<%BANNER%>

Record for a UF thesis. Title & abstract won't display until thesis is accessible after 2011-12-31.

Permanent Link: http://ufdc.ufl.edu/UFE0024874/00001

Material Information

Title: Record for a UF thesis. Title & abstract won't display until thesis is accessible after 2011-12-31.
Physical Description: Book
Language: english
Creator: Lee, Seonmi
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: Journalism and Communications -- Dissertations, Academic -- UF
Genre: Mass Communication thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Statement of Responsibility: by Seonmi Lee.
Thesis: Thesis (Ph.D.)--University of Florida, 2009.
Local: Adviser: Chan-Olmsted, Sylvia M.
Electronic Access: INACCESSIBLE UNTIL 2011-12-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0024874:00001

Permanent Link: http://ufdc.ufl.edu/UFE0024874/00001

Material Information

Title: Record for a UF thesis. Title & abstract won't display until thesis is accessible after 2011-12-31.
Physical Description: Book
Language: english
Creator: Lee, Seonmi
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: Journalism and Communications -- Dissertations, Academic -- UF
Genre: Mass Communication thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Statement of Responsibility: by Seonmi Lee.
Thesis: Thesis (Ph.D.)--University of Florida, 2009.
Local: Adviser: Chan-Olmsted, Sylvia M.
Electronic Access: INACCESSIBLE UNTIL 2011-12-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0024874:00001


This item has the following downloads:


Full Text

PAGE 1

1 TRIPLE-PLAY STRATEGY IN THE CA BLE AND TELECOMMUNICATIONS INDUSTRIES: AN EMPIRICAL STUDY OF ITS USE AND IMPACT ON THE US COMMUNICATIONS MARKET By SEONMI LEE A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2009

PAGE 2

2 2009 Seonmi Lee

PAGE 3

3 To my parents

PAGE 4

4 ACKNOWLEDGMENTS First, I thank God for his unconditional love for m e. Without his love I may not complete my study and dissertation in the doctoral progra m. Second, I am sincerely grateful to my dissertation chair, Dr. Sylvia Chan-Olmsted. Sh e supported my study during last four years while I am studying in the telecommunication policy, management, and media economics field at University of Florida. Without her support, help s, and advices, I may not complete this long journey of acquiring the Ph.D. Third, I thank Dr. David Ostroff. He gave me theoretical approaches about new media policy and new media system, which contribute s to my dissertation. Fourth, I appreciate Dr. Marilyn Roberts for her cares. Also, I si ncerely appreciate Dr. Sanford Berg that he gave me insightful comments and id eas, which were very infl uential for dissertation. He also introduced many experts in the field of Economics, which made it possible to write better dissertation. I also truly thank those who always helped me, especially Jae Yoo, Sangwon Lee, and Bumsub Jin. Finally I thank my parents and my brothe r, for their support and encouragement.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ............................................................................................................... 4LIST OF TABLES ...........................................................................................................................8LIST OF FIGURES .......................................................................................................................10ABSTRACT ...................................................................................................................... .............11CHAPTER 1 INTRODUCTION ..................................................................................................................13The Triple-Play Strategy and Its Impact on the Market .........................................................13Significance of the Study ........................................................................................................18Purpose of the Study .......................................................................................................... .....212 LITERATURE REVIEW .......................................................................................................24Two Perspectives for Analyses of the Triple-Play Strategy ...................................................24The Industrial Organizat ion (IO) Approach ....................................................................24The Resource-Based (RBV) View .................................................................................. 25Factors Contributing to the Use of Triple-Play ...................................................................... 26The Impact of Competition on the Use of Triple-Play .................................................... 26The Impact of Information Communication Technology on the Use of Triple-Play ...... 29The Impact of Market Potentia l on the Use of Triple-Play ............................................. 31The Impact of Firm Size on the Use of Triple-Play ........................................................ 32The Impact of Cost on the Use of Triple-Play ................................................................ 33Strategic Orientation ........................................................................................................35The Impact of Triple-Play Strategy on Tr iple-Play Providers Market Performance ............37The Impact of Bundling on Market Performance ............................................................ 37The Impact of Switching Costs on Market Performance ................................................ 40The Impact of Competition on Market Performance ...................................................... 42The Impact of Information Communicat ion Technology on Market Performance ........ 45The Impact of Market Potential on Market Performance ................................................ 46The Impact of Firm Size on Market Performance ........................................................... 47The Impact of Cost on Market Performance ................................................................... 48Resource-Based View (RBV) of Strategy ....................................................................... 48The Impact of a Triple-Play Bundle on Market Entry ............................................................ 50The Impact of Bundling on Market Entry .......................................................................51The Impact of Competition on Market Entry .................................................................. 52The Impact of Local Loop Unbundling Regulation on Market Entry ............................. 54The Impact of Information Commun ication Technology on Market Entry .................... 55The Impact of Network Effects on Market Entry ............................................................ 56The Impact of Market Potential on Market Entry ........................................................... 58

PAGE 6

6 3 RESEARCH MODELS ..........................................................................................................60Research Model I ....................................................................................................................60Market Factors .................................................................................................................61Firm Factors .................................................................................................................. ...62Strategic Orientation ........................................................................................................63Research Model II ...................................................................................................................63Strategic Factors ..............................................................................................................65Market Factors .................................................................................................................65Firm Factors .................................................................................................................. ...66Resource-Based View (RBV) of Strategy ....................................................................... 67Research Model III .................................................................................................................67Rationale for the Analysis of Hi gh-Speed Fixed Broadband Market .............................. 68Potential Determinants of Market Entry .......................................................................... 704 METHODS ....................................................................................................................... ......73Secondary Analysis ............................................................................................................ ....73Panel Data Analysis ........................................................................................................... .....73Sampling ...................................................................................................................... ...........74Measurement, Data, and Statis tical Methods for Model I ...................................................... 75Measurement and Data Sources ......................................................................................75Statistical Methods ..........................................................................................................78Measurement, Data, and Statis tical Methods for Model II ..................................................... 81Measurement and Data Sources ......................................................................................84Statistical Methods ..........................................................................................................85Measurement, Data, and Statis tical Methods for Model III ................................................... 87Measurement and Data Sources ......................................................................................87Statistical Methods ..........................................................................................................895 RESULT ........................................................................................................................ .........91The Use of Triple-Play ........................................................................................................ ...91Data and Descriptive Statistics for Model I ....................................................................91Regression Analysis of the Use of Triple-Play ............................................................... 93Market Performance of the Triple-Play Providers ................................................................109Data and Descriptive Statistics for Model II ................................................................. 109Regression Analysis of the Impact of Triple-Play on Market Performance ................. 113Analysis of Variance (ANOVA) Results of the Impact of Triple-Play on Market Performance ............................................................................................................... 135Analysis of Variance (ANOVA) Results of the Impact of Triple-Play on Market Performance ............................................................................................................... 139The Impact of Triple-Play on Market Entry ......................................................................... 152Data and Descriptive Statistics for Model III ................................................................152Regression Analysis of the Impact of Triple-Play on Market Entry ............................. 1536 DISCUSSION AND CONCLUSION .................................................................................. 156

PAGE 7

7 Summary Results and Analyses ............................................................................................ 156The Drivers of the Use of Triple-Play ........................................................................... 156The Impact of Triple-Play on Market Perf ormance of the Triple-Play Providers ........ 168Resource-Based View (RBV) of Strategy ..................................................................... 179The Impact of Triple-Play on Market Entry in the High-Speed Fixed Broadband Market ........................................................................................................................ 181Implications .................................................................................................................. ........183Theoretical Implications ................................................................................................ 183Strategic Implications .................................................................................................... 187Policy Implications ........................................................................................................189Limitations and Suggestions for Future Research ................................................................ 191LIST OF REFERENCES .............................................................................................................194BIOGRAPHICAL SKETCH .......................................................................................................207

PAGE 8

8 LIST OF TABLES Table page 1-1 Growth trends of the top triple -play providers in the US (2002-2007) ...................................161-2 Growth trend in the number of advanced broadband providers (2001-2007) .........................174-1 Variables, measurement, and data sources fo r Model I, the use of a triple-play strategy ....... 764-2 Variables, measurement, and data s ources for Model II, market performance ....................... 834-3 Variables, measurement, and data sources for Model III, market entry .................................. 885-1 Descriptive statistics for the use of triple-play in the cable industry .......................................925-2 Descriptive statistics for the use of triple-play in th e telephone industry ...............................925-3 Descriptive statistics for the use of triple-play in both industries ...........................................935-4 Regression analysis of the use of triple-play in the cable industry .........................................955-5 The impact of competi tion on cables triple-play .................................................................... 965-6 Regression analysis of the use of triple-play in the telephone industry ..................................985-7 The impact of competi tion on telcos triple-play .................................................................... 985-8 Discriminant function analysis of the us e of triple-play in the cable industry ...................... 1015-9 Discriminant function analysis of the us e of triple-play in the telephone industry ............... 1025-10 Results of ANOVA of co st differences by industry ............................................................ 1045-11 Summary table for the use of triple-play .............................................................................1075-12 Descriptive statistics for the market performance in the cable industry .............................. 1115-13 Descriptive statistics for the market performance in the telephone industry ......................1125-14 Results of regression of the impact of triple-play on subscriber performance in the cable industry ................................................................................................................ ...1165-15 Significant factors of subscriber performance in the cable industry ...................................1175-16 Results of regression of the impact of tr iple-play on financial performance in the cable industry ...................................................................................................................... ......1195-17 Significant factors of financial performance in the cable industry ...................................... 121

PAGE 9

9 5-18 Results of regression of the growth rates of m arket performance in the cable industry ..... 1235-19 Significant factors of grow th in the cable industry ..............................................................1245-20 Results of regression of the impact of triple-play on subscriber performance in the telephone industry ............................................................................................................1265-21 Significant factors of subscriber performance in the telephone industry ............................1285-22 Results of regression of the impact of triple-play on financial performance in the telephone industry ............................................................................................................1305-23 Significant factors of financial pe rformance in the telephone industry ............................... 1315-24 Results of regression of the growth ra tes of market performance in the telephone industry ...................................................................................................................... ......1335-25 Significant factors of growth in the telephone industry ......................................................1345-26 Results of ANOVA of market performance by years of triple-play in the cable industry .. 1365-27 Results of ANOVA of mark et performance by years of tr iple-play in the telephone industry ...................................................................................................................... ......1375-28 Results of ANOVA of individual service revenue share ..................................................... 1415-29 Summary table for the use of triple-play .............................................................................1435-30 Descriptive statistics for market entry in the high-speed fixed broadband market ............. 1525-31 Results of regression of market entry .................................................................................. 1545-32 Significant factors of market entry in the high-speed fixed broadband market ..................1555-33 Summary of significant fa ctors of market entry .................................................................. 1556-1 Significant factors of the use of triple-play ...........................................................................1586-2 Significant factors of market performance ............................................................................ 169

PAGE 10

10 LIST OF FIGURES Figure page 3-1 Model I: The use of a triple-play strategy ............................................................................... 613-2 Model II: The impact of a triple-pla y strategy on triple-pla y providers market performance ................................................................................................................... ....643-3 Model III: The impact of the triple-play bundle on market entry ...........................................685-1 Capital expenditures as sh are of revenue by industry ........................................................... 1055-2 Marketing expenses as sh are of revenue by industry ............................................................ 1066-1 Discrimination of firms by firm variables in the cable industry. ........................................... 1636-2 Discrimination of firm s by firm variables in the telephone industry. ...................................1646-3 Subscriber and revenue by years of triple-play in the cable industry ....................................1746-4 Subscriber and revenue by years of triple-play in the telephone industry ............................. 1746-5 Individual service revenue as a share of total revenue by industry ....................................... 1806-6 Market entry by years of the triple-play strategy .................................................................182

PAGE 11

11 Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy TRIPLE-PLAY STRATEGY IN THE CA BLE AND TELECOMMUNICATIONS INDUSTRIES: AN EMPIRICAL STUDY OF ITS USE AND IMPACT ON THE US COMMUNICATIONS MARKET By Seonmi Lee December 2009 Chair: Sylvia M. Chan-Olmsted Major: Mass Communication Increasing competition in the recent telecommunications market has encouraged cable and telephone companies to engage in the triple-play strategy th rough which television, telephone, and broadband Internet services are marketed in a bundle. As this practice becomes more prevalent in both sectors, this study examines the potential determinants of its use by those industries. In addition, this st udy explores how the triple-pla y strategy benefits cable and telephone companies, assesses market performance of triple-play providers, and investigates how the triple-play bundle influences market entry in the communications market due to potential effects of bundling on deterring entr y of rivals. A panel data anal ysis using a US dataset from 2000 to 2007 was employed in this study. For the drivers of the use of the triple-play strategy, the results of the regression analysis indicate that market co mpetition, platform competition, and market potential factors influence the practice of triple-p lay in the cable industry. Also, market competition, platform competition, market potential, and firm size factors influenc e the practice of triple -play in the telephone industry. In the application of Po rters generic strategy in the tr iple-play market, cable firms tend

PAGE 12

12 to employ a differentiation strate gy while telephone firms tend to remain in a cost leadership position. For the impact of triple-play on market performance of the triple-play providers, the result of the regression indicates that the triple-play strategy generally enhances market performance in the cable and telephon e industry. In addition, switching costs tend to improve market performance for the triple-play providers. Increased market competition is likely to negatively influence market performance of triple-play prov iders, while platform competition contributes to improve data service performance in the cable and telephone industry. In addition, cable and telephone companies show superior performance in their core services co mpared to their other two services, which support the re source-based view of strategy. For the impact of triple-play on market entry, th e result of the regres sion indicates that the triple-play bundle influen ces market entry positively but at decreasing rates in the high-speed fixed broadband market. This implies that th e triple-play bundle enab les firms to have a competitive advantage over single service providers, which may result in deterring entry to some degree. Local loop unbundling regulation has an imp act on market entry jointly with the impact of triple-play.

PAGE 13

13 CHAPTER 1 INTRODUCTION The Triple-Play Strategy and Its Impact on the Market A triple-play offering is the bundled servic e that com bines television, fixed telephone service, and broadband acce ss over a single access network at a discount price (Trend Communications, 2006). Generally, the triple-play bundl e is anticipated to bring about synergies in that it will help raise the Average Revenue Per User (ARPU)1, which represents a prime indicator of profit poten tial; increase the number of subscrib ers; and reduce subscribers churn rate2 (Nortel, 2005; UT Starcom, 2006). This is b ecause the triple-play bundle is assumed to aggregate consumers who have different preferences for individual services as well as to lock in the consumer within a bundle (Bauer, 2006; Bughin & Mendonca, 2007; Crampes & Hollander, 2006). For example, triple-play providers can at tract consumers who have a high willingness to pay for video and data service (say $40 for vide o, $20 for data service) but a low willingness to pay for voice service (say $5 for voice service) by setting the lower price of the bundle (say $55) rather than the summed prices of video, voice, and data service ($65, if assumed that the price of video, voice, and data service is $40, $15, and $10, re spectively). If the thr ee services are offered individually, the consumer would not subscrib e to voice service because the price of voice service ($10) is higher than the consumers willingness to pay ($5). In fact, Kagan Research (2003) showed that the telecommunications comp anies reduced the churn rate to 1.2% at the beginning of the provision of the triple-play service (Cisco Systems, 2005). In 2006, triple-play service providers, such as Comcast, Cox, Time Warner Cable, AT&T, and Verizon, successfully 1 ARPU, which is commonly calculated by dividing the aggregate amount of revenue by the total number of users who generate that revenue, is usually seen as a prime indicator of profit potential. 2 The consumers churn rate is the rate of attrition that subscriber-based consumers unsubscribe of the consumer base, mainly employed in cable and satellite television and the wireless telephone industries.

