Price, Programming and Potential Competition in US Cable Television Markets

Journal of Regulatory Economics; 27:1 25–46, 2005© 2005 Springer Science+Business Media, Inc. Manufactured in The Netherlands.

Price, Programming and Potential Competition inUS Cable Television Markets∗

SCOTT J. SAVAGEUniversity of Colorado

Department of Economics, 256 UCB, Boulder, CO 80309-0256, USAE-mail: [email protected]

MICHAEL WIRTHUniversity of Denver

Department of Mass Communications and Journalism Studies, 2490 South Gaylord Street, Denver, CO80208, USA

E-mail: [email protected]

AbstractPotential competition in the US cable television industry is empirically examined for 385 markets.Predicted entry probabilities are included in a supply-demand model to permit empirical investigationof the effect of potential competition on incumbent cable operator’s price and channel programmingdecisions. Estimation results show incumbents offer more channels to consumers in markets fac-ing greater potential competition from broadband service provider (BSP) wireline overbuilders and/orincumbent local exchange carriers. In particular, when the probability of entry rises to about 42%, theaverage cable system provides six more channels, and price per channel declines from US$ 0.77 toUS$ 0.66.

Key words: cable television, competition, entry, price, quality

JEL Classification: L8, L9, L11

1. Introduction

The Federal Communications Commission (FCC 1994 and 2001a) suggests themarket for delivery of multichannel subscription television to households is highly

∗ We thank Ben Compaine, Shane Greenstein, Anne Hoag, Donald Waldman, Bradley Wimmer, theeditor of this journal, an anonymous referee, and participants at the 30th Research Conference onCommunication, Information and Internet Policy (TPRC 2002), September 28–30, 2002, HiltonHotel, Alexandria, Virginia, for comments. Armando Galarraga and Nana Puangpathumanondprovided research assistance, and Jane Frenette and Jonathan Levy provided excellent help withFCC cable entry and competition data. The usual disclaimer applies.

26 SCOTT SAVAGE AND MICHAEL WIRTH

concentrated and characterized by substantial barriers which increase the costof entry. These barriers include regulations that delay rivals’ access to publicrights-of-way, technology limitations (e.g., direct broadcast satellite (DBS) andmultichannel multipoint distribution service (MMDS) line-of-sight problems), andincumbent operator behaviours that raise rivals’ costs (e.g., limiting programmingavailability). Cable ownership attribution and vertical integration also have implica-tions for costs, prices and entry as a few multiple-system-operators (MSOs) accu-mulate an increasing share of national cable systems and programming. As notedby Owen and Wildman (1992), Chipty (1995), and Ford and Jackson (1997), ahigher concentration of cable systems and programming can increase an MSO’ssubscriber base, and its ability to bargain down the prices paid for programs.Beard and Ford (1999) also find higher monthly subscription prices in cable com-munities served by MSOs which own cable systems serving a relatively large num-ber of subscribers on a national basis.

Technology change and deregulation of cable television and telecommunicationsduring the 1990s sees many incumbent cable operators facing competition from DBS,MMDS, broadband service provider (BSP) wireline overbuilders, and/or incumbentlocal-exchange carriers (ILECs). Several studies examine the effect of actual compe-tition on cable pricing and programming. Beard and Ford (1999) show that monthlysubscription prices for monopoly cable communities are 13–17% higher than theiroverbuilt, duopoly counterparts. The United States General Accounting Office (GAO2000) finds lower prices, and a greater number of channels, in markets where BSPwireline overbuilders, MMDS providers (also called wireless cable), and/or ILECsare present. Further, in response to an increase in DBS market share, cable incum-bents increase the number of channels offered in their service packages, but chargehigher prices. More recently, the FCC (2002) estimates a 6.3% subscription pricedifferential between “competitive” and monopoly communities. The FCC’s resultsshow that operators tend to offer more channels at a lower price when there is com-petition from overbuilders. However, these same results show no significant effect oncable demand or cable prices from DBS competition.

While both the GAO (2000) and FCC (2002) acknowledge non-price competitionbetween the incumbent and new entrant (particularly increases in the number ofchannels in service packages) is a key feature of multichannel subscription televisionmarkets, analysis of the effect of potential competition on incumbent’s pricing andprogramming is scarce. Potential competition has long been recognized as an impor-tant, albeit imperfect, control for market power, and failure to examine the effectsof potential competition does not provide a complete picture of entry and compet-itive behaviour in US subscription television markets (Bain 1956; Clark 1902). Asnoted by Gilbert (1989), most studies of the role of potential competition in indus-trial organization focus on the magnitude of entry/exit and the persistence of profits.However, what is needed is a better understanding of the extent to which price andnon-price behaviour of established firms is conditioned on the threat of entry. RecentUS deregulation provides an opportunity to empirically assess the strategic behav-iour of incumbent cable operators to deter entry and protect their established market.Casual observation of markets in the new “liberal” environment suggests the threat

PRICE, PROGRAMMING AND POTENTIAL COMPETITION 27

of entry may force incumbents to reduce subscription prices directly, and/or indi-rectly, by adjusting quality through changes in channel programming. Such behav-iour may be revealed by regression analysis of the effect of potential competition oncable operator’s price and channels, respectively.

This paper examines potential competition for 385 US cable television marketsduring the mid to late 1990s. By contrast to prior studies, which approximatepotential competition with measures of actual competition (typically, market con-centration), we use the probability of market entry as a proxy for potential com-petition.1 A probit model is estimated where the BSP wireline overbuilder and/orILEC’s decision to enter a cable market reflects underlying cost, demographic,and industry structure variables. Predicted entry probabilities are included in asupply-demand model to permit empirical investigation of the effect of potentialcompetition on monopoly operator’s pricing and channel programming decisions.Simultaneous equations estimation of the supply-demand model shows that incum-bent cable operators offer more channels to consumers in markets facing greaterpotential competition from BSP wireline overbuilders and/or ILEC’s. In particular,when the probability of entry rises to about 42%, the average cable system providessix more channels and price per channel declines from US$ 0.77 to US$ 0.66.

The paper is organised as follows. An industry overview of US multichannelsubscription television markets, regulation and policy is provided in section 2. Sec-tion 3 outlines the empirical approach for testing the effect of potential compe-tition on cable pricing and programming. Data used for estimation are describedin section 4, while estimation results are discussed in section 5. Section 6 containsconcluding remarks and discusses future avenues of research.

2. Industry Overview

The cable business began in 1948 as a retransmission service (i.e., no originalprogramming was provided), initially delivering 1–4 retransmitted over-the-air TVstations via wire to subscribers who paid a monthly fee (Baldwin and McVoy, 1988).

Because cable systems developed as local franchised monopolists, cable oper-ators rarely faced competition from other multichannel subscription televisiondistributors.2 When cable was a local/regional broadcast television reception andretransmission business, the number of local over-the-air television stations that sub-scribers could receive had a significant impact on local system penetration. However,

1 Many studies of airline behaviour conclude that the threat from potential competitors does notpromote much price competition. Such conclusions are typically drawn from regression analyseswhich reveal a positive correlation between prices and market (and airport) concentration—a proxyfor potential competition (See Reiss and Spiller, 1989).

