller and
more rural areas. Additionally, the DBS penetration rate is higher in
areas that require a relatively higher angle or elevation at which the
satellite dish is mounted and is lower in areas where there are more
multiple dwelling units. These two factors can be associated with the need
of DBS satellite dishes to "see" the satellite: A dish aimed more toward
the horizon (as opposed to aimed higher in the sky) is more likely to be
blocked by a building or foliage, and people in multiple dwelling units
often have fewer available locations to mount a satellite dish.

Additionally, we found that several factors related to competitive
conditions influence the DBS penetration rate. As shown in table 3, our
model results indicate that in cable franchise areas where local broadcast
stations are available from one or both DBS providers, the DBS penetration
rate is approximately 12 percent higher than in areas where local stations

                                  Appendix III
                       Cable-Satellite Econometric Model

are not available via satellite.12 This finding suggests that in areas
where local stations are available from one or both DBS providers,
consumers are more likely to subscribe to DBS service and, therefore, DBS
appears to be more competitive with cable than in areas where local
stations are not available from a DBS provider.

We did not find that DBS companies' provision of local broadcast stations
is associated with lower cable prices. In table 3, the estimate is, in
fact, positive, although not statistically significant, and we therefore
cannot reject the hypothesis that provision of local broadcast stations
has no impact on cable prices. However, we found that cable prices were
approximately 16 percent lower in areas where a second cable company-
known as an overbuilder-provides service. Finally, cable prices are higher
in areas where the cable company provides more channels, indicating that
consumers are generally willing to pay for additional channels and that
providing additional channels raises a cable company's costs.
Additionally, we found that DBS penetration rates are lower in cable
franchise areas where a second wire-based competitor is present; in these
areas, the DBS penetration rate is 37 percent lower compared with similar
areas where a second wire-based competitor is not present.

Alternative Specifications	We considered alternative specifications under
which we expanded the definition of local broadcast stations to account
for (1) whether one or both DBS companies offer local stations and (2) the
length of time that DBS companies have provided local stations. To conduct
this analysis, we included several additional variables: "Both DBS
companies provide" equals 1 if both DBS companies offer local stations in
the cable franchise area, "One DBS company provides" equals 1 if only one
DBS company offers local stations, "Long-term" equals 1 if either or both
DBS companies have offered local stations in the cable franchise area for
more than 3 years as of January 2004, "Short-term" equals 1 if local
stations have been available for less than 3 years, "Both long-term"
equals 1 if both DBS companies have offered local stations in the cable
franchise area for more

12This magnitude is less than we found previously; using 2001 data, we
found that the DBS penetration rate was about 40 percent higher in cable
franchises in which local stations were offered by both DBS companies. The
somewhat smaller impact of local stations could be because the DBS
companies have introduced this service into many more areas, including
some smaller television markets.

                                  Appendix III
                       Cable-Satellite Econometric Model

than 3 years as of January 2004, and "Both otherwise" equals 1 if local
stations have otherwise been available from both DBS companies.

We report the results of these alternative specifications only for the DBS
penetration equation because we are primarily interested in their affects
on DBS penetration and we found little impact on the other equations in
the model. We present the results for four different specifications in
table 4. In general, there is evidence that the longer that local stations
have been available in a local area, the larger will be the increase in
the local DBS penetration rate, and that the increase in the local DBS
penetration rate is greater in those areas in which both DBS companies
provide local stations.

                   Table 4: Alternative Specification Results

                         DBS penetration                     DBS          DBS 
                                                     penetration  penetration 
                               equation:         DBS   equation:    equation: 
                                         penetration             
                         main              equation: long term   duration and 
                         specification               and short         number 
         Variable         (from table 3) both or one        term of providers 
                                         provided                
      Cable price per             0.4582      0.4646      0.4690       0.5103 
          channel                                                
                               [0.0002]a   [0.0002]a   [0.0002]a    [0.0001]a 
     DBS provision of             0.1131                         
      local stations                                             
                               [0.0770]c                         
    Both DBS companies                        0.1562             
          provide                                                
                                           [0.0409]b             
      One DBS company                         0.0862                   0.1106 
         provides                                                
                                            [0.2127]                 [0.1135] 
    Long-term (greater                                    0.2207 
       than 3 years)                                             
                                                       [0.0096]a 
Short-term (less than                                  0.0929 
         3 years)                                                
                                                        [0.1493] 
      Both long-term                                                   0.2852 
                                                                    [0.0030]a 
      Both otherwise                                                   0.1347 
                                                                    [0.0793]c 
     Median household             0.2006      0.1977      0.1919       0.1921 
          income                                                 
                               [0.0026]a   [0.0031]a   [0.0040]a    [0.0039]a 
        Presence of                                                           
       nonsatellite              -0.4607     -0.4646     -0.4703      -0.4615
        competitor                                               
                               [0.0001]a   [0.0001]a   [0.0001]a    [0.0001]a 
Nonmetropolitan areas          0.3460      0.3476      0.3417       0.3417 
                               [0.0001]a   [0.0001]a   [0.0001]a    [0.0001]a 
       Percentage of                                                          
     multiple dwelling           -0.2485     -0.2469     -0.2474      -0.2448
           units                                                 
                               [0.0001]a   [0.0001]a   [0.0001]a    [0.0001]a 