PAGE 14

14 added subscribers of broadband, Internet Protocol TV (IPTV), and Voice over Internet Protocol (VoIP) services owing to the trip le-play service (Forbes, 2007). Furthermore, it is presumed that the triple-play service providers can deter a sing le service providers entry in the communication market because the bundled service is offered, in ge neral, at a lower price, which would lead to the reduction of a new entrants profits (Bauer, 2006; Nalebuff, 2004). The triple-play bundle emerged under pressure from increased market competition. There are two main external drivers for an increase in market competit ion. First, recent developments of Internet Protocol (IP) technology has helped media and telecommunications services to converge. As a result, innovative an d advanced IP based services su ch as VoIP and IP TV have emerged in the marketplace (Pernet, 2006). These ne w services are expected to take away from the market share of traditional video and telephone servi ces (Zimmerman, 2007). Second, the Federal Communications Commission (FCC) has made an effort to boost competition in the telecommunications market by allowing cross-entr ance of cable and telephone industries into each others market (Shim & Oh, 2006). Cable operators have begun to offer telephone services using the IP network, and traditional telephone op erators have started offering video services over the IP network (Crampes & Hollander, 200 6). Consequently, competition in the US telecommunications market has exponentially increased as video and telephone providers compete for the same customers within a single geographic area (Pernet, 2006). Besides, the market environment has become more complicat ed as cable operators compete with telephone companies that mostly team with direct broadcast satellite (DBS) in offering triple-play services (Seo, 2007). Such changing market dynamics will ha ve a marked impact on the strategies of the US cable and telecomm unications firms.

PAGE 15

15 Motivated by current market dynamics, both telecommunications and cable firms see the triple-play services as an essential strategy in their pursuit of market growth (Krauss, 2006; OECD, 2006). In response to the firms efforts to distribute the triple-play service, the triple-play market has been growing rapidly in the US. It is projected to grow from $19.5 billion (USD) in 2005 to $119.5 billion (USD) in 2010 (Parks Associat e, 2006). The subscriber performance of triple-play providers continuously enhances. As seen in Table 11, top triple-play providers, on average, have kept adding subscribers of vi deo, voice, and data services from 2005. Note that Cablevision first launched the triple-play serv ice in 2003 (Higgins, 2004) (for a closer look at growth trends in individual firms, see Table 1-1). Despite the current success of attracting additional subscribers, the long-term benefits of the triple-play bundle still are uncertain. As s een in Table 1-1, a lthough cable and telephone companies continuously gained subs cribers of each three services after the implementation of the triple-play strategy (the year of 2004, on average), the growth rate of the subscribers of three services declined from 2006. Except for Time War ner, other firms in Table 1-1 experienced a declining growth rate of video, voi ce, and data services (in all or at least two of the services) from 2006 to 2007. Considering that the triple-play market is in the growth stage3, the future growth and profitability of the triple-play bundle are st ill questionable. This study presumed that competition may influence a reduction in growth rates. In other words, competition can weaken the positive effect of the tr iple-play bundle (Bugh in & Mendonca, 2007; Crampes & Hollander, 2006). 3 According to IDC (2006), it is estimated that 39 percent of US households will use a triple-play service in 2010 (Kim, 2007).

PAGE 16

16 Table 1-1 Growth trends of the top tr iple-play providers in the US (2002-2007) Compan y Servi ce Year 2002 2003 2004 2005 2006 2007 No of Sub Growth rate (%) No of sub Growth rate (%) No of sub Growth rate (%) No of sub Growth rate (%) No of sub Growth rate (%) No of sub Growth rate (%) Comcast Video 29200 186.3 27500 -5.8 28600 4.0 29400 2.8 35500 20.75 39300 10.70 Voice 1438 1267 -11.9 1100 -13.1 1200 9.1 2400 100.00 4600 91.67 Data 3620 281.9 5000 38.1 6600 32.0 8100 22.7 11000 35.80 13200 20.00 Cablevision Video 2963 -1.5 3850 29.9 4446 15.5 4990 12.2 5574 11.70 5751 3.18 Voice 40.00 282 605.0 739 162.1 1214 64.28 1592 31.14 Data 770.1 52.0 1057 37.3 1353 28.0 1694 25.2 2039 20.37 2282 11.92 Charter Video 9261.6 2.0 8789.1 -5.1 8666 -1.4 8681.1 0.2 8106.5 -6.62 8140.3 0.42 Voice 22.8 0 24.9 9.2 45 82.3 121.5 167.6 446.3 267.33 959.3 114.95 Data 1138.1 184.5 1527.8 282.0 1884 371.1 2196.4 449.1 2393.4 498.35 2682.5 570.63 Cox Video 6280.9 0.7 16455.7 162.0 21061.8 28.0 21503.8 2.1 Voice 718.42 58.4 988.4 37.6 1305.4 32.1 1683.6 29.0 Data 1408.0 252.0 1987.2 396.8 2571.2 542.8 3143.3 685.8 Mediacom Video 1078 2.0 1032.6 -4.2 1019 -1.3 1062 4.2 1055 -0.66 1037 -1.71 Voice 71 0 106 49.30 data 110 -72.5 157.8 -60.6 205 -48.8 266 -33.5 320 -20.00 359 -10.25 AT&T Video 513 0 692 34.9 1510 118.21 2347 55.43 Voice 27528 0 26683 -3.1 25308 -5.15 35047 38.48 Data 6921 0 8538 23.4 12170 42.54 14156 16.32 Time Warner Video 6884 1.0 6887 1.0 13403 0.5 13979 1.0 20672 0.68 21273 0.97 Voice 206 0 998 0.2 1860 0.54 2895 0.64 Data 620 0.5 954 0.7 3368 0.3 4141 0.8 6644 0.62 7620 0.87 *Source: SEC Filing ** No. of Sub= the number of subscribers *** Unit of the number of subscribers: thousand

PAGE 17

17 In addition, it is expected that triple-pla y bundles will impact the structure of the telecommunications market. At the macro leve l, firms will expand their businesses to new advanced services by either developing their own technologies or coll aborating with other service firms with the triple-p lay strategy. Thereby two integrating firms will compete against each other in a single geographic area (C rampes & Hollander, 2006; Zimmerman, 2007). Consequently, it is plausible for the structure of the telecommunications market to become a duopoly. At the micro level, triple-play providers can deter a single servic e providers entry into the market through either an aggressive prici ng strategy or the provision of more attractive services (Bauer, 2006). In this case, whether the bundling is pro-competitive is always an issue because the bundling strategy can prevent fair competition (Bughin & Mendonca, 2007). Table 1-2 shows the trends in market entry in the high-speed broadband market. Note that the triple-play service requires advanced broadba nd lines to be delivered because components of the triple-play bundle such as IP TV and VoIP are advanced IP based services. As seen in Table 1-2, the number of providers continuously incr eased from 160 in 2001 to 1360 in 2007. However, as an increasing number of cable and telephone companies offer triple-p lay services after 2003, the growth rate of the number of providers decreased from 2.62% in 2005 to 1.03% in 2007. Table 1-2 Growth trend in the number of advanced broadband providers (2001-2007) Year Advanced broadband line The number of advanced broadband providers Growth rate 2001 5,571,605 160 1.38 2002 10,029,042 237 1.48 2003 15,863,169 378 1.59 2004 22,966,048 485 1.28 2005 37,332,557 1,270 2.62 2006 51,101,510 1,326 1.04 2007 69,556,081 1,360 1.03 *Source: Yearly FCC High-Speed Services for Internet Access Reports. ** For data through 2004, only those providers with at least 250 lines per state were required to file.

PAGE 18

18 Significance of the Study New m arket environment where competiti on accelerates often pushes communications firms to formulate new business d ecisions and engage in new stra tegies. The triple-play strategy has especially been one of the most significan t strategic moves for both telecommunications and cable operators in the last few years (Tim e Warner, 2007). As an increasing number of communications firms executes the triple-play stra tegy, prior studies have examined the drivers of the triple-play bundling. Such an alyses provide implications on the firms strategic options in different market environment, and give insight on how future competition might proceed in a converging communications market. Few prior studies have discussed the contributing factors of the use of the triple-play strategy at the market and firm level. At th e market level research, it is suggested that communications firms, in general, use the triple-play strategy because of increased competition incurred by technological convergence and the development of new technology (ITU, 2006; Seo, 2008). As technological convergence enables the co mmunications firms to expand their services toward new advanced IP-based services such as IP TV and VoIP, it is possible for a firm to provide a full range of video, voice, and data services to the consum er. To bundle all three services, firms can attract different segmen ts of customers (Larsson, 2007; Seo, 2007). In addition, the pro-competition regulatory environment played an important role in the emergence of the triple-play strategy (Cra mpes & Hollander, 2006). At the firm level, prior studies claimed that if a firm is able to enjoy economies of scal e and scope, it will use th e triple-play strategy to reduce costs (e.g., production and transaction costs) for the provision of various services (Baranes & Le Blanc, 2006; Bauer, 2005; Ovum 2008). Only one empirical study examined the factors affecting the use of the triple-play strategy in the cabl e industry, and found that larger local cable systems are more likely to o ffer triple-play services (Seo, 2007).

PAGE 19

19 In spite of the growth of trip le-play services in both the cable and telephone industries, previous studies peripherally discussed the poten tial determinants of th e triple-play strategy. Thus, little is known about the rela tive importance of the factors a ffecting the use of triple-play. Further literature in the relationship between re levant factors and the use of the triple-play strategy is likely to provide insightful interpretation of the firms current strategic behaviors, suggest the firms future strategic decisions, and forecast future competition dynamics among communications firms. Furthermore, this study will compare contribut ing factors of the use of the triple-play strategy between the cable and telephone indus tries. Cable and telephone companies have different competitive advantages based on th eir market and operational backgrounds. Cable companies have the ability to exploit synerg ies from video entertainment whereas telephone companies have core capabilities to provide traditional telephone se rvices (Bauer, 2006; Crampes & Hollander, 2006). In addition, they have operated under different regulatory regimes (Crampes & Hollander, 2006). A comparison of the factors between th e two industries will provide other explanations on how these two industri es behave differently with their resources in a competitive market. Regarding the impact of the triple-play stra tegy, a triple-play bundle has the potential to affect from a companys market performance to market structure. Traditionally the effects of bundling have been examined from two perspectiv es, i.e., a firms market performance and the market structure the firm is in. For market perf ormance, prior research suggest that bundling can be utilized for maximizing a firms profit in that it decreases the degree of difference in consumers preferences, the so called sor ting effect (Adams & Yellen, 1976; Bakos & Brynjolfsson, 1999; Salinger, 1995; Schmalensee, 1984; Stremers ch & Tellis, 2002). To date,

PAGE 20

20 few studies have suggested how the triple-play b undle would influence market performance (Yankee Group, 2003; Krauss, 2006). They have asse rted that triple-play will have a positive impact on the firms profitability in terms of increasing ARPU and decreasing consumers turnover. Nevertheless, recent studies have show n that the bundling act itself generates a new type of competition when two competing firms offer the same kinds of bundle (Nalebuff, 2004; Reisinger, 2004). The reduction in consumers valuation incurred thro ugh bundling would lead to indifference in the characteristics of the two different bundles. Thereby it would result in declining profits. Therefore, an examination of the profitability of bundling offers utility in assessing whether a triple-play bundle will have a positive impact on market performance under the circumstance where cable and telephone triple-p lay providers compete for the same customer (Baranes & Le Blanc, 2006). Three aspects of market perfor mance are proposed to interpret the impact of a triple-play bundle. The first aspect is whethe r the triple-play provi ders enhance subscriber performance. If the triple-play providers successfully aggregate the consumer, the number of subscribers would increase. The triple-play market is still in the gr owth stage, so the firms have room to attract subscribers. However, an increase in subscribers does not guarantee the improvement of a firms financial health. Thus the second aspect of market performance is whether the triple-play providers enhance their financial performance. Du e to competition in the market, it is possible that the increasing marketing expenses might not be compensated by the acquisition of new subscribers. In other words, if the competition effect outweighs the sorting effect of the tripleplay bundle, it would negate the firms profits. Another aspect to examine is the growth rate of subscriber and financial performance. The growth rate usually points to th e future trend of a product or se rvice (Reda, 2004). It also makes

PAGE 21

21 it possible to make a comparison of the product/service development between two consecutive time periods. For these reasons, the growth rate is sometimes employed to forecast the diffusion of technologies (Michalake lis et al., 2008). The growths of s ubscriber and financial performance are closely related to the success of the triple-play strategy. Thus, an examination of growth rates is useful in assessing how the triple-play strate gy has been developed over time, and predicting the future growth of triple-play providers. Regarding the triple-play impact on market stru cture, the bundling strategy is expected to prevent potential rivals, especially single se rvice providers, from coming into the market (Nalebuff, 2004; Whinston, 1990). One of the explan ations for this notion is that the market power of a firm in one market can protect the firms position in the other market. Moreover, the firm also has the choice to fix the price of the bundle far lower than the rivals prices. Based on this notion, some prior research proposed that the triple-play strategy can pl ay a role in deterring new entrants (Baranes & Le Bl anc, 2006; Bauer, 2005; ITU, 2006; Papandrea et al., 2003). The impact of the triple-play bundle on market entry is a sensitive issu e for regulators because there is a risk that the triple-play providers can le verage their market power into a new market (Baranes & Le Blanc, 2006). Therefore, it is essential to examine how the triple-play strategy influences market structure. Such an analysis will provide insights into whether the triple-play bundle is anti-competitive. Purpose of the Study This study has three goals. The first goal is to exam ine the factors associated with the use of the triple-play strategy in the cable and tele phone industries. The seco nd goal of this study is to analyze how the triple-play bund le influences the triple-play providers market performance, and the last goal is to assess how the bundling practice impacts market entry.

PAGE 22

22 This study will examine which spec ific factors relate to the use of the triple-play strategy in both the cable and telephone indus tries. First, this study will explore the relationship between the factors and the use of the triple-play st rategy in each cable and telephone industry, respectively, at the firm level. Then, this study w ill compare the impact of factors associated with the use of the triple-play strategy between the cable a nd telephone industries. The bundling strategy is expected to enhance market perforamance because it might attract different segments of consumers within a bundle. However, when two rivals that offer similar bundled services compete against each other, bu ndling, by nature, can incur another type of competition because bundling enables two bundled se rvices to be substitutable. Consequently, bundling would negatively influence market perfor mance. This study will examine the impact of triple-play bundles on the providers market performance. Fo r this analysis, three aspects of the market performance will be addressed: subscriber performance, financial performance, and growth rates. To clearly test the impact of the triple-play bundle, this study will control for other market variables associated with the firms mark et performance regardless of the impact of the triple-play bundle (e.g., the grow th of infrastructure, and other firm characteristics). Another possible impact of the triple-play bundle is that it can deter rivals entry into the marketplace. For this reason, debates have rece ntly emerged on whether triple-play bundles would prevent fair competition. This study w ill examine how triple -play bundles influence market entry in the communications market. The analysis will control for other variables associated with market entry regardless of the impact of the triple-play bundles (e.g., the growth of infrastructure, network effects, etc.). In all three analyses, this study will capture the changes in market and firm conditions over time from 2000 to 2007. The first two analyses (the drivers for the practice of the triple-play

PAGE 23

23 strategy, and its impact on the firms market perf ormance) will be addressed at the firm level, and the last analysis (the impact of the triple -play bundle on market entry) will be at the state level. Also in the first analysis this study categorizes all relevant factors into two categories, the market factor and the firm factor. A firms strate gic behaviors, in general, are influenced by the external market environment as well as the fi rms specific characteristics. Therefore, the categorization of the relevant variables will help interpret the interacti on between the external market factors and the internal fi rm factors. In the second analysis this study divides the relevant factors into three categories: strategic factors, mark et factors, and firm fact ors. Since the strategic factors are the primary focus of this study, othe r factors will be care fully controlled. In the similar fashion, other factors influencing market entry regardless of the strategic aspect will be controlled in the th ird analysis. This dissertation is organized as follows: In Chapter 2, the relevant literature on factors affecting the use of the triple-play strategy, market performance, and market entry will be reviewed with an emphasis on the impact of bundling. Relevant research questions will be also proposed in this chapter. In Ch apter 3, a research model will be presented. In Chapter 4, data gathering methods and procedures will be discus sed. In Chapter 5, the empirical analysis and results of the study will be pr esented. Lastly, in Chapter 6, implications of the findings, limitations, and suggestions for further research will be discussed.