2 A small number of traditional cable overbuilders have had an impact on incumbent cable operatorsin a limited number of local communities. However, the economics of laying cable down the samestreet as the incumbent cable provider to compete as an overbuilder continue to be difficult (SeeNoam, 1985; Owen and Greenhalgh, 1986; Johnson, 1994; Rizzuto and Wirth, 1998; Neel, 2002).


with the advent of satellite delivered basic and premium cable networks, consumerdemand for cable was no longer strictly tied to the availability (or lack thereof ) ofover-the-air TV stations in a market. If consumers “wanted their MTV,” there wasusually only one place to get it—their local cable company. As a result, local cableoperators became the “gatekeepers” for a large quantity of non-broadcast program-ming. In the absence of any meaningful competition for cable, policy makers becameincreasingly concerned with cable’s market power (Wirth, 1990).

In spite of the fact that most cable operators did not face meaningful multi-channel competition, the FCC was a strong proponent of cable deregulation dur-ing the 1980s. Part of the FCC’s reasoning was based on its belief that, as a resultof competition from broadcasters, no more than about 50% of US television house-holds would ultimately subscribe to cable. Additionally, the FCC believed thatmarket entry, by entrepreneurs using alternative distribution technologies, wouldultimately provide a significant degree of competition (FCC 1979; Parsons andFrieden 1998).3 Potential alternative multichannel delivery technologies includedDBS (also called direct to the home or DTH); MMDS, and Satellite MasterAntenna Television (SMATV or private cable).

A number of events took place during the 1990s, which have had a significantimpact on the multichannel television competitive landscape including: passage ofthe 1992 Cable Act; development of digital signal compression; the rise of the DBSindustry; and passage of the 1996 Telecommunications Act.

2.1. The 1992 Cable ActThe 1992 Cable Act established Program Access Rules, which require that all

MSO-owned satellite delivered cable program networks be made available to allmultichannel subscription television distribution competitors at a reasonable price.These rules have provided DBS and other cable competitors with affordable andguaranteed access to a multitude of national cable programming networks, signifi-cantly reducing the risk of launching services to compete with incumbent cableoperators (Shiver 1993; Baldwin et al. 1996; Southwick 1998). Without the Pro-gram Access Rules, it is highly unlikely that meaningful multichannel televisioncompetition would have developed.4

2.2. Digital Signal CompressionDramatic advances in digital signal compression technology are revolutionizing

the communications industry (Van Tassel 1995 and 2001). DirecTV (along withStanley Hubbard’s USSB), launched the first successful high powered DBS system

3 In this 1979 report, the FCC indicated, “all of the available information suggests that underforeseeable circumstances cable penetration is unlikely to exceed about 48% of the nation’stelevision households.”

4 Baldwin et al. (1996, 263) indicate that failure to gain access to cable’s key program brands (e.g.,HBO, MTV, ESPN, etc.) would leave competitors with little desirable programming to offer theirsubscribers.


in 1994 as an all-digital system allowing it to deliver four or more video chan-nels per satellite transponder for a total of 150 channels (Southwick 1998). Morerecently, cable entrepreneurs have utilized digital signal compression technology tolaunch a number of additional services such as digital cable (designed to directlycounter DBS’s competitive threat), near video on demand/video on demand, highspeed Internet access, and a variety of other interactive television services (Kang2002; Rizzuto and Wirth 2002). Over-the-air television broadcasters are in the pro-cess of utilizing digital signal compression technology to convert the over-the-airbroadcast system from analog to digital (Van Tassel, 2001). Telephone companiesare also using digital compression technology to experiment with the delivery ofmultichannel video programming over the telephone network using very high bit-rate digital subscriber line (VDSL) (2001).

2.3. DBSHughes’ launch of “DBS 1” in December 1993 was the beginning of the

first successful high-powered DBS system (Eckhouse 1993; Paul Kagan Associates1999). It took a risky up front investment of over US$ 1 billion dollars to comeinto a marketplace littered with failed attempts to enter (e.g., STC, USCI, SkyCable, etc.) and facing a highly entrenched incumbent (Shiver 1993; Owen 1999).From the outset, it was assumed in the industry that high-powered DBS wouldbecome a competitor for cable.

“The ball game appears to be not only that of home satellite purveyors draw-ing from rural and underserved suburban markets but also moving into directcompetition with the cable industry. . . The real question is whether satellitecould be a competitor to cable. Our conclusion is that it certainly can be (MoneyTalks 1994, 1).”

Satellite TV is provided throughout the US by DirecTV and EchoStar. BothDirecTV and EchoStar compete with each other, and with cable companies, at thenational level by offering service plans with the same prices and channel offeringsin all markets. As of September 2003, there were 21 million DBS subscribers in theUS (11.9 million DirecTV subscribers and 9.1 million EchoStar Dish subscribers),making DBS by far the largest multichannel television rival to cable (DTH Counts2004). Prospects for continued DBS growth appear to be good, particularly in viewof the recent merger between DirecTV and News Corporation (Farrell and Hearn2004; FCC 2004).

2.4. Telephone Companies and the 1996 Telecommunications ActAfter Judge Harold Greene allowed the Regional Bell Operating Companies

(RBOCs) to enter the information-services business on July 25 1991, the feelingwas that it would be “only a matter of time before Congress allows phone com-panies to compete directly with cable by offering television programs (Washington1991).” On July 16, 1992, the FCC decided, “to allow RBOCs to compete withcable [by carrying], but not originating, video signals (Dominello 1995, 47).” This


signalled the creation of what became known as video dialtone (VDT) service, andthe FCC eventually granted a small number of VDT applications. However, pas-sage of the Telecommunications Act of 1996 eliminated the need for telephone com-panies to act as common carriers when distributing video programming (Parsonsand Frieden 1998).5

A number of RBOCs initially moved forward to aggressively compete with cable.For example, Ameritech (now part of SBC Communications) obtained over 100cable franchises within its local telephone market area and aggressively “over-built” the incumbent cable systems. Ameritech had approximately 250,000 cablesubscribers and its systems passed about 1.5 million homes prior to its decisionto exit the multichannel television business by selling its cable systems to Wide-OpenWest (WOW) in December 2001 (Cho 2000; Neel 2002). The RBOCs havealso almost completely abandoned MMDS as a means of providing video pro-gramming. For example, Bell South decided to partner initially with EchoStar,and more recently with DirecTV, to convert all of its MMDS customers (in NewOrleans, Atlanta, Orlando, Jacksonville, and Daytona Beach) to DBS subscribers(FCC 2004, 76). Verizon (formerly Bell Atlantic) also abandoned attempts to pro-vide video programming via MMDS or cable overbuilds and opted instead to part-ner with DirecTV to deliver the video portion of its telecommunications bundle.SBC has entered into a partnership with EchoStar to provide the video portionof its bundled telecommunication package. The final RBOC, Qwest (formerly USWest) has entered into a similar agreement with both DirecTV and EchoStar (FCC2004). Qwest is also the only RBOC, which has commercially deployed VDSL(Brown 2000). Qwest is currently using VDSL to deliver multichannel televisionon an experimental basis in Phoenix, AZ, Highlands Ranch, CO and Omaha, NE(Qwest 2002). As of May 2002, Qwest has over 41,000 multichannel television sub-scribers in Phoenix (Figler 2002).