Appendix III
Cable-Satellite Econometric Model

                         (Continued From Previous Page)

                       DBS penetration                       DBS          DBS 
                                                     penetration  penetration 
                             equation:         DBS     equation:    equation: 
                                       penetration               
                       main              equation: long term and duration and 
                       specification               short               number 
        Variable        (from table 3) both or one          term of providers 
                                       provided                  
      Age of cable             -0.1332     -0.1319       -0.1296      -0.1322 
        franchise                                                
                             [0.0011]a   [0.0013]a     [0.0015]a    [0.0012]a 
      Cable system             -0.2406     -0.2346       -0.2341      -0.2240 
        megahertz                                                
                             [0.0118]b   [0.0142]b     [0.0140]b    [0.0189]b 
    Television market                                                         
          size                  0.0848      0.0687        0.0445       0.0234
                             [0.0003]a   [0.0138]b      [0.1513]     [0.4981] 
        Angle (or                                                             
     "elevation") of            1.0697      1.0514        1.0202       1.0173
     satellite dish                                              
                             [0.0001]a   [0.0001]a     [0.0001]a    [0.0001]a 
        Intercept              -1.2538      -1.114       -0.8147      -0.7400 
                              [0.2452]    [0.3054]      [0.4559]     [0.5003] 
       Sample size                 624         624           624 

Source: GAO. Notes: P-values are in square brackets. aSignificant at the 1
percent level. bSignificant at the 5 percent level. cSignificant at the 10
percent level.

GAO's Mission	The Government Accountability Office, the audit, evaluation
and investigative arm of Congress, exists to support Congress in meeting
its constitutional responsibilities and to help improve the performance
and accountability of the federal government for the American people. GAO
examines the use of public funds; evaluates federal programs and policies;
and provides analyses, recommendations, and other assistance to help
Congress make informed oversight, policy, and funding decisions. GAO's
commitment to good government is reflected in its core values of
accountability, integrity, and reliability.

Obtaining Copies of The fastest and easiest way to obtain copies of GAO
documents at no cost

is through GAO's Web site (www.gao.gov). Each weekday, GAO postsGAO
Reports and newly released reports, testimony, and correspondence on its
Web site. To Testimony have GAO e-mail you a list of newly posted products
every afternoon, go to

www.gao.gov and select "Subscribe to Updates."

Order by Mail or Phone	The first copy of each printed report is free.
Additional copies are $2 each. A check or money order should be made out
to the Superintendent of Documents. GAO also accepts VISA and Mastercard.
Orders for 100 or more copies mailed to a single address are discounted 25
percent. Orders should be sent to:

U.S. Government Accountability Office 441 G Street NW, Room LM Washington,
D.C. 20548

To order by Phone:	Voice: (202) 512-6000 TDD: (202) 512-2537 Fax: (202)
512-6061

  To Report Fraud, Contact:
  Waste, and Abuse in Web site: www.gao.gov/fraudnet/fraudnet.htm

E-mail: [email protected] Programs Automated answering system: (800)
424-5454 or (202) 512-7470

Congressional	Gloria Jarmon, Managing Director, [email protected] (202)
512-4400 U.S. Government Accountability Office, 441 G Street NW, Room 7125

Relations Washington, D.C. 20548

Public Affairs	Paul Anderson, Managing Director, [email protected] (202)
512-4800 U.S. Government Accountability Office, 441 G Street NW, Room 7149
Washington, D.C. 20548
*** End of document. ***