PAGE 24

24 CHAPTER 2 LITERATURE REVIEW This chapter reviews scholarship on factors c ontributing to the practice of the triple-play strategy. In addition, this chapter reviews litera ture on the impacts of the triple-play bundle on market performance and market entry. Literature on other factors that affect the triple-play bundle or independently, the commun ications market will also be reviewed for the purpose of isolating the impact of the trip le-play bundle. Relevant research questions will follow the review of literature on each factor. Two Perspectives for Analyses of the Triple-Play Strategy This study derives factors influencing the use of the triple-play strategy and its im pact on the market from two perspectives: the Industr ial Organization (IO) approach and the ResourceBased View (RBV). The IO approach emphasizes the role of external factors while the RBV focuses on the internal firm fact ors. These two perspectives are commonly used to investigate drivers of firms strategic decisions as well as to analyze the relationship between firms strategies and their perfor mance (Gaddard et al., 2009). The Industrial Organization (IO) Approach Industrial organization theory is an approach used to understand the relations am ong structure, conduct, and market pe rformance. The structure of a market is mostly dependent on the number of sellers. Market conduct refers to a firms behavior toward its customers and competitors. Market performance is evaluated based on the ability of a firm to achieve the such goals as organizational e fficiency or progress. A firms strategic choice is often supported by a sensitive, ongoing analysis of the external environment (Aaker, 1984). The industrial organi zational (IO) view of strategy focuses on the linkage between a firms strategy and its extern al environment (Chan-Olmsted, 2006). In the IO

PAGE 25

25 view of strategy, a firms strate gic decision is influenced by the industry-level determinants of competitive conditions such as the degree of competition, which influences the firms performance (Porter, 1991; Chan-Olmsted, 2006). T hus early IO studies i nvestigated the effects of industry structure and firms conduct on firms performance (Goddard et al., 2009). Porter (1980) formulated five exogenous influences of the fi rms strategy, called Porters Five Forces Model. This Model employed concepts in IO economics to derive five forces that determine the competitive intensity and subsequent attractiveness of a market. The five identified forces are the extent and intensity of competition, the treat of entrants, the threat of substitutes, the power of buyers, and the power of suppliers. Porters Five Fo rce Model is widely used to analyze a firms status in the current market as well as its strategic decisions. From this perspective, th e general environment of a firm including economic, political/legal, and technological factors influences a firms strategic moves in the current communications market (Chan-Olmsted, 2006; L ee & Lee, 2008). To as sess the triple-play strategy, this study considers the industry-level determinants of competitive conditions. The Resource-Based (RBV) View The Resource-Based View (RBV) approach em phasizes an individual firms unique capabilities and their impact on the firms business strategy in the market (Chan-Olmsted, 2006). From the RBV perspective, a firms heterogeneous resources can generate competitive advantages, which lead to s uperior performance (Chan-Ol msted, 2006). The RBV approach helps analyze the internal resources of a firm at industry level, and to closely assess the firms advantages and disadvantages. The firms resources are placed in two cate gories: property-based and knowledge-based resources (Barney, 1991; Chan-Olmsted, 2006). Prope rty-based resources are inimitable because of the property rights or physical facilities based on the complexi ty of network configuration.

PAGE 26

26 Knowledge-based resources are the firms intellec tual skills related to the firms cumulative experiences which operate as knowledge barriers Landers and Chan-Olmsted (2002) proposed that property-based resources in clude station ownership, copyright and top content, and that knowledge-based resources embrace audience expert ise, content multi-purposing expertise, and technology management. Barney (1991) suggested when resources are rare, difficult to imitate, and valuable, such resources lead to superior performan ce when used strategically. From this perspective, the firm-level specific features influence an organizations strategic moves in the current communications market. To assess the triple-pla y strategy, this study considers firm-level factors. Factors Contributing to the Use of Triple-Play The Impact of Competition on the Use of Triple-Play Market competition Cram pes and Hollander (2006) propos ed that a main driver of th e triple-play strategy is the need to mitigate competition. Usually, bundling is assumed to be an effective means of relaxing market competition. When two competing firms se ll two or more differen tiated products in a single package, they are able to differentiate themselves. Consequently, the firms can avoid direct price competition (Carbajo et al., 1990; Chen, 1997). Liao and Tauman (2002) demonstrated that in a market where competiti on exists, the bundling strategy can be the only device to improve benefits to both sellers and consumers at the same time. Without bundling, price competition between multiproduct firm s cannot be relaxed. Moreover, Bakos and Brynjolfsson (2000) showed that the bundle provider can capture most of the market share from sellers that provide a ll products independently. The aforementioned benefits of bundling are likely to drive cable and telephone companies to bundle their services in th e communications market, whic h has experienced increased

PAGE 27

27 competition since the Telecommunications Act of 1996. Note that the primary purpose of Telecommunications Act of 1996 was to create a free and competitive market where commercial telecommunication providers would compete fo r both residential and commercial accounts. Section 706 of the Telecommunications Act of 1996 places emphasis on high-speed, switched broadband telecommunications capability that enables users to originate and receive high-quality voice, data, graphics, an d video telecommunications using any technology. Traditionally, cable and tele phone industries have been characterized as natural monopolies4 (Grubesic & Murray, 2004). However, the Federal Communications Commission (FCC) supported cross-entrance by the cable and telephon e industries into each others market, enhancing competition in a single geographica l market. Stated differently, the natural monopolistic market structure has gradually transf ormed into a competitive market (Crampes & Hollander, 2006; Shim & Oh, 2006; Zimmerman, 2007). In a competitive market, cable and telephone companies s ee triple-play as a strategy of continuing market growth. Platform competition W ith changes in the market environment, recent technological convergence5 has generated additional competitive pressure on the cable and telephone industries (Tardiff, 2007). The phenomenon of convergence makes it possible to deliver the same or similar services over different networks (Bauer, 2006; Seo, 2008). As the quality of converged services and overall 4 The market where cable and local tele phone providers serve has the characteri stic of a natural monopoly. Although cable video services compete against satellite video and other broadband video services, in each local area, one cable system operates in most areas. Incumbent local exchange carriers (i.e., local telephone service providers) do the same although they compete with other competitive local exchange carriers (CLECs) in the local telephone market. 5 According to Bauer (2006), technological convergence is defined as development s affecting the technological basis of communications at the level of networks, applications and services. As a result of technological convergence, three changes, digitiza tion, increased processor speed, and the migration to higher transmission capacity, are made. Also Bauer clarified the notion of conve rgence and classified it into three areas: technological, market, and organizational convergence.

PAGE 28

28 network infrastructure can be compared by c onsumers, competition is propelled in the communications market (Baranes, 2006; Bauer, 2006). Though the triple-play service may increase inter-modal competiti on through platform competition in the market, platform competiti on contributes to faster broadband diffusion (Crandall 2005; Lee, 2006), which implies a speedy increase in size of the market and thus an increase in demand for new services (Lee, 2006). Hence firms may have an opportunity to generate new revenue streams if they develop network infrastructure create value-added services, and/or differentiate themselves in the triple-play market. Three platforms are mainly used by cable a nd telephone operators to deliver broadband related services such as IP TV and VoIP in a tr iple-play service: Digita l subscriber loop (DSL), cable modem, and fiber optic (ITU, 2005). Cable ope rators provide broadband related services in a triple-play service mostly using cable modems while telephone operato rs use either DSL or fiber optic (OECD, 2006). Because functionally similar triple-p lay bundles are available on one of three platforms that have different features respectively,6 triple-play providers have made an effort to attract subscribers by offering enhanced and differentiated services and/or investing in upgrading network infrastructure (Grubesic & Murray, 2004; ITU, 2005; OECD, 2006). Lee and Lee (2008) found significant group differences among different platforms (DSL, cable, and fiber optic) in triple-play se rvice pricing, video service differen tiation, and service speed. Bughin and Mendonca (2007) demonstrated that telecommunicat ions providers with fiber optics serve the triple-play bundle at a hi gher price than their rivals that have a cable or DSL network. 6 DSL and cable modem offer 1) high bandwidth to consumers, and 2) fast speed to transmit data, video, and voice connections. However, DSL has drawbacks in terms of 1) limitation of geographical distance to serve, and 2) lower quality of the copper line. Cable modem embraces a prob lem of congestion during peak-hours. Fiber optic has advantages such as 1) fastest speeds, and 2) the highest bandwidth demand and installation cost.

PAGE 29

29 The practice of triple-play Despite the new type of com petition cause d by technological conve rgence, Crampes and Hollander (2006) concluded that triple-play bundling is an essen tial strategy for communications firms because it enables the cable and telephone industries to protect th eir market share or maintain their dominant market positions (Bau er, 2006). Baranes and Le Blank (2006) asserted that communications firms made a decision to o ffer triple-play bundles be cause they can offset the losses from competition. Pernet (2006) point ed out that telephone operators aggressively offer such bundles as they face competition with cable operators. She also claimed that when a market is highly competitive, the number of bundles increases. Empirically, Bughin and Mendonca (2007) showed that as market become s competitive, communications firms are more likely to offer triple-play bundles. Based on the literature reviewed, the follo wing research questions (RQs) are proposed: RQ1a. Does competition influence the use of the triple-play strategy in the cable industry? RQ1b. Does competition influence the use of the triple-play strategy in the telephone industry? RQ1c. Does competition influence differently the use of the triple-play strategy between the cable and tele phone industries? The Impact of Information Communication Technology on the Use of Triple-Play The developm ent of information communicati on technology benefits consumers when new and advanced services are available to them7 (Distaso et al., 2006). As such, the development of communications network infrastructure is an essential condition to introduce new advance 7 Crandall et al. (2002), for example, report that in US consumer benefits from universal broadband deployment, that is to say 94% of US households, which is the current level of telephone service, could easily be $300 billion a year, and that 50% deployment would resu lt in benefits of around $10 0 billion annually (benefits increase nonlinearly due to network effects). See also Distaso et al. (2006).

PAGE 30

30 services within a triple-play bundle (Pernet, 200 6; Lee & Brown, 2008). In offering a triple-play service, sufficient diffusion of advanced comm unication infrastructure is required (Lee & Brown, 2008). This is because many components of the triple-play service are IP-based communication services, such as IP TV and Inte rnet telephony, which are carried over advanced communication networks (Kim & Kim, 2007). These enhanced IP-based communication services are based on broadband access to the home a nd interconnection (Laffont & Tirole, 2001). Grubesic and Murray (2004) emphasized the impor tance of diffused broadband for advanced telecommunication services. From such a pers pective, the development of information communication technology is associated with th e deployment of triple-play services in the marketplace. In general, the development of communicati ons network infrastructure helps achieve efficiency in production and dist ribution processes in terms of time and costs (Antonelli, 1997). The extent of productivity realized, however, is associated with a firms business strategies. Koski and Majumdarm (2002) proposed that a firm implements different strategies depending on each stage of the relevant networks technological development. That is, a firms decision on entry timing, service prices, diffusion speed, and service areas needs to keep pace with the development of their networks. For example, Shim and Oh (2006) indicated that when consumers evaluate the broadband Internet servic e sufficiently high, cable operators decide to bundle video and broadband Internet services. Grubesic and Murray (2004) concluded that advanced telecommunications service providers ar e willing to operate in urban areas where the cost structure is more favorable, and thereby ad vanced telecommunications services are unlikely to be provided in rural areas.

PAGE 31

31 Based on the literature reviewed, the follo wing research questions (RQs) are proposed: RQ2a. Does the development of communications network infrastructure influence the use of the triple-play strategy in the cable industry? RQ2b. Does the development of communications network infrastructure influence the use of the triple-play strategy in the telephone industry? RQ2c. Does the development of communicati ons network infrastructure influence differently the use of the triple-play strategy between the cable and telephone industries? The Impact of Market Potential on the Use of Triple-Play It is generally believed that m arket potential influences suppliers behavior, and vice-versa. Market potential drives communi cations firms to improve existing services and develop new ones in order to increase their market share (Seo, 2008). Thus, potential growth of market demand is closely associated with a firms stra tegic decision to expand its business. The positive relationship between demand and the provision of new services has also been empirically supported by the following studies. Grubesic and Murray (2004) indi cated that broadband service providers are more likely to offer adva nced telecommunications services in urban areas than in rural areas. In addition, Seo (2008) found that cable operat ors are more likely to provide VoIP in high population density areas than in low population density areas. Regarding the provision of triple-play services, Seo (2007) revealed that in areas where market demand is higher, a firm is likely to offer triple-play services. Based on the literature reviewed, the follo wing research questions (RQs) are proposed: RQ3a. Does market potential influence the use of the triple-play strategy in the cable industry? RQ3b. Does market potential influence the use of the triple-play strategy in the telephone industry?

PAGE 32

32 RQ3c. Does market potential influence different ly the use of the triple-play strategy between the cable and telephone industries? The Impact of Firm Size on the Use of Triple-Play It is assum ed that firm size has a positive correlation with market power and/or the firms capabilities and resource s (Koski & Majumdarm, 2002). First, larger firms are better able to capitalize on production and marketing efficiency to effectively offer trip le-play packages. They are more likely to have the capacity to develop and market multiple products simultaneously (Mitchell & Skrzypacz, 2006). Larger firms also have much higher bargaining power than smaller firms (Vagliasindi et al., 2006). Moreove r, larger firms tend to quickly respond to external and environmental factors (Dean et al., 1998). On the contrary, smaller firms take small market positions and have limited resources. Small firms are less likely to consider envir onmental market conditions in an industry and technological development (Dean et al., 1998). In addition, smaller firms are less likely to enter the market if substantial investment in offering se rvices or products is n ecessary due to lack of capital resources (Dean et al., 1998). Small fi rms are therefore less likely to expand their business than larger firms (Gentry & Jamison, 2005). In the communications market, the constructio n of communications network infrastructure requires enormous expenditures, and at the sa me time scale economies exist. Large firms are therefore more likely to survive in the market because they have a sizable cost advantage over smaller firms (Ford et al., 2005). In addition, in the mobile industry, a firm having larger network size has a competitive advantage over a firm having smaller network size due to network effects (Kim & Kwon, 2003). Regarding the practice of th e triple-play strategy, Seo (2007) revealed that cable operators having a larg e video subscriber base are more likely to implement the tripleplay strategy.