Overall, telephone companies have provided a relatively small amount of actualcompetition for cable and other multichannel subscription television providers todate (FCC 2004). VDSL technology holds promise for allowing telephone compa-nies to become significant video distribution competitors. However, such competi-tion is likely to be slow to develop in the near term because of the upgrade costsof replacing billions of copper local loops with fiber, and a dispute over what linecode to use in modulating the VDSL signal (Brown 2000).

2.5. Current State of Multichannel Television CompetitionAccording to the FCC (2004), as of June 2003 almost 70.5 million of the

94.2 million multichannel subscription television households were cable subscribersgiving cable entrepreneurs a 74.9% share of the multichannel television market-place. DBS had 20.4 million subscribers for a 21.6% market share; BSPs (i.e., allnon-ILEC wireline cable overbuilders) had 1.4 million subscribers for a 1.49%-market share; MMDS had 0.2 million subscribers for a 0.2% market share;

5 In March 1996, the FCC ended VDT (Lucido 1996).


SMATV had 1.2 million subscribers for a 1.3% share; home satellite dish had 0.5million subscribers for a 0.5% share; and open video systems had 0.06 million sub-scribers for a 0.01% market share. As of June 2003, the largest cable overbuildersinclude RCN Corp. with 462,953 subscribers, WOW with 290,000 subscribers, andKnology Holdings with 132,163 subscribers (FCC 2004). RCN, WOW and Knol-ogy are the 9th, 13th, and 24th largest MSOs, respectively. The FCC (2001b) indi-cates that the

“BSPs . . . are overbuilding existing cable systems with state-of-the-art sys-tems that offer a bundle of telecommunication services, including video, voiceand high-speed Internet access. BSPs are carefully selecting which communitiesto serve, based on factors such as favorable demographics and high populationdensity. Their strategy is to increase per subscriber revenue and decrease churn.”

Significant challenges faced by BSP wireline overbuilders are: obtaining capitalto build these capital-intensive high-end broadband systems; and competing withentrenched cable incumbents.

In sum, though there has been a limited amount of competitive BSP entry intoexisting cable markets, incumbent cable operators no longer have the multichan-nel television marketplace to themselves. Between the actual competition providedby DBS and the actual and/or potential competition provided by BSP wirelineoverbuilders and ILECs, cable operators, such as Insight Communications’ CEOMichael Willner, indicate “I feel competitive pressure in every one of my markets,whether there is a wireline competitor or not (Neel 2002, 20).”

3. Empirical Model

Some measure of the competitive pressure discussed by Insight Communications’CEO Michael Willner is obtained from a two-step empirical investigation of theeffect of potential competition on cable prices and programming. The first-stepuses probit estimation, which explains the BSP wireline overbuilder and/or ILEC’sdecision to enter a market as a function of cost, demographics, and industrystructure, to construct an index of potential competition (PCOM) from the pre-dicted probability of entry (see Appendix A). In the second-step, a supply-demandmodel is estimated that includes PCOM as an explanatory variable in incum-bent cable operators’ price and channel equations, respectively.6 Estimated param-eters for PCOM in supply and channel equations permit empirical testing ofthe effect of potential competition on cable operator’s price and programmingdecisions.

6 Several studies estimate entry in telecom markets, but do not relate entry to potential competitionand pricing (Rosston and Wimmer 2000; Zolnierek et al. 2001; Abel 2002; Prieger 2003).


Estimation of the supply-demand model follows the new empirical industrialorganization approach outlined by Bresnahan (1989). Data limitations, however,necessitate the imposition of more structure on the model with respect to firmbehaviour. Accordingly, in the spirit of Ford and Jackson (1997), Blank et al.(1998), GAO (2000) and the FCC (2002), econometric specification is based ona generalized form of the dominant firm-competitive fringe (DFCF) model. Theindustry is defined as the delivery of multichannel subscription television to house-holds. The cable operator is the dominant firm with a large share of the multichan-nel subscription television market, and DBS operators are the fringe with about 7%market share at June 1997 (FCC 1998). Consumers perceive slightly differentiatedservices between cable and DBS because of higher DBS set-up costs and differ-ences in the quality of service delivery, both in terms of reception and provisionof channels. Following Blank et al., the dominant cable operator offers a differen-tiated subscription television service relative to a few fringe DBS operators whoseprice is determined by supply-demand conditions for the infra-fringe homogenousservice. By contrast to the GAO and FCC, cable operators are assumed to choosehow many channels to provide in a basic service package, and what to charge fora package of channels, rather than deciding on a price per channel.

The DFCF assumptions above suggest equilibrium cable price, quality andquantity are determined by the solution to the dominant cable operator’s residualdemand equation, supply relation and quality relation. Profit maximizing condi-tions can be used to obtain the following general-form demand, supply and pro-gramming quality equations for the cable operator:

SUB=f (PRICE,CHAN,HOMES,AGE, INC,NWHITE,COLL,GRAD,SAT) (1)

PRICE=g(SUB,CHAN,PLANT,AREA,WAGE,DEN,YEAR, θ) (2)

CHAN=h(CAP,SUB, INC, θ) (3)

where SUB is number of “basic” (basic plus expanded basic) cable subscribers,PRICE is the monthly subscription price for a basic service package, CHAN is thetotal number of channels in the basic package, HOMES is the number of homespassed by the cable operator (a proxy for the upper limit or potential size of themarket in the short-run), AGE is cable system age, INC is average income in thesystem’s primary county of service, NWHITE is the percentage of non-white pop-ulation in the system’s primary county of service, COLL is the percentage of theover 25 population with a college education in the system’s primary county of ser-vice, GRAD is the percentage of the over 25 population with a graduate educationin the system’s primary county of service, SAT is DBS market share, PLANT ismiles of coaxial plant in use, AREA is system geographical size, WAGE is hourlywage of cable system technical employees, DEN is population density of the sys-tem’s primary county of service, YEAR is the year the “average house” was con-structed in the system’s primary county of service, CAP is system capacity, and


θ measures the conduct of cable operators with respect to price and program-ming decisions. Note that (1) is the cable operator’s residual demand function asit takes into account fringe actions through the DBS market share variable SAT(Bresnahan 1989). As such, price and number of channels will affect cable demanddirectly though PRICE and CHAN, and indirectly through competitive fringe reac-tions in SAT, which in turn affects cable demand.7

Ideally, (1)–(3) should be estimated directly. However, data constraints preventsuch estimation and intermediate specifications consistent with economic behaviourare appealing. Since data for DBS market share by cable system are not available,and the number of competitors is typically constant across markets (i.e., EchoStarand DirecTV), we hypothesize that DBS penetration varies across different cablemarkets according to:

SAT=SAT(RUR,DWELL,NSYS) (4)

where RUR is the percentage of the population living in a rural area in thecable system’s primary county of service, DWELL is the percentage of multiple(10 or more) dwelling units in the system’s primary county of service, and NSYSis the incumbent cable operator’s share of national systems. RUR and DWELLreflect how subscribers in rural areas, and areas with a higher share of multi-ple dwelling units (MDUs), view the substitutability between cable and DBS. Weassume ∂SAT/∂RUR > 0 since traditionally, there has been less line-of-sight diffi-culties and greater marketing presence of direct-to-home satellite providers in ruralareas (FCC, 2001c). Similarly, ∂SAT/∂DWELL > 0 is assumed as DBS operatorsare likely to gain higher share in markets with a high percentage of householdsin MDUs since, historically, cable penetration among incumbent cable operatorsis lower in MDUs. This is due, in part, to the desire of MDU owners to obtainpayment for granting cable MSOs access to their private property to build andsell cable service to the MDU’s occupants. As a result, MDUs represent an under-penetrated market opportunity for DBS operators.8 NSYS reflects the effective-ness of DBS operators marketing service to cable markets served by large MSOs.Chipty (1995) argues that MSOs with a higher concentration of national systemsand subscribers obtain lower cost deals from program suppliers, and supply morecable subscriptions and channels. DBS operators may be less inclined to aggres-sively market service to areas served by an incumbent MSO with a large nationalfootprint and associated strategic advantages. As such, ∂SAT/∂NSYS < 0.

7 Blank et al. (1998) use the number of firms to measure the competitive presence of the fringe inintraLATA toll markets. Since the number of DBS operators does not vary across markets, wefollow the GAO (1999) and measure competitive presence in subscription television markets withDBS market share.

8 For example, see DSS Moves into Multiple Dwelling Unit Market (1996), Wireless One andDirecTV Sign Strategic Alliance to Provide Service to Multiple Dwelling Units (MDUs) (1997).


Let cable operator conduct depend on actual competition from fringe DBSoperators, and potential competition from new cable entrants:

θ = θ(SAT,PCOM) (5)

where PCOM is potential competition from BSP wireline overbuilders and/orILECs. Substituting (4) and (5) into (1)–(3), respectively, provides a partiallyreduced structural model of demand, supply and quality suitable for estimating theeffect of potential competition on cable pricing and programming. Given linearfunctional forms, the system of equations for market i is:

SUBi = α0j +α1PRICEi +α2CHANi +α3HOMESi

+α4AGEi +α5INCi +α6NWHITEi +α7COLLi +α8GRADi

+α9RURi +α10DWELLi +α11NSYSi + ei (6)

PRICEi = δ0j + δ1SUBi + δ2CHANi + δ3PLANTi + δ4(PLANT/AREA)i

+δ5WAGEi + δ6DENi + δ7DEN2i + δ8YEARi + δ9PCOMi

+δ10RURi + δ11DWELLi + δ12NSYSi +ui (7)

CHANi = φ0j +φ1CAPi +φ2SUBi +φ3INCi +φ4PCOMi

+φ5RURi +φ6DWELLi +φ7NSYSi +νi (8)

where α, δ, and φ are parameters to be estimated, α0j , δ0j and φ0j vary by “typeof incorporation” of the cable system according to FCC classification, and e, u,and v are disturbance terms.

Theory and prior studies guide a priori expectations for demand, supply andquality arguments in (6)–(8). Consumer behaviour theory predicts α1 < 0, α2 > 0and α5 > 0. Larger markets, in terms of number of homes passed, have poten-tial for larger cable subscription so α3 > 0 is expected. Both the GAO (2000) andFCC (2002) argue that older cable systems have greater brand loyalty as house-holds are more aware of service availability and service quality. A finding of α4 >0would support this conjecture. Previous studies suggest that household preferencesfor communications vary by demographic variables such as race and education(Duffy-Deno 2001; Eisner and Waldon 2001). While the inclusion of NWHITE,COLL and GRAD in the residual demand equation account for these effects,the signs for these variables remain an empirical question. Greater DBS marketshare implies lower cable demand, so ∂SAT/∂RUR > 0 and ∂SAT/∂DWELL > 0in (4) suggest α9 < 0 and α10 < 0, respectively, and ∂SAT/∂NSYS < 0 suggestsα11 >0.

Since (7) is a supply relation, the coefficients for SUB and CHAN shouldbe positive, so δ1 > 0 and δ2 > 0. δ3 < 0 implies cable costs decrease with plantsize, although δ4 > 0 would suggest these economies moderate in cable mar-kets of “smaller” geographical size. Costs should be higher in high wage mar-kets, but lower in densely populated markets so δ5 > 0 and δ6 < 0 are expected.


Further, a finding of δ7 > 0 for DEN2 indicates the cost advantages dampen inmarkets with very high population density. YEAR controls for possible variationin subscription prices from incumbent marketing strategies that set installationcharges at, or below, cost, and recoup revenues through higher monthly sub-scription prices. To the extent that installation costs are higher in markets com-prising of older homes (due to older internal wiring, poor access to conduits,etc.), we expect the sign for δ8 to be positive. For PCOM, δ9 < 0 shows incum-bents lower their price when facing entry and potential competition to protecttheir established subscriber base.9 The DFCF model suggests that when fringesize (i.e., DBS penetration in the current analysis) increases, the dominant firm’sprofit maximizing price will decrease. Given ∂SAT/∂RUR > 0, ∂SAT/∂DWELL > 0,∂SAT/∂NSYS < 0, and assuming cable and satellite are “close” substitutes, a find-ing of δ10 <0, δ11 <0 or δ12 >0 would be consistent with this type of pricing behav-iour.

Systems with larger capacity have greater flexibility to increase the numberof channels offered so φ1 > 0 is expected in (8). As the number of subscribersincreases, cable companies may offer more channels to satisfy the programminginterests of their large, diverse, subscriber base, and φ2 >0 would support this con-jecture (FCC 2002). Higher income households are able to pay for service withmore channels, and φ3 > 0 would support cable operator’s provision of packageswith more channels in areas with relatively high income. φ4 > 0 suggests incum-bent cable providers facing a higher probability of competition from ILECs andBSP wireline overbuilders will increase channels to provide a higher quality servicethan they would if they could be sure that actual entry would not occur. A findingof φ5 > 0 or φ6 > 0 suggests cable operators provide more channels to rural sub-scribers, and to markets with a high share of multiple dwelling units, to make theirpackages more attractive than those offered by fringe DBS operators. By contrast,φ7 ≤0 would indicate there is not sufficient DBS competition in markets served bylarge MSOs to warrant the incumbent cable operator providing more channels tosubscribers.

4. Data and Variable Descriptions

To construct PCOM, and estimate (6)–(8), a random sample of 500 cable systemswas drawn without stratification from the FCC’s list of US cable communi-ties located at www.fcc.gov/mb. Data for cable operator’s cost, size, demand,programming, and industry structure are sourced from the 1999 Television andCable Factbook (Warren Publishing Inc. 1999). Census-based demographic datafor each system’s primary county of service are obtained from The Right Site

9 By offering a low price (or price per channel), the cable operator signals to potential competitorsthat it can supply low cost cable subscriptions and continue to earn economic profits.