PAGE 33

33 Based on the literature reviewed, the follo wing research questions (RQs) are proposed: RQ4a. Does the firm size influence the use of th e triple-play strategy in the cable industry? RQ4b. Does the firm size influence the use of the triple-play stra tegy in the telephone industry? RQ4c. Does the firm size influence differently the use of the triple-play strategy between cable and telephone industries? The Impact of Cost on the Use of Triple-Play In general, bundling enables suppliers to achie ve efficiency in that it can reduce production costs and d istribution costs, save storage sp ace, or simplify production logistics (Evans & Salinger, 2008). In this context, the triple-play bundling strate gy may reduce transaction costs by consolidating three different se rvices into one package (Papa ndrea et al., 2003). Moreover, the triple-play strategy can reduce consumers educa tion costs that are necessary in launching new services. When a firm creates the bundle combining new services with existing services, the degree of consumers uncertainty about new services decreases, and consequently a firm can spend less money to persuade and edu cate consumers (Kim & Kim, 2007). Nevertheless, the triple-play strategy also re quires additional costs. This study considers two cost factors related to the implementation of triple-play strategy: marketing expenses and capital expenditures. Marketing expenses and capital expenditures ta ke the largest portions of the cable and telephone industries spending (Papandrea et al, 2003). The introduction of triple-play bundles requires a large amount of marketing expe nses (Net Insight, 20 06; Time Warner, 2008). When a firm launches a new service, investme nt in marketing is important because it will positively influence brand equity, contribute to shortand long-term profits, and maintain customer relationships (Moorman & Rust, 1999; Seggie et al., 2007). Marketing is also necessary when a firm seeks differentiation of serv ices or products from its competitors (Keller,

PAGE 34

34 2004). This is because differentiation creates a stable customer base and raises barriers to entry (Porter, 1980). Thus marketing expenses are highly related to a firms strategic decision to offer new services. For capital expenditures, the construction or upgrades of a communications network which enables the offering of bundled products requires substantial capital expend itures. The faster pace of investment can accelerate growth in exis ting services, higher productivity, overall lower prices, and new services (W ellenius & Townsend, 2005). This is also true when competition is intense in the market. Firms have an incentive to invest in innovation in order to have competitive advantage by offering new technologies or services (Janssen & Mendys-Kamphorst, 2008). In pa rticular, it is a widely held belief that intermodal competition through platform compe tition promotes investment in innovation (DotEcon & Criterion Economics, 2003). Recently accelerating platform competition through the provision of the triple-play is pos itively associated with the grow th of investment in innovation. As technological convergence enables multiple plat forms to carry functionally similar triple-play bundles, triple-play providers have made an e ffort to attract subscr ibers by investing in upgrading network infrastructure in order to re lax competition (Grubesic & Murray, 2004; ITU, 2005; OECD, 2006). Despite these benefits of investment in infr astructure, firms are reluctant to invest in building and/or upgrading infrastructure because th ey may not necessarily achieve expected sales or returns on expenditures (W ellenius & Townsend, 2005). Regulat ed firms are known to invest much less than unregulated firms (Crandall, 2009; Al esina et al., 2003). This implies that there is a difference in investment between cable (le ss regulated) industry and telephone (heavily regulated) industry.

PAGE 35

35 Moreover, Janssen and Mendys-Kamphorst (2008) pointed out that there is a technical difference between networks of cable and tele phone companies in delivering the triple-play service, and that this leads to different tim ing and strategies. Comp ared to video service transmission of telephone firms, cable compan ies can more easily deliver voice service via network infrastructure as voice transmission re quires less capacity than video transmission. Upgrading network is therefore es sential, especially for telephone companies, to carry video services that take high er bandwidth capacity, which delays the introduction of the triple-play service due to substantial cost of upgrades. Hence cable companies are at onset ready to offer the triple-play while telephone companies lag behind. Regarding timing of investment, Janssen and Mendys-Kamphorst (2008) summarized the relationship between investment and competition as an inverted U curve. That is, the investment in innovation initially increases as competition incr eases but the investment decreases later. Based on the literature reviewed, the follo wing research questions (RQs) are proposed: RQ5a. Does cost influence the use of the tr iple-play strategy in the cable industry? RQ5b. Does cost influence the use of the trip le-play strategy in the telephone industry? RQ5c. Does cost influence differently the use of the triple-play strategy between cable and telephone industries? Strategic Orientation Porter (1980) proposed three ca tegories of generic strategies : overall cost leadership, differentiation, and focus. The overall cost lead ership strategy require s the construction of efficient facilities, the ability of cost reductions through e xperience curve effects, and cost m inimization in all areas in order to mainta in a low-cost position in the market. The differentiation strategy centers on perceived uniqueness of servi ces or products. Thus this strategy requires strong capability in research, strong marking abilities, and highly elaborated

PAGE 36

36 labors that can combine various skills from ot her businesses. If the differentiation strategy succeeds, a firm can increase its earnings with a premium pricing strategy by creating brand loyalty among customers, which results in en try barriers. The focus strategy deals with a particular customer group, segment of the product line, or geographic market. The focus strategy serves its narrow strategic target in very eff ective or efficient ways, and consequently this strategy be advantageous over riva ls that target broader customer s, product lines, or geographic areas. Lee et al. (2008) applied Porters generic stra tegy to the Mobile Virtual Network Operators market. They suggested that the multiple play strategy used by media and telecommunication companies can be categorized as a differentiated co st leadership strategy. This is because media and telecommunication companies, in general, implement the strategy of offering bundling service at a discounted price, whic h can be a differentiation point. In addition, Lee and Lee (2008) showed how cab le and telephone companies are likely to choose different modes of multiple play strategy. They examined differences in price, speeds, and TV channels of the triple -play service by industry, and propos ed that cable companies tend to employ a greater differentiation strategy ba sed on video services, while telecommunication companies tend to employ a low cost strate gy to compete with cable companies. In conclusion, though triple-play strategy can be considered a differentiated cost leadership strategy, there are differences in the degree of differentiation and low cost strategy based on industries and platforms. Based on the literature reviewed, the follo wing research question (RQ) is proposed: RQ6. Is there a difference in strategic behavior s of the triple-play service providers by industry in terms of th e generic strategies?

PAGE 37

37 The Impact of Triple-Play Strategy on Triple-Play Provid ers Market Performance This section will review th e literature on the potential im pact of the bundling strategy. One of the primary goals of this study is to analyze the impact of tr iple-play bundles on the triple-play providers market performance. Based on the literature review, this study will propose a relationship between the triple-play practice and market performance. Literature on other factors that affect the triple-play bundle or i ndependently, the communica tions market, will be reviewed for the purpose of isolating the impact of the triple-play strategy. The Impact of Bundling on Market Performance Bundling ref ers to the practi ce of selling products in a si ngle package (Stremersch & Tellis, 2002). Bundling has two opposite impacts on market performance of the bundle provider (Bughin & Mendonca, 2007; Fay & MacKie-Mason, 2001; Reisinger, 2004). On the positive side, bundling can increase a sellers profit in a monopoly market if consum ers valuations of two goods are negatively correlated. This is because bundling typica lly reduces in heterogeneous consumers valuations, called the sorting e ffect (Adams & Yellen, 1976; Salinger, 1995; Schmalensee, 1984). The sorting effect is sti ll effective in more extended markets where multiple products and a number of consumers exis t. In such settings, bundling can also attract different segments of consumer s who have different valuations for products (Schmalensee, 1984; Stremersch & Tellis, 2002). In th is case, bundling helps the firm to extract consumers surplus, by leading demand to be inelastic and enabling demand for a bundle to be predictive (Adams & Yellen, 1976; Bakos & Brynjolf sson, 1999; Stremersch & Tellis, 2002). Empirically, Crawford (2008) discovered that bundling in the cab le network industry reduced the degree of heterogeneity, and thereby increased profits by 6. 0%. On the consumers side, bundling can also raise consumers surplus when a bundle is provide d at a discount price (Liao & Tauman, 2002).

PAGE 38

38 Other bundling benefits exist besides the sorting effect. When introducing new technology or services, if a firm decides to bundle existing technology or services with a new technology or service, the firm can reduce the degree of consumers uncertainty about those new technology or services (Eppen et al ., 1991). In order to retain existing subscribers and to attract new subscribers, Shapiro and Varian (1999) sugg ested that a communications firm should offer bundled complementary services to its installed base of consum ers. Moreover, by engaging in bundling, it is possible to lock in the subscribers by creating sw itching costs in order to avoid consumers turnover and guarantee future prof its (Eppen et al., 1991; Gans, 2000; Klemperer, 1995; Shapiro & Varian, 1999; Shy, 2001). Based on such potential impacts of bundling, triple-play provid ers can expect 1) to raise ARPU, 2) to see an increase in the number of subscribers, and 3) to retain existing consumers of their traditional services in the provision of a triple-play bundle (Nortel, 2005; UT Starcom, 2006). Triple-play bundles are expected to reduc e production and transaction costs due to a decrease in the number of tran sactions (Baranes & Le Blank, 2006). Furthermore, triple-play bundles help incumbents to have a competitive advantage over new entrants through the use of cross-branding and cross-promotions (ITU, 2006). Nevertheless, bundling does not always have a positive impact on a sellers profit. Bundling, by nature, can generate new types of competition, which would result in a decrease in profits (Fay & MacKie-Mason, 200 1). When two firms compete in a multi-product market, the firms necessarily implement bundling even if bundling creates a negativ e situation called the prisoners dilemma situation (Matutues & Re gibeau, 1992). Specifically, if 1) two firms compete against each other, 2) two firms each offer two different products, and 3) one firms two different products can be substitutable for a ri vals two products, then the two competing firms

PAGE 39

39 decide to bundle their services at a discount price. This is be cause the firm providing products independently would fail to attr act consumers if its rival offers the bundled product at a largely discounted price. However, the bundled products will lead to homogeniza tion of customers in that the consumers preferences for each product w ould be averaged out due to the nature of the bundle. The homogenization of customers woul d conversely intensify price competition, and directly erode profits, called the bus iness-stealing effect (Reisinger, 2004). In a comparison between the sorting effect and the business-stealing effect, Reisinger (2004) concluded that the busines s-stealing effect dominates the sorting effect if consumers reservation values are negatively correlated and, as a result, the values ca n be averaged out. Recently, Bughin and Mendonca (2007) examined tw o effects of triple-play bundles in Western Europe and found that the sorting effect outwe ighs the business-stea ling effect, though the sorting effect cannot fully compensate for the competition effect incurred the triple-play bundles. In the US communications market, cable a nd telephone companies provide triple-play bundles that are similar to each other in terms of functionality from the consumers perspective8 (TNT Telecom, 2005). To attract subscribers, cab le and telephone compan ies have the incentive to offer triple-play bundles at a significant discount otherwise they would fa il to gain subscribers of all three services (Crampes & Hollander, 2006; Lee, 2008). Such pric ing strategies would likely impact the firms profit negatively. Empi rically, Lee and Lee (200 8) revealed that the triple-play strategy led to a higher level of ARPU among tel ecommunications companies. In the long run, however, Tardiff (2007) claimed that an increased competition among the triple-play services will erode a firms market share and diminish its profits. 8 According to a survey by TNS Telecoms in 2005, the consumer is indifferent to who provide the triple-play offering and has no preferences for triple-play bundle providers.

PAGE 40

40 Based on the literature reviewed, the follo wing research question (RQ) is proposed: RQ7. Do triple-play bundles influence triple-p lay providers market performance such as subscriber performance, financial performance, and growth rates? The Impact of Switching Costs on Market Performance In this sec tion, literature on the relations hip between switching costs and market performance will be reviewed. In brief, switching costs have the potential to influence market performance jointly with the triple-play practice because triple-play provid ers artificially create switching costs when consumers s ubscribe to triple-play services. Based on the literature reviewed, this study will propose a relationshi p between switching co sts and triple-play providers market performance. The impact of switching costs on trip le -play providers market performance Switching costs are incurred when a consumer sw itches to a different service or adopts a new technology9 (Shapiro & Varian, 1999). For network competitors, switching costs soften competition because they prevent consumers from moving out, which generate barriers to entry (Gans, 2000; Klemperer, 1995). Furthermore, switc hing costs can guarantee future profits in that a firm can obtain a reliable consumer base durin g the term of the contract (Shapiro & Varian, 1999). The creation of switching costs helps a firm to raise long-run profits rather than short-run profits. A telecommunications firm usually has to endure temporary losses when it introduces new services if it has no installed base of subscribers (Cabral et al., 1999; Rohlfs, 1974). To build a base of subscribers, Rohlfs (1974) propos ed introductory pricing when a firm offers a 9 In the telecommunications industry, all information activities begin and complete using a system composing various hardware and software. The system can survive by locking in subscribers. When the consumers are locked in, switching costs emerge.

PAGE 41

41 new communications service at a low price (though the firm sacrific es initial profits). Similarly, Cabral et al. (1999) suggested that a firm sets a lower first-period price, i.e., introductory pricing, and continuously raises its price after the firm lo cks in the subscribers. The firm is thereby able to increase its profits. In a competitive market, firms tend to spend a large amount of money to attract new customers, and then try to recover losses from locked-in customers. Consequently, a firms profits would be no more than a normal rate of return on its investments (Shapiro & Varian, 1999). However, for firms having no locked-in customers, they would provide services at a discount through promotions to at tract new consumers even without any revenue stream that compensates for the huge expenditure (Shy, 2001). As a way of creating switch ing costs, Shapiro and Vari an (1999) proposed that a communications firm is able to lock in subscr iber through contracts. In this case, switching costs are the amount of compensation that contr act-breakers should pay. Using contracts, a firm is able to retain subscribers unt il the contract ends. Increasingl y, triple-play providers lock in subscribers by making contracts when they o ffer triple-play bundles to new subscribers10. When the amount of compensation and the term of cont ract are designed in an appropriate way, the firm can enhance market performance. Based on the literature reviewed, the follo wing research question (RQ) is proposed: RQ8. Do switching costs influence triple-play providers market performance such as subscriber performance, financial performance, and growth rates? 10 For example, Verizon Communication charges $199 when subs cribers of the FiOS triple bundle cancel a two-year contract before the term of contract ends.

PAGE 42

42 The Impact of Competition on Market Performance As mentioned, increasing competition is likely to be a main driver for the emergence of triple-play services. Competitive market dynamic s, jointly with the triple-play strategy, is expected to influence the triple-play serv ice providers market performance. However competition, by nature, can influence a firms market performance over time in spite of the strategies employed. Thus, this section will re view literature on how competition generally influences market performance of communications firms as well as how competition has a potential impact on the market perfor mance of the triple-play providers. General competition effect on market perfo rmance of communication service providers The industrial organization approach asserts that market performance is very much influenced by market structure (Albarran, 1996; Kaserman & Mayo, 2002). Market structure is heavily dependent on the number of sellers. The sellers market performance can be defined as how successfully the seller achieves organiza tional efficiency and progress (Albarran, 1996; Sherman, 1995). It is believed that competition relieves m onopoly power in a market (Loomies et al., 2005; Shapiro & Varian, 1999). Regarding the relations hip between market competition and market performance of communications firms, two opposite claims have been made to date. Some researchers suggest that only a concentrated market structure can provide a firm with enough financial resources to innovate and offer new services, and consequently enhance market performance (Burnett, 1992; Schumpeter, 1950). In a market where the number of competitors increases, the profit of its mark et leader is expected to decrease (Economides, 1996a). On the contrary, some have shown that in a competitive market, firms would be able to enhance market performance through market segmentation. A comp etitive market would not allow a single firm to acquire the largest market share (Li, 2006; Li, 2004; Litman, 1979), yet reduced prices due to

PAGE 43

43 competition would give the firms incentives to develop new products, which results in higher profits (Fay & Mackie-Mason, 2001). The relationship between market competition an d market performance is still inconclusive in the communications market. Regarding the ne gative relationship between market competition and market performance, Kaserman and Mayo (2002) showed that as market competition increases in the long-distance market, the ma rket share of telecommunications providers decreases over time. They also implied that co mpetitive pressure drove a reduction in rates of telephone services, which would result in nega tive profits. Woroch (2002) concluded that increased market competition in a local teleco mmunications market negatively influenced local network providers profits. As for network compet ition, Shy (2001) showed that in a market with new entrants, the incumbent experiences a reduc tion in profits whereas an entrant can raise profits. From a different perspective, Hazlett (2006) pointed out that as video markets have become competitive due to the FCCs deregula tion policy, cable operators continue to add subscribers and successfully increa se their revenues. Godes et al (2009) pointed out that media firms have two different revenue sources, i.e., content and advertising (called, the two-sided market structure), and showed that media firm s increase content prices as the video market becomes competitive, while they decrease advert ising prices. This happens because increasing market competition leads to intense competition for advertisers, which reduces the return per customer impression from the advertisement and thus advertising prices. Consequently, competition for advertising gives media firms an incen tive to increase content prices in order not to reduce total profits. Li (2007) found that organi zational efficiency of cable television systems is more likely to improve when competition exists in the franchise areas than in areas where no

PAGE 44

44 competition exists. She found that both the net-pr ofit rate and the penetr ation rate of cable systems in a competitive market are much higher than in a monopolistic market. Competition effects on the market p erformance of triple-play providers Com petition within components of triple-play serv ices is likely to influence the triple-play providers market performance (Grubesic & Murray, 2004; Shim & Oh, 2006). Cable and telephone operators have services within a triple-play bund le that can be substituted by a rivals products. A typical triple-play serv ice of the cable operator consists of cable televi sion, Internet telephony, and broadband services. Of these servi ces, Internet telephony has the potential to cannibalize the market share of fixed-line telephony service of telephone companies (Ferguson, 2002; Grubesic & Murray, 2004; Loomies et al., 2005; Shim & Oh, 2006; Woroch, 2002). Faulhaber (2003) predicts that VoIP will us urp telephony platforms after mobile becomes a dominant medium by substitu ting fixed-line telephones.11 The triple-play service of telephone operators is composed of IP TV and/or satellite video servic es (DBS), fixed-line telephony services, and broadband services Of these services, new video entertainment such as IP TV and/or satellite video entertainment12 might take away from the multichannel video distributors market share (Bauer, 2006; Crampes & Hollande r, 2006; Grubesic & Murray, 2004; Loomies et al., 2005; Shapiro & Varian, 1999). As similar broadband related services in the triple play service are carried over a variety of platforms, platforms encounter competition as well (OECD, 2006). In general, platform competition is likely to impact the broadband deployment (Lee, 2006), which leads to increases 11 To defend Internet telephony, some telephone companies also offer Internet telephony in addition to traditional fixed-line telephony. For example, Verizon Communication offers FiOS telephone service as Internet telephony at a higher price in addition to traditional fixed-line telephony at a lower price. 12 According to the FCC (2008), cables share of MVPD subscribers fell from 69.4 percent in 2005 to 68.2 percent in 2006. On the other hand, DBS increased its share from 27.7 in 2005 to 29.2 percent in 2006, a net increase of 2.6 percent.