Market Analysis Edition (Easy Analytic Software Inc. 1998). These data are col-lected from 1995 to 1999, a period that captures useful information about cableoperator behaviour following the opening up of their markets to entry from BSPwireline overbuilders, ILECs, and DBS providers. After eliminating systems withincomplete data the sample reduced to 399 communities representing 49 statesand the District of Columbia. Three hundred and eighty five systems are cable“monopolies” (i.e., they are the single cable operator in the market) and 14 havebeen awarded “competitive status” by the FCC during 1998–2001 as a result ofBSP (i.e., independent, municipal, and competitive local exchange carrier wirelineoverbuild) and/or ILEC entry. These 14 systems, located in Alabama, Colorado,Florida, Georgia, Ohio, Texas, Utah and Vermont, provide the basis for predict-ing the probability of cable entry in Appendix A, and calculation of the poten-tial competition index used in the estimation of (6)–(8) for the 385 monopoly cablesystems.

Table 1 provides selected characteristics for the 385 monopoly cable systems. Theaverage sample system is 310 months old and serves 22,308 basic subscribers. Onaverage, basic service packages have 32 channels and retail at a monthly subscrip-tion price of US$ 23.28, or US$ 0.77 per channel. Selected characteristics com-pare well with FCC (1999) data for the same sample period. The FCC survey783 systems at July 1 1997, and find the average system serves 18,383 subscriberswith 48 channels at a monthly subscription price of US$ 26.14, or US$ 0.64 perchannel. There are 58 different incumbent cable system owners in the present sam-ple, and all systems were operated by MSOs except Sumner Cable TV Companyin Wellington, Kansas. TCI owned the most national systems, 1151, and had thehighest share of national subscribers at 19.46%. The potential competition indexranges from 0 to 42.34%, and the mean value of 2.973 indicates that, on aver-age, monopoly markets faced about a 3% probability of entry from a BSP wirelineoverbuilder and/or ILEC competitor. A full description of the variables and dataused for estimating the market entry equation and the supply-demand system areprovided in Appendix B.

Table 1. Cable Sample Characteristics

Characteristics Mean SD Min Max

Number of subscribers 22,308 41,665 155 447,909Monthly subscription price (US$) 23.28 5.900 7.934 65.28Number of channels 32.07 9.567 11 65Subscription price per channel (US$) 0.769 0.244 0.252 2.522System age (months) 309.5 113.9 1 582National systems for each incumbent 524.5 481.4 1 1151Incumbent’s percent of national subscribers 9.091 8.766 0.001 19.46Potential competition index (percentage) 2.973 6.150 3E−86 42.34

Notes. n=385. SD, standard deviation; Min, minimum value; Max, maximum value.


5. Estimation Results

Table 2 reports three stage least squares (3SLS) estimates of the supply-demandsystem for 385 cable markets. Since PCOM is a predicted value from first-step pro-bit estimation, its inclusion in the system results in biased estimates of coefficientvariances and smaller standard errors. Accordingly, corrected standard errors arederived from Maddala’s (1983) covariance matrices of two-step probit methods forsimultaneous equations models.

Price (PRICE), number of channels (CHAN), market size (HOMES) and systemage (AGE) are important determinants of basic cable demand. The price elastic-ity of demand, calculated at the mean, is −1.90 and reasonably similar to priorstudies.10 Demand is relatively inelastic with respect to the number of channelsoffered, 0.72, and is noticeably smaller than the GAO (2000) estimate of 3.22and the FCC (2002) estimate of 3.67. As expected, demand increases with thenumber of homes passed (HOMES) and the age of the system (AGE), with thelatter finding supporting the GAO and FCC conjecture that demand is higherin older cable communities because consumers are more aware of the availabil-ity and quality of service. Model estimates also suggest cable demand is higherin markets with higher household income (INC), but there is no significant rela-tionship between cable demand and the percentage of the non-white population(NWHITE), and education (COLL and GRAD). Coefficient estimates for quantityof subscribers (SUB) and number of channels (CHAN) in the supply equationare both positive and significant. Estimated coefficients for several cost proxiesin the price equation conform to expectations. Prices decrease with plant size,but decreases are moderated in systems of smaller geographical size.11 Further,prices increase with the hourly wage of telecom equipment installers and repairers(WAGE), and decrease with population density (DEN). Quality equation esti-mates show that system capacity (CAP) and SUB are positive determinants of thenumber of channels provided by cable operators. The latter results being consis-tent with FCC findings that cable operators provide more channels as the num-ber of subscribers increases to satisfy the diverse interests of a larger subscriberbase.

The signs for potential competition in (7) and (8) provide an empirical test ofthe effect of potential competition from BSP wireline overbuilders and/or ILECson prices and programming in multichannel subscription TV markets. Potentialcompetition has no direct effect on incumbent’s prices (PRICE). However, aftercontrolling for demand, prices and (indirectly) DBS market share, a positive sign

10 Mayo and Otsuka (1991) estimate −0.69 to −1.51, Rubinovitz (1993) −1.46, Chipty (1995) −1.05to −2.34, the FCC (2002) −2.19, GAO (2000) −3.22, and Goolsbee and Petrin (2004) −0.43 to−0.55.

11 For instance, cable markets with the minimum AREA of 27.2 miles2 have∂PRICE/∂PLANT =−US$ 0.001 and markets with the maximum AREA of 8162 miles2 have∂PRICE/∂PLANT =−US$ 0.329.


Table 2. Cable Operator Supply-Demand Estimates

Residual Demand Supply Relation Quality RelationSUB (6) PRICE (7) CHAN (8)

Variable Coefficient Coefficient Coefficient

SUB 0.00004b 0.00004a

(0.00002) (0.00001)PRICE −1,821.3a

(477.79)CHAN 501.44a 0.2780a

(189.93) (0.0825)CAP 0.2202a

(0.0262)HOMES 0.6635a

(0.0113)AGE 28.363a

(5.6204)INC 1686.8a 0.4623

(755.04) (0.4398)NWHITE 796.67

(3171.8)COLL 20,107

(16,948)GRAD 19,704

(27,717)PLANT −0.0033a

(0.0015)PLANT/AREA 0.0894b

(0.0519)WAGE 0.2766a

(0.1257)DEN −2.4690a

(1.2368)DEN2 0.3294

(0.2006)YEAR −0.2466

(0.1512)PCOM 0.9411 13.978a

(5.4692) (6.8444)RUR 2112.2 0.4352 1.0246

(3181.3) (1.6811) (2.1340)DWELL −71,005a −4.2489 33.632a

(17,848) (10.563) (11.133)NSYS 7408.1a 1.3237 1.4803

(2536.0) (1.2768) (1.6698)INTERCEPT 6071.3 11.584 12.012

(12,314) (7.0692) (7.9073)

Notes. n = 385. Standard errors, in parentheses, are derived from Maddala’s (1983) covariance matrices oftwo-step probit methods for simultaneous equations models.a Significant at the five percent level.b Significant at the ten percent level.


for PCOM in the quality equation suggests that incumbents, facing greater poten-tial competition, will improve programming quality by providing more channelsfor a given monthly subscription price. By reducing the price per channel, theincumbent may pre-empt entry by signalling it can supply subscriptions at lowercost while maintaining profitability. Some insight into how this strategic behaviouraffects the price per channel faced by subscribers is obtained by comparing thenumber of channels offered to subscribers in markets with the average level ofpotential competition with the number of channels offered in markets with thehighest level of potential competition. An increase in potential competition from2.973% to 42.34% would see average subscribers offered about six more channelsat the higher monthly price of US$ 24.81, with price per channel declining fromUS$ 0.77 to US$ 0.66. Variables that indirectly determine satellite operator mar-ket share, RUR, DWELL and NSYS, have mixed impact on cable demand, pric-ing and channels. Cable demand is lower in areas with a high share of multipledwelling units, and subscribers in these areas may view DBS as a close substi-tute to cable. Accordingly, cable operators provide more channels at a lower priceper channel in response to actual DBS competition. By contrast, demand is rela-tively high in systems served by large MSOs, and satellite does not appear to bea close substitute in these markets during the period of study. There is no signifi-cant relationship between NSYS and cable prices, and NSYS and the number ofchannels.