PAGE 45

45 in the overall market size (DotEcon & Crit erion Economics, 2003). In addition, platform competition in triple-play bundles is likely to influence the speedy deployment of new technologies and services by forcing firms to of fer improved quality of services and diverse service choices (Zimmerman, 2007). Speedy diffusion of new services helps the market share of the firm in the communications market (Denni & Gruber, 2007). To relax competition, the triple -play service providers differe ntiate themselves. Currently cable companies use cable modems to carry broadband related services while telephone operators employ digital subscrib er loop (DSL) or fiber optic (G rubesic & Murray, 2004; OECD, 2006). Depending on the capacity of these networks, firms provide various prices and quality of service, which relates to a tr iple-play providers market perf ormance (ITU, 2003; Lee & Lee, 2008; OECD, 2006). Of all platforms, Bughin a nd Mendonca (2007) found that the price of the triple-play service delivered over fiber optics is the highest among the three, which would influence positively future revenue. Despite firms efforts to attract or retain cu stomers, once a consumer switches platforms, it directly influences the firms financial performa nce. For a more detailed analysis of replacement among platforms, Loomis et al. (2005) revealed that the displacement of local telephone operators by their own DSL netw ork increases in revenue but incurs huge costs, and that the displacement by a rivals DSL or cable modem directly reduces profits. Based on the literature reviewed, the follo wing research question (RQ) is proposed: RQ9. Does competition influence triple-play providers market performance, such as subscriber performance, financial performance, and growth rates? The Impact of Information Communicati on Technology on Market Performance Regardless of the im pact of triple-play bundles, informa tion communication technology (ICT) may influence a firms market performan ce as well. Thus, this variable needs to be

PAGE 46

46 controlled. This section will review the impact of ICT on a general communications firms market performance. The development of communications network infr astructure is closely related to new uses of communications technology and services (A ntonelli, 1997). The uses of communications technology by the service providers increase productivity because new technology helps achieve efficiency in production and dist ribution processes in terms of time and costs (Antonelli, 1997). In addition, the transaction costs of doing busin ess fall with the development of ICT, which enhances a firms output (Roller, 2001). The speed of diffusion also causes an increase in market share of the firms in the communications mark et (Denni & Gruber, 2007). Few studies have empirically shown a relationship between th e development of communications network infrastructure and market performance. Fa ulhaber and Hogendorn ( 2000) found that broadband deployment is profitable in both large and sma ll areas. However, they found that broadband deployment is more profitable in large areas than small areas. On the contrary, Denni and Gruber (2007) suggest that as the diffusi on rate of broadband becomes higher, the market share of the incumbent decreases as other competitors enter the market. The aforementioned impacts of growth in ICT on firms market performance are independent of the impact of triple-play strategy, which may explain the change in market performance of triple-pla y service providers. Therefore this study controls for the impact of ICT. The Impact of Market Potent ia l on Market Performance Similar to the factor of ICT, this study assumes that market potential can relate to market performance of the triple-play providers besides the impact of the triple-play strategy (the primary focus of this study). Thus market pote ntial is another control factor for this study. Market potential is a po werful driver for a firm to offer new services because the firm can anticipate the generation of a new revenue stream (Hitt et al., 2005). As the size of market

PAGE 47

47 potential increases, the firm can enhance market performance either by increasing its market share or by reducing costs (Seo, 2008). In additio n, a firm can expect high er potential returns in larger markets as well as less risk for the inve stment (Hitt et al., 2005). For these reasons, communications firms prefer to offer their servic es in larger areas rath er than smaller areas (Grubesic &Murray, 2004). Therefor e, market potential influences firms market performance independently of the triple-pla y strategy, which may explain the change in market performance of triple-play service providers. Hence this study controls for th e impact of market potential. The Impact of Firm Size on Market Performance This study controls for firm size as well. It is assumed that firm size has a positive correlation with market power and/or the firms capabilities and resource s (Koski & Majumdarm, 2002). Generally, larger firms can enjoy the benef its of market performance. For example, the firm with the largest network is more profitable than firms with smaller competitive networks (Faulgaber & Hogendorn, 2000; Ford & Jackson, 1997) In that the event two competitors offer similar but incompatible products, the profitability (earni ngs) of the firm with the largest network is higher than firms with smaller compe titive networks (Faulhaber & Hogendorn, 2000; Liebowitz & Margolis, 2002). Small firms are less likely to respond to the dynamics of competition in the market and technological deve lopment, which implies that creating a large revenue base is unnecessary for them (Gentry & Jamison, 2005). In addition to profitability, studies have suggested that a la rger firm has the power to bloc k rivals entry (Savage & Wirth, 2005). Therefore, firm size influences firms ma rket performance indepe ndently of the tripleplay strategy, which may explain the change in market performance of triple-play service providers. Hence this study controls for the impact of firm size.

PAGE 48

48 The Impact of Cost on Market Performance This study also controls for the factor of fi r m-level cost. From the perspective of cost, marketing expenses and capital expenditures will be considered. Investment in marketing is highly likely to enhance a firms performance (S eggie et al., 2007). Marketing investment has a long-term positive impact on product choice (Jedid i et al., 1999). It also helps deliver better value to customers and enhance customer rela tionships (Achrol & Kotler, 1999). Consequently, marketing impacts a firms financial outcome (Amb ler et al., 2004; Jedidi et al., 1999). Capital expenditures generate advanced and innovative technologies. Cons equently, capital expenditures are likely to positively impact a firms perfor mance by increasing productivity and demand as well (Wellenius & Townsend, 2005). However, return on investment is not ac hieved in the short term, which may impact firms overall financial performance negatively. Due to direct impacts of marketing and capital expend itures on firms market performa nce but not jointly with the triple-play strategy, this study controls these two cost factors. Therefore, cost influences firms market performance independently of the triple-play strategy, wh ich may explain the change in market performance of triple-play service provide rs. Hence this study contro ls for the impact of cost. Resource-Based View (RBV) of Strategy The RBV approach em phasizes an individual fi rms unique capabilities and their impact on the firms business strategy in the market (Chan-Olmsted, 2006). In a competitive market, the RBV implied that the incumbent has resource, know ledge, or experience to use in defeating its competitors (either new or potential). Such re source, knowledge, or experience requires a sustained investment for a long time, which makes it difficult for the entrant to learn or imitate the resource, knowledge, or experience in a sh ort time (Dierickx & C ool, 1989). Barney (1991) suggested when resources are rare, difficult to imitate, and valuable, such resources lead to

PAGE 49

49 superior performance when used strategically. Thus the incumbent has a competitive advantage over its competitors that newly enter the mark et without resource/knowledge/experience, which leads to performance to the entrants (Maijoor & Van Witteloostuijin, 1996). In competing with the triple-play service, the triple-play provider between the cable and telephone industry has heterogeneou s resources. Traditionally cable companies have engaged in the TV business (video) in their franchising area s since they began their services, and have set the record in terms of the number of subs cribers among all multichannel video programming distributors (MVPD) (FCC, 2009). In addition, the cable industry also provides high-speed Internet service, either combining with video services or selling it independently using the network infrastructure. Recently, the cable i ndustry began expanding its business to telephony with Voice over Internet Protocol (VoIP) service. Telephone companies have their core service, telephone services, and offered hi gh-speed Internet services e ither adding it to telephony or supplying it separately. Now telephone companies have expanded to incl ude video services. Cable companies have competitive advantages over telephone companies in providing video services. A long history of video provi sion strengthens the cable industry. Cable companies can effectively segment their audi ence based on consumers preferences and set prices to maximize revenue. They can also effi ciently manage their own system based on their experience with trained human re sources. More importantly, cable operators have built up loyal customers and enjoyed scale economies based on ex isting subscribers. Cable systems also have their own video contents to offer or are closely affiliated with program providers, which lead to lower costs of programs. Telephone companies have served in relativel y larger service areas compared to cable firms that are locally oriented. As a result, they developed a relati vely larger network

PAGE 50

50 infrastructure of telephony and data services and enjoyed scal e economies. In addition, their telephone network is more reliable and stable than cables in deliver ing telephone services, which is mostly preferred to business customers who could lose significant amount of dollars if they lost communications services (Cra mpes & Hollander, 2006; Tomlinson, 2000). In sum, firms in each industry have long provi ded a traditional service, i.e., video service for cable and voice service for telephone firms, and they cumulate resources, experiences, or knowledge in operating this core service for a lo ng time. In recent times, they have expanded their businesses by entering each others service market, which requires a prolonged investment in developing the depth of managerial know-how and physical resources (Zollo & Winter, 2002). In the past, firms in each industry would demonstr ate superior performance in its core service relative to start-up services or rivals new se rvices because of a competitive advantage. However, with the triple-play strategy, firms are able to launch new services with an established core service customer base by offering price discount s or/and attractive value-added services (Lee, 2009), which may shorten the time needed to deploy new services (Kim & Kim, 2007). Based on the literature reviewed, the follo wing research question (RQs) is proposed: RQ10. Are there differences in market perfor mance among three services (i.e., video, voice, and data) between cable and telephone fi rms with the practice of triple-play? The Impact of a Triple-Play Bundle on Market Entry This section will review l iterature on the potential impact of the bundling strategy. The last focus of this study is to analyze the impact of triple-play bundles on entry in the communications market. In particular, this study will test the impact of a triple-play bundle on market entry in the high-speed fixed broadband ma rket. The most important reason is that highspeed fixed broadband is fundamental to the implementation of the triple-play strategy because

PAGE 51

51 new IP-based services within a triple-play bundle are provided through high-speed fixed broadband networks (Grubesic & Murray, 2004; Shy, 2001). Detailed justification will be provided in Chapter 3. Based on the literature review, this study w ill propose a relationship between triple-play bundling and market entry. Litera ture on other factors that af fect triple-play bundling or independently, the communications market will be reviewed for the purpose of isolating the impact of the triple-play practice. The Impact of Bundling on Market Entry Zimmer man (2007) predicted that competiti on among triple-play service providers will lead to structural changes in the telecommunications market in the future. From a regulatory perspective, whether the triple-play strategy can be used for the communications firm to leverage its dominant position in a service into a new market is an important issu e (Pernet, 2006). Peitz (2008) supported that bundling can effectively block entry whereas unbundling cannot do the same. Bauer (2005) suggested that triple-play providers are able to prevent single service providers from entering the market, consequently increasing their market share. The triple-play practice enhance competitive advantages of th e incumbents over new entrants with crossbranding and cross-promotions, which results in entry deterrence to some degree (ITU, 2006). Bundling theories have generally agreed with the role of bundling in deterring competitors entry. The firm which implements a bundling strate gy can take advantage of its rival. When a firm exercises a monopoly power for X product in X market, and it plans to introduce a new product, Y, in Y market, the firm ties X and Y t ogether in order to le verage its power in X market and in Y market. Consequently, the mono polist effectively deters rivals entry to Y market because bundling lessens competition in Y market (Whinston, 1990). In addition, a monopolist is able to maintain its monopoly pow er in a newly emerging market when the

PAGE 52

52 monopolist engages in bundling w ith complementary products. Th is is because the bundling strategy raises the entry cost of rivals and/or the strategy ge nerates positive netw ork externalities (Carlton & Waldman, 2002). In a te chnology market, if a firm offe rs a bundle that consists of innovation-based complementary components, the fi rm can stop a single complementary product seller from entering the market because such complementary components are valuable only when consumed altogether (Choi & Stefanadis, 2001). Furthermore, in a competitive market, if a firm has market power for two different products and bundles them together, it can effectiv ely prevent new single se rvice providers from entering the market, even without a discount on a bundle. This is becau se the entrants predict that the firm would lower the price of the bundle as soon as rivals ente r the market. Consequently, the firm is able to raise its profit (Nalebuff, 2004). Moreover, a seller is able to differentiate itself from other competitors by bundling differentiated products, which leads to an avoidance of aggressive competition. The seller can thereby ge nerate greater profits by preventing the price from dropping to unit cost (Carbajo et al., 1990). Based on the literature reviewed, the follo wing research question (RQ) is proposed: RQ11. Do triple-play bundles influence market entry in the broadband market? The Impact of Competition on Market Entry Com petition can influence market entry jointly with the practice of triple-play. This section will review literature on how competition generally influences market entry. This study will then propose a relationship between competition and mark et entry in consideration of the impact of triple-play bundles. Shy (2001) predicted that when all communication service providers that traditionally served a single market have the opportunity of offering telecommunica tions, broadcasting, and Internet service altogether, ther e would be no market foreclosure. As the communications market

PAGE 53

53 is open to competition, an increase in the num ber of providers is anticipated (Economides, 1996b). In large geographic areas wher e the degree of competition is hi gh, entry is more likely to take place (Cooper & Kimmelman, 1999; Grube sic & Murray, 2004; Hubbard & Lehr, 1998). The incentive to enter largely populated areas is th at firms are able to reach more subscribers at lower costs (Gabel & Gabel, 1997), and that usua lly in large areas, demand for the service is much higher (Grubesic & Murray, 2004). In a competitive market, communications firms ha ve incentives to extend their services to a new market (Koski & Kretschemer, 2005; Seo, 2008; Woroch, 2006). As technological convergence provides an opportunity to develop new services, communica tions firms rapidly step into a new technology mark et in order to attract greater segments of consumers by diversifying services (S eo, 2007; Seo, 2008). Another incentive to enter a new market is that an entry generates a new revenue stream and consequen tly recovers the firms losses in the current competitive market (Koski & Kretschemer, 2005 ; Woroch, 2006). Moreover, firms expect to leverage their existing market power in the new market (Nalebuff, 2004; Woroch, 2006). Cost reduction in offering services owi ng to the development of digital technology is also a big driver for market entry in the communications market (A ntonelli, 1997). In the network industry, as competition increases, greater entry occurs because a firm can reduce the fixed cost (Economides, 1996b). Denni and Gruber (2007) co nfirmed that in areas wher e the degree of competition among platforms is higher, the number of providers increases. In response to rivals entry, firms are likely to improve their facil ities in order to discourage entry because the number of entrants has an impact on profitability (Bourreau, 2004). Based on the literature reviewed, the follo wing research question (RQ) is proposed:

PAGE 54

54 RQ12. Does competition influence market entry in the broadband market when the tripleplay strategy is practiced? The Impact of Local Loop Unbundling Regulation on Market Entry Local loop unbundling (LLU) regulation can influence m arket entry jointly with the practice of triple-play. This section will review literature on how LLU regulation influences market entry. This study will then propose a relationship between LLU regulation and market entry in consideration of the impact of triple-play bundles. A primary goal of the Telecommunications Act of 1996 is to encourage the deployment of advanced telecommunication capabilities in a reasonable and timely manner by creating a free and competitive market where all telecommuni cation providers can compete altogether (Grubesic & Murray, 2004). In order to promote competition in the broadband market, the Act requires that the incumbent local exchange carri ers (ILECs) to unbundle thei r local loops and to lease them to competitors. Entrants can use variou s forms of resale of the ILECs facilities and LLU to access broadband customers (Bauer, 2005) LLU is designed to foster intra-modal competition for broadband service (OECD, 2003). In the broadband market, LLU regulation is impor tant in that it has impact on market entry and network investment (Baranes & Bourreau, 2005). In the provisi on of the triple-play service, it is essential to upgrade network infrastructure to transmit digitalized signal s (Bauer, 2006). Cable and telephone firms can enter each others ma rket via the upgraded network. They are also able to offer innovative and attr active services requiri ng high bandwidth to deliver over this network (Janssen & Mendys-Kamphrost, 2008). Ho wever, upgrading network infrastructure requires large investment, and the firms incentive to invest is influenced by LLU regulation in the broadband market.