6. Conclusion

Recent deregulation of the US cable and telecom sectors seeks to open up thesetraditional monopoly markets to entry and encourage competition, both withinand across sectors. While the GAO (2000) and FCC (2002) show that non-pricecompetitive behaviour, particularly increases in the number of channels in servicepackages, is a key feature of contemporary multichannel subscription televisionmarkets, actual competition from BSP wireline overbuilders and ILEC’s has rolledout very slowly. Because most cable markets continue to be served by a monopolywireline provider, or at best, are highly concentrated around the incumbent, analy-sis of the effect of actual entry does not provide a complete picture of the state ofcompetition in the cable industry as a whole. With recent market deregulation it ispossible that the threat of entry from a potential competitor may force incumbentsto reduce subscription prices directly, and/or indirectly, by adjusting programmingquality through changes in the number of channels provided.

This paper provides a two-step empirical investigation of the effect of potentialcompetition on cable prices and programming. First-step probit estimates indicatenew BSP wireline overbuild and ILEC entrants are more likely to enter monopolymarkets with high population density, income, and household growth. Entry is lesslikely in regulated city markets, and markets served by an incumbent MSO witha large base of national subscribers. These estimates can be used to profile howcable industry competition is developing in the evolving liberal environment, and


to make predictions about where BSP overbuild and ILEC market entry is likely toarise in the future. Predicted entry probabilities are included in a supply-demandmodel to permit empirical investigation of the effect of potential competition onincumbent cable operator’s prices and channel programming decisions. Second-stepestimation of the supply-demand system shows incumbent cable operators offermore channels to consumers in markets facing greater potential competition fromBSP wireline overbuilders and/or ILEC’s. In particular, when the probability ofentry rises to about 42%, the average cable system provides six more channels, andprice per channel declines from US$ 0.77 to US$ 0.66.

Limitations of the current study include: the lack of system-by-system DBSsubscriber data; the need for better control variables to measure horizontal andvertical integration; and the use of list price (as opposed to actual) data asobtained from the 1999 Television and Cable Factbook. Future work in this areashould obtain and incorporate system-by-system DBS subscriber data, if possible,to directly control for and measure the impact of DBS on the competitive decisionmaking of local cable systems. Additionally, use of more recent data would allowfor the incorporation of cable operators’ new revenue stream strategies (i.e., highspeed Internet access, residential telephony, and digital cable services) to evaluatetheir response to actual and potential competitive entry by multichannel competi-tors.

Appendix A: Market Entry and Potential Competition

Theory suggests firms enter markets when they expect to earn above normal profit.As such, a new cable operator’s marginal entry decision is based on the expectedprofits (π ) from entering cable market i:

πi =β ′xi + εi (A.1)

where β is a vector of coefficients, x is a vector of exogenous variables whichaffect the expected revenues and costs of operation, and εi is a disturbance term.While the expected profitability of the new entrant is not observable, it is pos-sible to observe whether a new entrant has entered (ENTRY = 1) or not entered(ENTRY = 0) market i:

πi =1 when π∗i >0, and πi =0 when π∗

i ≤0 (A.2)

where π∗i is unobserved profit from market entry, and β ′ xi is an index function.

Given a symmetric distribution, the underlying empirical model for the probability(P ) of entry is:

P(π∗i >0)=P(εi <β ′xi )=F(β ′xi ) (A.3)

where multichannel competitive entry has occurred leading the FCC to award amarket “competitive status”, and F (·) is the normal distribution. Exogenous vari-ables included in x are: growth rate of households in the cable systems’ primary


county of service (HHG); income in the systems’ county (INC); population den-sity of the system’s county (DEN); hourly wage of cable system technical employ-ees (WAGE); system age (AGE); an indicator of whether the cable market is inan incorporated city (CITY); incumbent’s share of national subscribers (NSUB);incumbent’s share of national systems (NSYS); number of systems owned by theincumbent in the state corresponding to market i (SYS); and the age distributionfor the population in the systems’ county (POP).

Variables in x are selected to approximate the new entrant’s determination ofprice, quantity, and profits. HHG and INC measure expected subscriber demandafter entry. Since demand should be higher in markets with higher householdgrowth and larger income, a positive relationship is expected between HHGand probability of entry, and INC and probability of entry. DEN and WAGEapproximate cost variables associated with network deployment and operation.New entrant costs should be lower in markets with high population densityand low wage rates for cable employees. As such, a positive sign is expectedfor DEN and a negative sign for WAGE. AGE, NSUB, NSYS, and SYS arestrategic variables that approximate any competitive advantages held by theincumbent cable operator. Older cable systems with less sophisticated infrastruc-ture and service delivery may attract entry from BSP wireline overbuilders andILECS with more efficient technologies and marketing strategies, and as such,a positive relationship is expected between AGE and the probability of entry.Incorporated cities typically have more local rules and regulations that makeit relatively harder for new entrants to obtain and operate a cable franchise.Accordingly a negative relationship is expected between CITY and ENTRY. Thebargaining power hypothesis implies a cost advantage to incumbents with largenational presence, and new entrants are less likely to enter these markets. Con-trolling for regional size (SYS), a negative sign on NSUB and/or NSYS maybe interpreted as evidence of bargaining power. SYS and POP are controls forregional size and subscriber age profile. Their signs remain an empirical ques-tion.

Probit estimates of the entry equation (A.3) for 399 markets are provided inTable A1. BSP wireline overbuilders and ILECs are more likely to enter marketswith high household growth, income, and population density. The quadratic termsINC2 and DEN2 indicate the probability of entry moderates at higher levels ofhousehold income and density. Entry is also likely in markets where the percentageof the over 18 population (POP) is relatively high. The positive sign for AGE sug-gests that new entrants are attracted to markets where the incumbent has a systemof older vintage. Here, BSP wireline overbuilders and ILECS may have a relativeadvantage by deploying new infrastructure and technology, and/or by providingservice to subscribers that are well aware of the availability and quality of ser-vice provided through cable TV. Entry is less likely in high-wage markets, marketswhere the incumbent MSO has a large base of national subscribers, and incor-porated cities. The latter findings suggest bargaining power advantages and exces-sive city rules and regulations, may dissuade potential new operators from enteringcable markets.


Table A1. Probit Estimates of Market Entry Equation

Variable Coefficient Standard Error

HHG 0.0330a 0.0131INC 11.958a 4.7241INC2 −1.6544a 0.6645DEN 3.8469a 1.6360DEN2 −1.8677b 0.9750WAGE −0.1260 0.0785AGE 0.0048a 0.0019CITY −0.9672a 0.3737NSUB −0.0422b 0.0258NSYS −1.0006 0.8439SYS 0.0071 0.0063POP 10.409b 5.3230INTERCEPT −24.940a 8.7894Log likelihood −41.053

Notes. n=399. Dependent variable is ENTRY.a Significant at 5% level.b Significant at 10% level.

Appendix B: Variable Descriptions

Table B1. Variables in Market Entry Equation

Variable Description and Data Source Mean

ENTRY Equals one when the community has BSP wireline overbuild and/or 0.0316ILEC entry during 1998–2001, and zero otherwise. Source: FCC(2001d).

HHG Projected household growth during 1990s. Source: Easy Analytic 0.1161Software Inc. (1998).

INC Median household income ($) (104). Source: Easy Analytic Software 3.4918Inc. (1998)

DEN Persons per square mile (103). Source: Easy Analytic Software Inc. (1998). 0.4156WAGE Hourly wage ($) for telecom equipment installers and repairers. Source: 19.950

Bureau of Labor Statistics (1999).AGE System age in months. Source: Warren Publishing Inc. (1999). 310.01CITY Equals one when the community is an incorporated city, and zero 0.7143

otherwise. Source: FCC (2001d).NSUB Incumbent’s percentage of national subscribers. Source: Warren 9.04

Publishing Inc. (1999).NSYS Incumbent’s percentage of national systems. Source: Warren 17.96

Publishing Inc. (1999).SYS Number of systems owned by the incumbent in the state corresponding 33.582

to market i. Source: Warren Publishing Inc. (1999).POP Percentage of population aged under 18 years. Source: Easy Analytic 26.32

Software Inc. (1998).


Table B2. Variables in Supply-Demand Model

Variable Description and Data Source Mean

SUB Number of basic and expanded basic subscribers to the system. Source: 22,308Warren Publishing Inc. (1999).

PRICE Monthly subscription price ($). Source: Warren Publishing Inc. (1999). 23.28CHAN Number of channels. Source: Warren Publishing Inc. (1999). 32.07CAP Number of 6-MHz channels a system can carry. Source: Warren Publishing 52.26

Inc. (1999).HOMES Number of homes passed by cable operator. Source: Warren Publishing Inc. 33,277

(1999).AGE System age in months. Source: Warren Publishing Inc. (1999). 309.52INC Median household income ($) (104). Source: Easy Analytic Software Inc. 3.4941

(1998).NWHITE Percentage of non-white population. Source: Easy Analytic Software 16.72

Inc. (1998).COLL Percentage of population over 25 years of age with college education. Source: 14.01

Easy Analytic Software Inc. (1998).GRAD Percentage of population over 25 years of age with post-graduate education. 6.41

Source: Easy Analytic Software Inc. (1998).PLANT Miles of coaxial plant in use by the cable operator. Source: Warren Publishing 398.12

Inc. (1999).AREA Square miles of the county corresponding to cable system i. Source: Easy 979.47

Analytic Software Inc. (1998).WAGE Hourly wage ($) for telecom equipment installers and repairers. Source: 19.97

Bureau of Labor Statistics (1999).DEN Population density (103). Source: Easy Analytic Software Inc. (1998). 0.4195YEAR Age in years of the “average house”. Source: Easy Analytic Software Inc. 17.11

(1998).PCOM Predicted probability of entry from probit estimation of (A1). 2.973RUR Percentage of population in rural areas. Source: Easy Analytic Software Inc. 44.56

(1998).DWELL Percentage of households in multiple (10 or more) dwelling units. Source: 4.10

Easy Analytic Software Inc. (1998).NSYS Incumbent’s percentage of national systems. Source: Warren Publishing Inc. 17.96

(1999).

References

Abel, J. 2002. “Entry in Regulated Monopoly Markets: The Development of a Competitive Fringe inthe Local Telephone Industry.” Journal of Law & Economics 45: 289–316.

Bain, J. 1956. Barriers to New Competition. Cambridge, MA: Harvard University Press.Baldwin, T., and D. McVoy. 1988. Cable Communication, second edition. Englewood Cliffs, NJ: Prentice

Hall.Baldwin, T., D. McVoy, D. Stevens, and C. Steinfield 1996. Convergence: Integrating Media, Information

& Communication. Thousand Oaks, CA: Sage Publications.Beard, T., and G. Ford. 1999. “Competition Between Wireline Networks as Fragmented Duopoly with

an Empirical Application to the Cable Television Industry.” Phoenix Center for Advanced Legal andEconomic Public Policy Studies.


Blank, L., D. Kaserman, and J. Mayo. 1998. “Dominant Firm Pricing with Competitive Entry and Reg-ulation: The Case of IntraLATA Toll.” Journal of Regulatory Economics 14: 35–53.

Bresnahan, T. 1989. “Empirical Studies of Industries with Market Power.” In Handbook of IndustrialOrganization edited by R. Schmalensee and R. Willig. Amsterdam: Elsevier Science Publishers.

Brown, K. 2000. “Telcos Breathe Life into VDSL.” CableWorld May 8: 61–2.Bureau of Labor Statistics. 1999. “Occupational Employment Statistics.” Retrieved June 19, 2002 from

http://stats.bls.gov/oes/1999/oessrcst.htm.Chipty, T. 1995. “Horizontal Integration for Bargaining Power: Evidence from the Cable Television

Industry.” Journal of Economics and Management Strategy 4: 375–397.Cho, J. 2000. “Ameritech Left Wondering?” CableWorld May 30. Retrieved July 17, 2001 from

http://www.cableworld.com.Clark, J. 1902. The Control of Trusts New York: Macmillan.Dominello, J. 1995. “Video Dialtone.” In Communication Technology Update, fourth edition, edited by

A. Grant. Boston, MA: Focal Press.DSS Moves into Multiple Dwelling Unit Market. 1996. Satellite Week. January 15. Retrieved March 6,

2004 from http://www.lexis-nexis.com/universe.DTH Counts. 2004. Sky Research. February.Duffy-Deno, K. 2001. “Demand for Additional Telephone Lines: An Empirical Note.” Information Eco-

nomics and Policy 13: 283–299.Easy Analytic Software, Inc. 1998. The Right Site Market Analysis Edition. Vineland, NJ: Easy Analytic

Software, Inc.Eckhouse, J. 1993. “Satellites Threaten Cable TV Operators.” The San Francisco Chronicle. December

16: 1. Retrieved September 14, 2001 from http://www.lexis-nexis.com.Eisner, J., and T. Waldon. 2001. “The Demand for Bandwidth: Second Telephone Lines and On-line

Services.” Information Economics and Policy 13: 301–309.Farrell, M., and T. Hearn. 2004. “Murdoch gets DirecTV.” Multichannel News January 5: 1.FCC. 1979. Inquiry Into The Economic Relationship Between Television Broadcasting And Cable Televi-

sion [71 FCC 2d 632]. Washington, DC: FCC.FCC. 1994. “1994 Report” [FCC Rcd at 228–245]. Washington, DC: FCC.FCC. 1998. “Fourth Annual Report: In the Matter of Annual Assessment of the Status of Competition

in the Market for the Delivery of Video Programming” [MB Docket No. 02–145]. Washington, DC:FCC.