PAGE 55

55 In general, mandatory unbundling delays entran ts incentive to invest in their facilities (Crandall et al., 2002). Lower LLU price discourages the ILECs in centive to invest in advanced infrastructure as well (Gabel & Huang, 2003; Pindyck, 2004). Hausman (2001) claimed that LLU regulation prevents the de ployment of DSL network faci lities, which results in the displacement by DSLs rival network, cable modem. In addition, the impact of LLU regulation on broadband deployment has been widely discussed after the Act of 1996. Typically, unbundling stimulates the deployment of advanced technology in the wireline sector (Bauer, 2005) Baranes (2005) asse rted that under the unbundling regime, potential entrants can offer hi gh bandwidth services without substantial investment in advanced infrastructure, which l eads to an increase in market entry in the broadband market. Nevertheless, Jorde et al (2000) contended that unbundling impedes facilities-based entry because delaying entrance will contribute to savings in technology costs and consumer learning cost on new services or new technologies. Bauer (2005) concluded that LLU regulation contri butes to increase mark et entry in a short time while it impedes the deployment of adva nced infrastructure by reducing investment. Based on the literature reviewed, the follo wing research question (RQ) is proposed: RQ13. Does LLU regulation influence market entry in the broadband market when the triple-play strategy is practiced? The Impact of Information Communica tion Technology on Market Entry Regardless of the impact of the triple-play practice, information communication technology (ICT) may influence market entry. Thus, it will have to be controlled in this study. This section will review the impact of IC T on entry in the communications market. Few studies have shown an empirical re lationship between the development of communications network infrastructure and mark et entry. Denni and Grube r (2007) suggest that

PAGE 56

56 as the diffusion rate of broadband becomes higher, the market share of the incumbent decreases as other competitors enter the market. In additi on, competitive entry is predicted to occur when the growth of communications network infrastr ucture is at mature stages (Faulhaber & Hogendorn, 2000). Denni and Gruber (2007) confirmed th at in areas where the diffusion rate of broadband is higher, the number of providers increases. The aforementioned impacts of growth in IC T on market entry are independent of the impact of triple-play strategy, which may explain the change in ma rket entry in the high-speed fixed broadband market. Therefore this study controls for the impact of ICT. The Impact of Network Effects on Market Entry This sec tion will review the literature on how network effects can influence entry in a telecommunications market. Network effects have the potential to influence market entry independently of the triple-play practice, which implies that it must also be controlled. The impact of network effects on market entry To analyze the telecommunications m arket, network effects are fundamentally important in that they directly relate to the success or failure of comm unications service (Oren & Smith, 1981). Network effects refer to the situation in wh ich the utility of each user increases when the number of users increases (Katz & Shapiro, 1985 ; Liebowitz & Margolis, 1994). Network effects are the inherent characteristi c of the network industry.13 The communications industry, including long-distance telephone, high-speed Internet, and IP-based tel ecommunications services, is 13 Shy (2001) pointed out network industries have four main characteristics: (1) complementarity, compatibility, and standards, (2) consumption externalities, (3) switching cost s and lock-in, and (4) significant economies of scale in production. Complementarity refers to consumers purchase systems rather than individual products. Compatibility and the same standards are conditions to offer complementary products. Consumption externalities mean the utility of each consumer is influenced by th e number of individuals using the same or complementary products. Sometimes they determine the adoption of new technology in th e communication market. Switching costs incur when the consumer switches to a different service or adopts a new technology. In order to prevent consumers from moving out, firms should lock in consumers in their system or technology. Significant economies of scale in production occur when the first copy bears large fixed costs while subsequent copies costs too low.

PAGE 57

57 highly related to network effects14 (Economides, 1996; Kim & Ki m, 2007; Shy, 2001; Soubeyran et al., 2007). In these industri es, a consumer decides to purch ase after considering product quality and price as well as the size of netw ork (Katz & Shapiro, 1985; Church & Gandal, 2005). Sometimes network effects determine the adop tion of new technology in the communications market. Shy (2001) proposed that as population grows, a new telecommunications technology is likely to diffuse rapidly because it is assumed that as the user size expand s, the utility of new adopters also increases. Network effects are relevant to entry in a communications market. Economides (1996b) showed a firm has incentive to invite rivals to the market where network effects are strong. Unlike a market without network effects, the firm s profit increases as the market size becomes larger in the presence of network effects. W ith rivals entry, the ma rket can support larger expected sales without the firm s own efforts to increase demand. Even when price competition takes place but product differentiati on is sufficient, the firm enc ourages rivals entry to expand the network size (Economides, 1996a). Kim (20 02) found under weak network effects, a firm promotes rivals entry when products are differen tiated. This is because rivals entry contributes to reduce a firms cost to expand the network size without cannibalizing its profits. However, under strong network effects, a fi rm would discourage rivals entr y regardless of the degree of product differentiation because rivals entry leads to decreases in the firm s profits (Kim, 2002). Gotz (2002) found that when the development of technology is evolut ionary, the level of entry is high because the entrants expect that they will increase revenue until the technology becomes irrelevant. When the number of entran ts no longer increases, incumbent firms reach a 14 The purpose of telephone services is to connect people to communicate with each other. Increasing in the number of phone users means that the user is able to communicate with more people using telephone. Thus the number of users increases, and the collective utility of users also increases. Similar to telephone service, high-speed Internet and Internet Protocol based telecommunications services have the characteristic of network externality in that the number of users affects th e utility of each user.

PAGE 58

58 break-even point. Beyond the break-even point, no more entry would occur. In addition, Faulhaber and Hogendorn (2000) found that competitiv e entry occurs over the base demand level in a broadband telecommunicati ons market being dominated by network effects. Shy (2001) discussed that firms continuously enter the mark et because entry into the telecommunications industry increases the utility of existing and new s ubscribers as well as the profits of the entering firm. Foros and Hansen (2001) showed that Inte rnet service providers can mitigate competition by increasing compatibility in the presence of network effects. The aforementioned impacts of network effect s on market entry are independent of the impact of triple-play strategy, which may explain the change in ma rket entry in the high-speed fixed broadband market. Therefore this study controls for network effects. The Impact of Market Potential on Market Entry This study assumes that market potential can relate to market entry, and so it considers market potential a control factor. Market potential is a critical determinant for new market entry because it provides an opportunity for firms to pursue greater market share and revenue (H ause & DuRietz, 1983; Burton et al., 1999). New entrants typically differentiate their services from the incumbent by catering to a specific segment of the consumer, a nd they expect to succeed in a market where demand continuously grows (Kaserman & Mayo, 2006). Bresnahan and Reiss (1990) confirmed that demand growth is a powerful attractor to ne w market entrants, and the number of entrants changes depending on market growth (Bresnahan & Reiss, 1991). In addition, Zolnierek et al. (2001) found that entry is likely to occur in urban areas rather than in rural in a local telephone market.

PAGE 59

59 The aforementioned impacts of market potentia l on market entry are independent of the impact of triple-play strategy, which may explain the change in ma rket entry in the high-speed fixed broadband market. Therefore this study cont rols for the impact of market potential. In sum, this study will examine the potential determinants of the us e of the triple-play strategy in cable and telephone industries. In addi tion, this study will explore how the triple-play strategy impacts market performa nce of triple-play providers by controlling for other factors. This study will also investigate how the triple -play bundle influences market entry in the broadband market due to potential effects of bun dling on deterring rivals entry. The next chapter will provide the proposed research models for this study.

PAGE 60

60 CHAPTER 3 RESEARCH MODELS This chapter discusses th e proposed research models for this study. Model I will present contributing factors of the use of the triple-pla y strategy. Model II will show the impact of the triple-play bundle on market performance of th e triple-play providers in addition to other relevant factors which will be controlled. Model III will examine the impact of the triple-play bundle on market entry in addition to other relevant factors whic h will also be controlled. Research Model I In Model I, this study will test factors contributing to the use of the triple-play strategy in the local US cable and telephone industry, respectiv ely. It will also com p are factors contributing to bundling between cable and telephone industries. Figure 3-1 provides an overview of the con ceptual framework. Essentially, the potential determinants of the use of the triple-play stra tegy are divided into two categories: market relevant factors and firm-specific factors. For ma rket factors, three variables are included: 1) competition, 2) information communication techno logy, and 3) market potential. Competition is divided into market competition and platform competition. In terms of the firm factor, firm size and cost are included. After testing contri buting factors in both the cable and telephone industries, this study will compare results to de termine which specific factors influence the use of a triple-play strategy differently between cable and telephone industries. In all three models, this study will capture changes in variables, including firm-speci fic variables, from 2000 to 2007. Brief restatements of the relationship between ea ch factor and the use of the triple-play bundling will be provided in the following section.

PAGE 61

61 Figure 3-1 Model I: The use of the triple-play strategy Market Factors Competition Pro-com petitive policies have facilitated comp etition in the telecommunications market since the Telecom Act of 1996. Recent technologi cal convergence has also generated additional competitive pressure for the cable and telephone i ndustries. In this situation, cable and telephone companies have emphasized the triple-play bund ling strategy to mitigate competition and to protect market share. Thus competition is an im portant market factor to assess the triple-play strategy. Information communication technology The developm ent of information communication technology is an essential external condition to introduce new advanced services wi thin the triple-play bund le. This is because components of the triple-play service are IP-based communications services, which are Notes: a) The relationship be tween relevant factors and the triple-p lay strategy in the cable industry; b) The relationship between rele vant factors and the triple-play st rategy in the telephone industry; c) Comparison of contributing factors be tween the cable and telephone industries Firm Size Cost FIRM FACTORS MARKET FACTORS Market Potential Competition Market Competition Platform Com p etition Information Communication Technology a ) a ) b ) b) The Use of a Triple-Play Strategy in Telephone Industry (2000-2007) The Use of a Triple-Play Strategy in Cable Industry (2000-2007) c ) Comparison of Factors between the Two Industries

PAGE 62

62 digitalized and carried over advanced communi cations networks. Thus, the development of communications network infrastructure is closely associated with a firms decision to offer the triple-play service. Market potential As m arket demand increases, communications firms have incentive to improve existing services and develop new services in order to increase their mark et share. Thus, the growth of market demand is closely associated with a firm s strategic decision to expand its business. Firm Factors Firm size Firm size has a positive correlation with the fi rms capabilities and reso urces. In terms of the provision of new services and/or complementar y services in a communications market, larger firms are more likely to be able to market many segments of the consumer markets and to aggressively expand their business. Cost From a cost perspective, this study sees two cost side factors to assess the triple-play strategy: marketing expenses a nd capital expenditures. Triple-pla y bundling is one example of a marketing strategy. Investment in the marketing of a new service is important because it has a positive effect on a brand equity and brand choice In addition, capital expenditures, which are required for the construction of or upgrades to a communications network in delivering the triple-play service, determine the timing of triple -play launch and mode of entry. Thus marketing expenses and capital expenditures are highly related to a firms strategic decision to offer new services.

PAGE 63

63 Strategic Orientation Porter (1980) proposed three ca tegories of generic strategies : overall cost leadership, differentiation, and focus. In the application of the generic strategies this study analyzes differences in the triple-play strategies betw een cable and telephone i ndustries after exam ining the relationship using firm-level factors. Research Model II In Model II, this study will exam ine the imp act of the triple-play bundle on triple-play providers market performance in the local US cable and telephone industry, respectively. To isolate the impact of bundling, this study will control for the general impact of other factors that are related to market performance, but not join tly with the triple-pla y. It will also compare contributing factors of bundling between the cable and telephone industries. Figure 3-2 provides an overview of the c onceptual framework. In Model II, market performance of the triple-play prov iders, as a dependent variable, is divided into three categories: subscriber performance, financial performance, and growth rates. Model II also divides the relevant factors into three sets: 1) stra tegic, 2) market, and 3) firm factor. For the strategic factors, the pr actice of the triple-play is in cluded as a primary variable in addition to switching costs. Regarding switching costs, only contra cts made by the triple-play providers will be taken into account because it is solely based on firms intention15 (Shapiro & Varian, 1999). For this reason, switching costs are included am ong the strategic factors. The market factors consist of three variables: 1) competition, 2) information communication technology, and 3) market potential. Competitio n is divided into market competition and 15 Sapiro and Varian (1999) classify switching costs into five types: 1) contracts, 2) training and learning, 3) data conversion, 4) search costs, and 5) loyalty costs. Out of these, only contracts are artificially designed by the firm.