FCC. 1999. “Report on Cable Industry Prices” [MM Docket No. 92–266] Washington, DC: FCC.FCC. 2001a. “Seventh Annual Report: In the Matter of Annual Assessment of the Status of Competi-

tion in the Market for the Delivery of Video Programming” [CS Docket No. 00–132]. Retrieved Jan-uary 18, 2001 from http://ftp.fcc.gov/Bureaus/Cable/Reports/fcc01001.pdf.

FCC. 2001b. “Eighth Annual Report: In the Matter of Annual Assessment of the Status of Competitionin the Market for the Delivery of Video Programming” [CS Docket No. 01–129]. Retrieved January21, 2002 from http://www.fcc.gov/mb/csrptpg.html.

FCC. 2001c. “Report on Cable Industry Prices” [MM Docket No. 92–266]. Washington, DC: FCC.FCC. 2001d. “Spreadsheet Describing Competitive Status of Cable Franchises 1998–2001.” Personal

correspondence with FCC.FCC. 2002. “Report on Cable Industry Prices” [MM Docket No. 92–266]. Washington, DC: FCC.FCC. 2004. “Tenth Annual Report: In the Matter of Annual Assessment of the Status of Competition

in the Market for the Delivery of Video Programming” [MB Docket No. 03–172]. Retrieved January19, 2004 from http://www.fcc.gov/mb/.

Figler, A. 2002. “Cox Aims to Get Back on Track in Phoenix.” CableWorld August 5: 36.Ford. G., and J. Jackson. 1997. “Horizontal Concentration and Vertical Integration in the Cable Televi-

sion Industry.” Review of Industrial Organization 12: 501–518.GAO. 1999. The Changing Status of Competition to Cable Television. Washington, DC: GAO.GAO. 2000. The Effect of Competition from Satellite Providers on Cable Rates. Washington, DC: GAO.Gilbert, R. 1989. “The Role of Potential Competition in Industrial Organization.” Journal of Economic

Perspectives 3: 107–127.


Goolsbee, A., and A. Petrin. 2004. “The Consumer Gains from Direct Broadcast Satellites and theCompetition with Cable TV.” Econometrica 72: 351–382.

Johnson, L. 1994. Toward Competition in Cable Television. Cambridge, MA: The MIT Press/Washington,DC: American Enterprise Institute.

Kang, Myung-Hyun. 2002. “Digital Cable: Exploring Factors Associated with Early Adoption.” Journalof Media Economics 15: 193–207.

Lucido, B. 1996. “Switched Video Services.” In Communication Technology Update, fifth edition, editedby A. Grant. Boston, MA: Focal Press.

Maddala, G. 1983. Limited Dependent and Qualitative Variables in Econometrics. New York: CambridgeUniversity Press.

Mayo, J., and Y. Otsuka. 1991. “Demand, Pricing, and Regulation: Evidence form the Cable TV Indus-try.” Rand Journal of Economics 22: 396–410.

Money Talks. 1994. “SkyReport”. first Quarter.Neel, K. 2002. “Deadend at the Headend.” CableWorld March 18: 17–20.Noam, E. 1985. “Economies of Scale in Cable Television: A Multiproduct Analysis.” In Video Media

Competition: Regulation, Economics and Technology, edited by Noam, E. New York: Columbia Uni-versity Press.

Owen, B. 1999. The Internet Challenge to Television. Cambridge, MA: Harvard University Press.Owen, B., and P. Greenhalgh. 1986. “Competitive Policy Considerations in Cable Television Franchis-

ing.” Contemporary Policy Issues 4: 69–79.Owen, B., and S. Wildman. 1992. Video Economics. Cambridge, MA: Harvard University Press.Parsons, P., and R. Frieden. 1998. The Cable and Satellite Television Industries. Boston, MA: Allyn and

Bacon.Paul Kagan Associates. 1999. The State of DBS 2000. Carmel, CA: Paul Kagan Associates.Prieger, J. 2003. “The Supply Side of the Digital Divide: Is There Equal Availability in the Broadband

Internet Access Market?” Economic Inquiry 41: 346–363.Qwest. 2002. “Qwest’s Current VDSL Markets.” Retrieved August 23, 2002 from http://www.qwest.com/

vdsl/learn/markets.html.Reiss, P., and P. Spiller. 1989. “Competition and Entry in Small Airline Markets.” Journal of Law and

Economics 32, 179–202.Rizzuto, R., and M. Wirth. 1998. Costs, Benefits, and Long-Term Sustainability of Municipal Cable Tele-

vision Overbuilds. Denver, CO: GSA Press.Rizzuto, R., and M. Wirth. 2002. “The Economics of Video on Demand: A Simulation Analysis.” Jour-

nal of Media Economics 15: 209–25.Rosston, G., and B. Wimmer. 2000. “From C to Shining C: Competition and Cross-Subsidy in Com-

munications.” SIEPR Discussion Paper No. 00–21. Stanford: Stanford Institute for Economic PolicyResearch.

Rubinovitz, R. 1993. “Market Power and Price Increases for Basic Cable Since Deregulation.” RandJournal of Economics 24: 1–18.

Shiver, J. Jr. 1993. “Beating Swords into Sitcoms? Defense Giant Hughes Takes on Cable with SatelliteTV.” Los Angeles Times November 21: 1. Retrieved July 17, 2001 from http://www.lexis-nexis.com.

Southwick, T. 1998. Distant Signals. Overland Park, KS: Primedia Intertec.Van Tassel, J. 1995. “Digital Video Compression.” In Communication Technology Update, fourth edition,

edited by A. Grant. Boston, MA: Focal Press.Van Tassel, J. 2001. Digital TV over Broadband: Harvesting Bandwidth. Boston, MA: Focal Press.Warren Publishing Inc. 1999. 1999 Television and Cable Fact Book. Washington, D.C.: Warren Publish-

ing Inc.Washington, M. 1991. “Why Cable Companies Are Playing So Rough.” Business Week August 12: 66.

Retrieved September 14, 2001 from http://www.lexis-nexis.com.Wireless One and DirecTV Sign Strategic Alliance to Provide Service to Multiple Dwelling Units

(MDUs). 1997. PR Newswire September 4. Retrieved March 6, 2004 from http://www.lexis-nexis.com/universe.

Wirth, M. 1990. “Cable’s Economic Impact on Over-the-Air Television Broadcasting.” Journal of MediaEconomics 3: 39–53.


Zolnierek, J., J. Eisner, and E. Burton. 2001. “An Empirical Examination of Entry Patterns in LocalTelephone Markets.” Journal of Regulatory Economics 19: 143–159.

Price, Programming and Potential Competition in US Cable Television Markets

Documents