PAGE 64

64 platform competition. Regarding the firm factor, firm size and cost are included. Factors such as information communication technology, market potential, firm size, and cost will be controlled. After testing the relevant factors of market performance of the trip le-play providers in consideration of control factors, this study will focus on how the use of the triple-play strategy influences market performance. In Model II, th is study will capture chan ges in variables from 2000 to 2007 as well as firm-specific variables. Br ief restatements of the relationship between each factor and market performance of the triple -play providers will be provided in the following section. Figure 3-2 Model II: The impact of a triple-p lay strategy on triple-play providers market performance Competition Market Competition Platform Competition ICT Market Potential MARKET FACTORS Cost Firm Size FIRM FACTORS STRATIGIC FACTORS Practice of Triple-Play Strategy (Primary Variable) Switching Costs Market Performance of the Triple-P lay Provider (2000-2007) Subscriber Performance Financial Performance Growth Rate: Financial Performance Growth Rate: Subscriber Performance

PAGE 65

65 Strategic Factors Practice of triple-play strategy Bundling has two opposite effect s on m arket performance of the bundle provider. One is the sorting effect when bundling reduces heteroge neous consumers valuations, which increase the sellers profits. Conversely, bundling has the business-steali ng effect when homogenization of customers intensifies price co mpetition and leads to a decrease in profits. As such, this study will examine how the triple-play strategy influences market performance. Switching costs Switching costs prev ent a subscriber from transf erring to a rivals service and consequently guarantee future profits in that a firm can obtain a reliable consumer base during the term of the contract. In general, the triple-play service is o ffered with a certain earl y termination fee and/or the term of contract. It is therefore proposed th at switching costs influenc e the triple-play service providers market performance jointly with the impact of triple-play bundle. Market Factors Competition In general, increasing market com petition infl uences the loss of a firms market share and/or profits. As suggested, competition within co mponents of the triple-play service is likely to influence the triple-play providers market pe rformance by cannibalizing each others market share. Platforms that deliver the triple-play bundle encounter competition as well. However, platform competition contributes to speedy broa dband deployment and thus it gradually expands market size, which may impact a firms market performance. Hence the impact of competition on the triple-play service provide rs market performance is comp licated. This study will consider the dynamics of competitive market conditions and propose a relationship between competition

PAGE 66

66 (market competition and platform competition) and the triple-play service providers market performance. Information communication technology This study controls the general im pact of information communication technology (ICT) on a firms market performance. The uses of communications technology being employed by the service providers increase producti vity because the new technology helps achieve efficiency in production and distribution processes in terms of time and costs. Because this growth occurs regardless of the triple-p lay strategy, this study controls for the general impact of information communication technology (ICT) on the firms market performance. Market potential This study controls the general im pact of mark et potential on a firms market performance. In general, when market potential is high, a firm is able to enha nce market performance either by increasing its consumer base or by reducing costs. This impact exists independently of the tripleplay strategy, which needs to be controlled. Firm Factors Firm size This study controls for firm size for the analys is of the triple-play providers financial performance, and corresponding growth rates. Firm size is positively correlated to market power, and larger firms can generally enjoy organizational efficiency. Consequently, it is more likely for larger firms to generate market share irrelevant of the impact of the triple-play strategy. Accordingly, this study controls for firm size. Cost This study also controls for the factor of fi r m-level cost. From a perspective of cost, marketing expenses and capital expenditures will be considered. Investment in marketing is

PAGE 67

67 highly likely to enhance a firms performance b ecause it helps retain customers in the long run. Capital expenditures are likely to positively impact a firms performance by increasing productivity and demand. On the contrary, dela yed return on investment may have negative impact on market performance. Because these two factors have direct impacts on market performance but not jointly with the triple-play strategy, this study controls for marketing expenses and capital expenditures. Resource-Based View (RBV) of Strategy The RBV approach focu ses on the individual firms unique capab ilities and their impact on the firms business strategy in the market. Using firm-level factors, this study analyzes differences in market performance among three se rvices (i.e., video, voice, and data) between cable and telephone firms with the practice of the triple-play from the RBV perspective. Research Model III In Model III, this s tudy will examine the imp act of the triple-pla y practicing on market entry in the communications market. This study will control for the general impact of other factors that are related to market entry, but not jointly with the triple-play strategy. Figure 3-3 provides an overview of the c onceptual framework. M odel III includes six explanatory variables: 1) practice of triple-play strategy as a prim ary variable, 2) competition, 3) information communication technology, 4) network effects, 5) market potential, and 6) local loop unbundling (LLU) regulation. Factors such as information communications technology, network effects, and market potential will be cont rolled. In Model III, this study will capture changes in variables from 2000 to 2007. As dependent variables, this study chooses to focus on the high-speed fixed broadband market where the potential impact of the triple-play bundle exists. The rationale for this choi ce is provided in the following s ection. Brief restatements of the relationship between each factor and market entry will be pr ovided after the section.

PAGE 68

68 Figure 3-3 Model III: The impact of th e triple-play bundle on market entry Rationale for the Analysis of Hi gh-Speed Fixed Broadband Market This study will test the im pact of a triple-pla y bundle on market entry in the high-speed fixed broadband market.16 The broadband market is chosen fo r its important role in analyzing triple-play services. The high-speed fixed broadb and market is one which makes it possible to deliver most advanced telecommunications services.17 In a sense, high-speed fixed broadband is fundamental to the implementation of the triple -play strategy because new IP-based services within a triple-pla y bundle are provided through high-speed fixed broadband networks18 (Grubesic & Murray, 2004; Shy, 2001). Sufficient diffusion of fixed broadband is required for the provision of enhanced IP-based co mmunication services (Lee & Brown, 2008). 16 The FCC defines high-speed as 200 kilobytes per second (Kbps) transmission speeds both downstream and upstream from provider to subscriber (G rubesic & Murray, 2004). All three plat forms cable and telephone use, i.e., cable modems, DSL, and fiber optic, have an ability to deliver advanced telecommunications services. 17 Advanced telecommunications service includes high-quality voice, data, graphics, and video telecommunications. These services can be provided via high-speed broadband (Grubesic & Murray, 2004; Telecommunications Act of 1996 Section 706). 18 In fact, new IP-based services can be delivered via wireless high-speed broadband (Shy, 2001). However, this study will only consider fixed broadband because a triple-p lay strategy is based on fixed-line broadband network (Shy, 2001).

PAGE 69

69 The second justification centers on data availability. Data on market entry (the number of providers from 2000 to 2007 as a measure) are only available for high-speed fixed broadband market, out of all three markets related to the triple-play bundle including video, telephone including local and long-distan ce, and broadband market. If other markets are considered, the proxy for market entry (for instan ce, the index of market concen tration) has to be utilized. However, the use of the proxy may generate a nother problem of interpretation due to the plausibility of othe r uncontrolled factors.19 In addition, because cable and telephone companies have offered data services before they provide triple-play bundles, changes in market trends incurred by the triple-play bundle woul d be relatively easily to trace. The third justification focuses on the attempt to examine the impact of the triple-play practice on entry in this market. Triple-play co mpetition takes place between cable and telephone companies having their own broadband network acce ss. Neither the cable nor telephone industry uses each others network. In addition, no stra tegic alliances between cable and telephone companies have been made to date. Consequently, it is more accurate to examine the competition structure between the two sectors. In the case of the video market where cable and DBS compete, telephone companies have started to provide video services. Some telephone companies provide video entertainment via IP TV and others allied with Direct Satellite Se rvice (DBS) providers, or both. In this case, the triple-play impact will be harder to assess because the performance of DBS is compounded as it is related to its own effort as well as synergies gained from the strategic 19 Market concentration and market entry is deemed to be positively correlated. That is, the higher market concentration, the lower level of market entry. Howeve r, there is the possibility of causation between the two variables. That is, high market concentration leads to lo wer level of market entry. When market concentration is used as the proxy of market entry, the impact of market c oncentration is not tested as well as other factors affecting market concentration, but not mark et entry can be interpreted as if they influence market entry.

PAGE 70

70 alliance with telephone companies, which needs to be isolated.20 Furthermore, the fixed broadband market is less likely to be regulated compared to telephony markets and, consequently, a relatively pure bundling effect can be interpreted. Thus, focusing on the fixed broadband market makes it possible to see clearer the impact of a triple-play strategy on market entry. In sum, this study will examine how th e triple-play strategy has influe nced market entry in the highspeed fixed broadband market stru cture at the state level. Potential Determinants of Market Entry Practice of triple-play strategy For regulators, whether the trip le -play strategy is pro-competitive is an important issue. The firm which implements a bundling strategy can take advantage of its rival in that (1) the firm can leverage market power from the first market to the second market, (2) the firm can lower the price of its rival, which will provide two produc ts as a discounted bundle to mitigate competition, and (3) the firm is able to maintain its market power in a newly emerging market by engaging in bundling of complementary products due to either en try cost or network externalities. In this context, there is the possibility that the triple-play bundle prevents single service providers from entering the market. Hence it is informative to examine the triple-play bundle on market entry. Competition As the communications m arket is open to competition, an increas e in the number of providers is anticipated. This happens because 1) communications firms have an incentive to 20 For example, the telephone market is highly regulated, an d thus it is probable that impacts of regulations outweigh the economic impact of a triple-play strategy. In the video market, the FCC has traditionally attempted to promote competition between cable television and satellite video providers. Although other independent video service providers exist in the market, the cable industry has recorded the largest market share in the video market being chased by DBS operators. With the recent emergence of a triple -play bundle in the video market, many telephone companies formed strategic relationships with DBS operato rs, and competed with cable television services. In addition, telephone companies provide IP TV and/or satellite video services. In this situation, competition structure is more complicated, and to see the pure impact of a triple -play strategy on entry into the video market may not be easy. In a fixed-line broadband market, however, the impact of a triple-play strategy will be explicit in that it is less likely to be regulated, and competition structure is more likely to be simple.

PAGE 71

71 extend their service to new market s to generate a new revenue str eam, 2) they can recover losses from the current competitive market by entering a new market, or 3) they can leverage market power in a new market. For platform competition, it is understood that in areas where the degree of competition among platforms is higher, the number of providers increases. As an important market condition, this study proposes the relationship between competition and market entry in consideration of the impact of the triple-play bundle. LLU regulation LLU regulation has an impact on market entry and network inve stm ent. In the provision of the triple-play service that consists of services requiring high bandwidth to deliver, firms need to invest in upgrading network infr astructure. The firms incentive to invest is influenced by LLU regulation in the broadba nd market. In general, LLU regulation contributes to an increase market entry in a short time while it impedes depl oyment of advanced infrastructure by reducing investment. Therefore, this study proposes a re lationship between LLU regulation and market entry in consideration of the im pact of the triple-play bundle. Information communication technology This study controls for the im pact of in formation communication technology on market entry. Competitive entry is predicted to occur when the growth of communications network infrastructure is in its mature stages. Because th is entry occurs independently of the impact of the triple-play bundle, this study controls for the ICT factor. Network effects A fir m has an incentive to invite rivals into the market where network effects are strong because the firms profit increases as the market size increases. Thus it is proposed that network effects have a positive relationship with market entry, independent of the impact of the tripleplay practice. Hence this study c ontrols for network effects.

PAGE 72

72 Market potential Market potential is another control f actor for this study. Market potential is positively correlated with new entry because high market pote ntial gives firms an opportunity to generate market share and profits. As such, this st udy controls for market potential factors.

PAGE 73

73 CHAPTER 4 METHODS This chapter discusses data, m easurement, st atistical methods and empirical models to examine factors contributing to the use of the triple-play strategy in the cable and telephone industries (Model I), triple-play providers mark et performance (Model II), and market entry in the US high-speed fixed broadband market (Model II I) outlined in previous chapters. This study will use secondary datasets and employs pane l data analyses in all three models. Secondary Analysis Secondary analysis is an efficient m ethod in th at 1) the researcher can obtain data easily and quickly, and 2) the dataset is less expe nsive to conduct a primary study (Beam, 2006; Stewart, 1984). In addition, a secondary analysis enables the researcher to expand the scope of the study (Stewart, 1984) and utilize internal financial data (Doyle & Firth, 2006). Although some secondary data is unreliable, the doc ument published by the government and/or the international organizations tends to provide re liable information (Doyle & Firth, 2006). For these reasons, in media economics and management of area, quantitative methods were employed in up to 60 percent of major journals such as the Journal of Media Economics and the Journal of Media Management over the past 15 years, and seconda ry analysis has become the most frequently used method of data coll ection in these areas (Beam, 2006). Panel Data Analysis A panel data analysis is a popular form of l ongitudinal data analysis in social science arenas (Yaffee, 2003). A panel data analysis is a m ethod that examines factors within multiple areas (or multiple firms), periodically observed over a defined time frame. A panel data analysis is helpful to observe firm behavior over time as well. Because a panel data analysis covers repeated observations of crosssections, it allows th e researcher to exam ine the dynamics of

PAGE 74

74 change even within a short time frame. Thus a panel data analysis will provide a rich interpretation of study factors due to its ability to cover both time and space dimensions (Kennedy, 2003; Yaffee, 2003). Most importantly, a panel data analysis help s control for unobserved heterogeneity in the model. Unobserved heterogeneity (including om itted variable bias) is a problem of nonexperimental research, which leads to unbiased es timators. With panel data analyses, however, it is possible to identify the true effect (Yaffee, 2003). Detailed regression methods using panel dataset are provided in the section of statistical anal ysis in each model. Sampling For Model I and Model II, the sam e datase t on firms was employed. Sampling began with stratification by firm size. For the cable industr y, this study selected the top 25 multiple system operators (MSOs, which represents approximately 90% of the nations cable households (Cable & TV Factbook, 2008)). For the telephone industry, a ll Incumbent Local Exchange Carriers (ILECs) and the large, facility-based Compe titive Local Exchange Carriers (CLECs) were included. All firm data sampled over 30 triple -play companies annually for 8 years (2000-2007). The number of observations was approximately 2 00 and varied by models. Detailed numbers of observations are provided in the results section of statistical analysis. For Model III, all US states were initially considered. However, several states were eliminated due to missing variables after careful considerations of the implications. Thus the dataset consisted of 42 states annually for 8 years (2000-2007) in this analysis.

PAGE 75

75 Measurement, Data, and Statistical Methods for Model I Drivers of the use of the triple-play strategy are analyzed in Model I. First, this study divides the triple-play providers into two industries: the cable and telephone industry. This study then ex amines which specific factors influence the use of the triple-play st rategy in each industry. To do so, this study employs panel data analyses using US datasets from 2000 to 2007. A couple of analyses are additionally performed in orde r to examine differences among firm-groups that are categorized by year(s) of the triple-play implementation using fi rm factors. Furthermore, this study compares how factors relate differently to the practice of the trip le-play strategy between the cable and telephone indus tries. For dependent variables, the unit of analysis is addressed at the firm level. Measurement and Data Sources Table 4-1 shows the variables, their m easurement, and the corresponding data sources for the use of a triple-play strategy. The data were mostly collected from th e annual reports of the Federal Communications Commission (FCC), each firms SEC filings, and the US Census Bureau. The dependent variable, the use of a triple-play strategy, is a continuous variable measuring how long triple-play providers have employed the triple-play strategy (unit: month). In a study that examined drivers of the use of the triple-play strategy, Seo (2007) used a dummy variable (1: Use of a tr iple-play strategy, 0: No triple-play strategy) to measure the use of the triple-play strategy in local cable systems with cross-sections. However, a dummy variable merely captures the existence of change, rather than the amount of changes of studying factors, which incurs loss of information. For this reason, use of continuous measures is recommended rather than dichotomous measures if data are available (Kerlinger & Pedhazur, 1973).

PAGE 76

76 Table 4-1 Variables, measurement, and data sources for Model I, the use of a triple-play strategy Variable Measurement Data sources DV Use of triple-play strategy (in cable and telephone, respectively) Continuous Months of the use of triple-play strategy SEC Filing IV Competition Market competition Market competition in video market Market competition in voice market Market competition in data market Competition with SMATV Continuous HHI in the MVPD market; HHI in the long-distance market; HHI in the ISPs market; % Share of SMATV in the MVPD market FCC biannual report on video market (2005); Television & Cable Factbook (2004-2008); FCC ARMIS database (2007); FCC annual report on telecommunications services (2000-2008); Market Research (2000-2008); Jupiter Research (2000-2008) Platform competition Continuous HHI for different fixed platforms FCC annual report on telecommunications services (2000-2008) ICT Continuous High-speed fixed broadband subscribers per 100 inhabitants in the service area of each firm FCC annual report on telecommunications services (2000-2008); SEC Filing Market potential Income Continuous Personal income per capita in the service area of each firm US Census Bureau (2000-2008); SEC Filing Population density Continuous Population density in the service area of each firm US Census Bureau (2000-2008); SEC Filing Firm size Continuous Basic service subscribers/lines SEC Filing Cost Continuous Ad & sales expense SEC Filing Continuous Capital expenditures SEC Filing As detailed in the literature review, there are proposed e xplanatory variables involving competition, information communication technol ogy, market potential, fi rm size, and cost. Competition includes two sub-categories: mark et competition and platform competition. To measure market competition, this study used th e Herfindahl-Hirshman-I ndex (HHI). The HHI is also commonly used to measure market co mpetition (Church & Gandal, 2006; Kaserman & Mayo, 2006), and this measure is commonly employed in the communications market (FCC, 2005; MacAvoy, 1995; Faulhaber & Hogendorn, 2000; Hovenkamp, 1990; Kaserman & Mayo, 2006). Video market competition was measured by the HHI in the Multiple Video Programming Distributors (MVPDs) Market. Voice market co mpetition was measured by the HHI in the longdistance market. This study initially consider ed both local market competition, and inter-

PAGE 77

77 platform competition in the telephone market (competition among wireline, wireless, and new technologies such as Internet telephony and VoIP). However it was difficult to measure local market competition and inter-pla tform competition due to the paucity of data. Thus this study only included competition in the long-distance market as the proxy of voice market competition. Data market competition was measured by the HHI in the Internet Service Providers (ISPs) market. Another video competition measurement was used: the percent share of SMATV in the MVPDs market to see how SMATV, as a cable competitor, has an impact on the strategic decision of the cable industry. For platform co mpetition, the HHI is also commonly used to measure platform competition (Church & Ganda l, 2006; Distaso et. al., 2006). Based on these justifications, this study also employed the HHI for market competition and the HHI between different fixed-broadband technol ogies for platform competition. ICT was measured by high-speed fixed broadba nd subscribers per 100 habitants in the service area of each firm. This is because the pr ovision of the triple-play service requires highspeed fixed broadband infrastructure. Market potential was measured by two variables: personal income per capita and population density per 100 habitants in the service area of each firm. Income per capita is used for the proxy of potential market demand (Karikari et al., 2003; Savege & Wirth, 2005; Seo, 2007; Seo, 2008) and population de nsity is the proxy of market size (Seo, 2007; Seo, 2008). Firm size was measured by basic service s ubscribers for the cable industry, and basic service lines for the telephone industry. The size of subscribers and lines are commonly used to measure firm size in the communication studie s (Karikari et al., 2003; Savege & Wirth, 2005; Seo, 2007; Seo, 2008). Cost was measured by ad vertising and sales expense (USD) and by capital expenditures (USD). Prior research meas ured cost as hourly wage (USD) for telecom

PAGE 78

78 equipment installers and repairers (Savage & Wirth, 2005; Seo, 2007; Seo, 2008). However, the triple-play strategy is one of marketing strategies which requires marketing expenses rather than physical labor expenses. In addition, firms ma y increase capital expend itures in implementing the triple-play because delivering parts of serv ices in the triple-play bundle requires higher capacity of bandwidth (Lee & Brown, 2008; Pernet 2006; Shim & Oh, 2006). Thus it would be appropriate to use marketing rele vant expenses, i.e., advertising and sales expense, and capital expenditures as the proxies of cost. Statistical Methods Regression analyses To exam ine the relationship between the proposed explanatory variable and the use of the triple-play strategy in each industry, this study employed two different regression analyses in order to obtain the robustness of the results: Pooled Ordinary Least Square (OLS) regression with time-fixed effect and the two-step model w ith firmand time-fixed effects. The pooled OLS regression method assumes homogeneous firm ch aracteristics among vari ables while it allows time-differences. For these reasons, pooled analys is is, in general, likely to suffer from heterogeneity bias of coefficient estimates. On the other hand, the two-step model allows both firmand time-differences among variables. One of benefits of two-step model is that it may control for unobserved firmand time-heterogeneity in the model, which helps obtain unbiased estimates compared to pooled OLS. In this study, pooled OLS regression, as an alternative method, was employed to show the robustne ss of results of the two-step model. This study conducted pooled OLS regression by taking time-fixed effects into account while not considering firm-fixed effects to exam ine the impact of proposed explanatory variables on the use of the triple-play st rategy in the cable and telephone industry. This study formulates the following regression model.

PAGE 79

79 Yit (USETPS) = 0 + 1(HHI_Video)t + 2(HHI_Voice)t + 3(HHI_Data)t + 4(HHI_Platform)t + 5(SMATV share)t + 6(Fixed Broadband)it + 7(Population Density) + 8 (Income)it + 9 (Firm Size)it + 10 (Marketing Expenses)it + 11(Capital Expenditures)it + 1t + it (4-1) where 0 is constant, and 1t represents time-dummies. In the empirical model, the dependent variable Yit is the months of triple-play use in firm i in year t in each industry. Expl anatory variables are the HHI for video market in year t, the HHI for voice market in year t, the HHI for data market in year t, the HHI for platform competition in year t, the percent penetration of SMATV in y ear t, high-speed fixed broadband subscribers per 100 habitants in the service area of firm i in year t, personal income per cap ita in the service area of firm i in year t, population density per 100 habitants in the service area of firm i in year t, firm size of firm i in year t, marketing expenses of fi rm i in year t, and capit al expenditures of firm i in year t. Another regression method this study employed is the two-step fixed effect model by including timeand firm-fixed effects at the same time. Because the fixed effect model cannot be used when firm-invariant variables (variables th at change only by time, not by firm) are included in the regression model (Yaffee, 2003), this st udy examined the proposed explanatory variables in separate two steps. First, this study conducted regression with the explanatory variables that varied by firm and time in addition to timeand firm-dummies representing timeand firm-fixed effects, respectively. In next step, this study estimated time-fixed effects using time-dummies in the previous step, and regressed firm-invariant variables (i.e., competition-related variables in this analysis) on this time-fixed effects estimate. This study formulates the following regression model in two steps.

PAGE 80

80 Step 1: Yit (USETPS) = 0 + 1(Fixed broadband)it + 2(Population Density) + 3(Income)it + 4(Firm Size)it + 5(Marketing expenses)it + 6(Capital Expenditures)it + 1t + Zt + it (4-2) where 0 is constant, 1t represents time-dummies, and Zt is firm-dummies. Step 2: t(Time Fixed-Effect Estimate) = 0 + 1(HHI_Video)t + 2(HHI_Voice)t + 3(HHI_Data)t + 4(HHI_Platform)t + 5(SMATV share)t + t (4-3) where 0 is constant, and t(Time Fixed-Effect Estimate) was estimated using 1t (timedummies) in Equation (4-2). In the empirical model, the dependent variable Yit is the months of triple-play use in firm i in year t in each industry. Expl anatory variables in Step 1 (i.e., Equation (4-2)) are the variables that change by firm and time. These variables are high-speed fixed broa dband subscribers per 100 habitants in the service area of firm i in year t, personal income per cap ita in the service area of firm i in year t, population dens ity per 100 habitants in the service area of firm i in year t, firm size of firm i in year t, marketing expenses of fi rm i in year t, and capit al expenditures of firm i in year t. Explanatory variables in Step 2 (i.e ., Equation (4-3)) are variables that change only by time including the HHI for video market in year t, the HHI for voice market in year t, the HHI for data market in year t, the HHI for platform competition in year t, and the percent penetration of SMATV in year t Multiple discriminant function an alysis As a part of an analysis of generic strategi es, this study additionally investigated specific firm characteristics in terms of firm size and co st factors by years of pr acticing the triple-play strategy. First, this study divides firms into seve ral groups on the basis of year(s) of the tripleplay strategy practice (i.e., 1 year of TPS, 2 years of TPS, 3 years of TPS, 4 years of TPS, and 5 years of TPS). Firms that do not employ the triple play strategy were eliminated. This study then conducted multiple discriminant function analysis to examine how the firm factors contribute to

PAGE 81

81 different group memberships. This analysis pr ovides insight on how di fferently the incumbent triple-play providers (say 4 or 5 years of TPS gr oups) and the start-up firms (say 1 or 2 years of TPS) use the triple-play strategy in terms of firm size and cost factors. Analysis of variance (ANOVA) analysis As a part of an analysis of generic stra tegies, this study also perform ed ANOVA to examine how the use of triple-pla y and industry (i.e., cable and telc o) are jointly associated with marketing expenses and capital expenditures. Fi rms were categorized by year(s) of the tripleplay practice (from non-triple-pla y-strategy to 4 or 5 years of tr iple-play). To control for firm size, this study divided marketing expenses by total operating revenue, and examined whether there are differences in marketing expenses as sh are of revenue between y ears of triple-play and between industries. This study also divided ca pital expenditures by tota l operating revenue, and examined whether there are differe nces in capital expenditures as share of revenue between years of triple-play and between industries. Measurement, Data, and Statistical Methods for Model II In Model II, this study exam ines the impacts of the triple-play stra tegy on the triple-play operators market performance, controlling for other variables. This study employs a panel data design and used the fixed effects regression estimation while accounting for timeand firmspecific features. This study uses the dataset of US triple-play companies from 2000 to 2007. In these cases, the dependent variables are addresse d at the firm level. Analyses are additionally performed in order to examine differences in market performance among firm-groups that are categorized by year(s) of the triple-play impl ementation. Furthermore, this study compares market performance between the ca ble and telephone industries. Table 4-2 shows the variables, their measurement, and the corresponding data sources for the use of the triple-pla y strategy. The data were mostly coll ected from the annual reports of the

PAGE 82

82 Federal Communications Commission (FCC), each firms SEC filings, and the US Census Bureau. For the dependent variable, this study investigates three categ ories of market performance: subscriber performance, financial performance, and growth rates. For subscriber performance, the numbers of each video, voice, and data subscr ibers were used as measures (Seo, 2007; Seo, 2008). For financial performance, the firms prof itability and revenue of video, voice, and data services were taken into account. For profitability, this study employed earnings before interest, taxes, depreciation, and amorti zation (EBITDA). EBITDA is commonly used to evaluate the firms profitability in media financial analyses (Ozanich, 2006). This is because other measures such as net income cannot reflect the actual cash flow incurred by operations in the media industry. First, communications firms need to in vest in technological infrastructure, which continues for a significant period of time. They also have a significant portion of noncash with respect to depreciation a nd amortization expenses. For the analysis of these firms, EBITDA is an appropriate measure for the firms profitability (Ozanich, 2006). For the growth rates, this study used the growth rate of the number of subscribers, revenue, and EBITDA as measures. Typicall y, the growth of market and/or technology is measured by the growth rates (Bell et al., 2001; Comin & Mulani 2005; Michalakelis et al., 2008). Thus, this study also used growth rates of the number of subscribers, revenue, and EBITDA as measures to capture the changing trend of subscriber and financial performance over time.

PAGE 83

83 Table 4-2 Variables, measurement, and data sources for Model II, market performance Variable Measurement Data sources DV Subscriber performance The number of video subscribers Continuous SEC Filing The number of voice subscribers Continuous SEC Filing The number of data subscribers Continuous SEC Filing Financial performance Profitabili ty: EBITDA Continuous SEC Filing Revenue: Video revenue Continuous SEC Filing Revenue: Voice revenue Continuous SEC Filing Revenue: Data revenue Continuous SEC Filing Growth rates Growth rate of the number of video subscribers Continuous SEC Filing Growth rate of the number of voice subscribers Continuous SEC Filing Growth rate of the number of data subscribers Continuous SEC Filing Growth rate of EBITDA Continuous SEC Filing Growth rate of video revenue Continuous SEC Filing Growth rate of voice revenue Continuous SEC Filing Growth rate of data revenue Continuous SEC Filing IV Practice of triple-play strategy Continuous Months of triple-play use SEC Filing Switching cost The term of contract Continuous Contract term: Month Annual reports; Each triple-play providers Website Terms & Conditions Early termination fee Continuous Early termination fee (USD) Annual reports; Each Triple-Play Providers Website Terms & Conditions Competition Market competition Market competition in video market Market competition in voice market Market competition in data market Competition with SMATV Continuous HHI in the MVPD market; HHI in the long-distance market; HHI in the ISPs market; % Share of SMATV in the MVPD market FCC biannual report on video market (2005); Television & Cable Factbook (2004-2008); FCC ARMIS database (2007); FCC annual report on telecommunications services (2000-2008); Market Research (2000-2008); Jupiter Research (2000-2008) Platform competition Continuous HHI for different fixed platforms FCC annual report on telecommunications services (2000-2008) ICT Continuous High-speed fixed broadband subscribers per 100 inhabitants in the service area of each firm FCC annual report on telecommunications services (2000-2008); SEC Filing Market potential Income Continuous Personal income per capita in the service area of each firm US Census Bureau (20002008); SEC Filing Population density Continuous Population Density in the service area of each firm US Census Bureau (20002008); SEC Filing Firm size Continuous Total operating revenue SEC Filing Cost Continuous Ad & sales expenses SEC Filing Continuous Capital expenditures SEC Filing

PAGE 84

84 Measurement and Data Sources As detailed in the literature review, explanat o ry variables involving the practice of tripleplay strategy as a primary explanatory vari able, switching costs, competition, information communication technology, market po tential, firm size, and cost are proposed. The main goal of this study is to examine the impact of the trip le-play bundle on market performance of the tripleplay providers. For the use of the triple-play strategy, th is study utilized the dependent variable in Model I as the primary explanatory variable in Model II. Thus, the measurement of this variable was months of the practice of the trip le-play strategy in each firm. For switching costs, contracts made by the firm are considered because a contract is the only switching cost the firm can create intentionally. Thus, switching costs were measured by the term of contract (month) and early termination fees that the contract-breaker should pay. Th e term of contract (month) can represent how long the triple-play providers can hold the subscriber. The longer the triple-play providers can retain the subscriber, the more securely the firm can enhance its market performance. In a similar vein, higher early term ination fees are likely to prevent a subscriber from switching. Thus this study used the term of contract and early termination fees as proxies of switching costs. For all competition variables, this study employed the Herfindahl-Hirshman-Index (HHI) as in Model I. Another video competition measurement was used: the percent share of SMATV in the MVPDs market. ICT was measured by hi gh-speed fixed broadband subscribers per 100 habitants in the service area of each firm. Mark et potential was measured by two variables: personal income per capita and population density pe r 100 habitants in the service area of each firm. Firm size was measured by total operating revenue (USD). Cost was measured by advertising and sales expenses (USD ) and by capital expenditures (USD).

PAGE 85

85 Statistical Methods Regression analyses To exam ine the impact of the triple-play strategy on triple-play providers market performance, this study employed the two-step model with firmand time-fixed effects. Unlike Model I, this study did not perform the pooled OLS as a validation method because a pair of dependent variables used in Model II, i.e., subscr iber and revenue, is expected to show similar results. Subscriber and revenue are usually clos ely related to market share, and consequently these two variables are presumed to have similar changing patterns. For this reason, the result of the impact of the triple-play st rategy on subscriber performance and the result on revenue can be compared with each other as validation. The fixed effect model cannot be used when firm-invariant variables (variables that change only by time, not by firm) are included in the regression model (Yaffee, 2003). Thus this study examined the proposed expl anatory variables in two separate steps. First, this study conducted regression with the expl anatory variables that changed by firm and time in addition to timeand firm-dummies representing timeand firm -fixed effects, respectively. In the next step, this study estimated time-fixed effects using time -dummies in the previous step, and regressed firm-invariant variables (i.e., competition-related variables in this analysis) on this time-fixed effects estimate. This study formulates th e following regression model in two steps. Step 1: Yit(Market Performance) = 0 + 1(Triple-play)it + 2(Early Termination Fee)it + 3(Contract Term)it + 4(Fixed broadband)it + 5(Population Density) + 6(Income)it + 7(Firm Size)it + 8(Marketing expenses)it + 9(Capital Expenditures)it + 1t + Zt + it (4-4) where 0 is constant, 1t represents time-dummies, and Zt represents firm-dummies. Step 2: t(Time Fixed-Effect Estimate) = 0 + 1(HHI_Relevant Market)t + 2 (HHI_Platform)t + 3(SMATV Share)t + t (4-5)

PAGE 86

86 where 0 is constant, t (Time Fixed-Effect Estim ate) was estimated using 1t (timedummies) in Equation (4-4), and 1(HHI_relevant market)t represents market competition that is closely related to the dependent variable at y ear t (e.g. for the dependent variable of video revenue, 1(HHI_relevant market)t is video market competition in year t). In the empirical model, the dependent variable Yit is the market performance in firm i in year t in each industry. Explanator y variables in Step 1 (i.e., Equa tion (4-4)) are the variables that change by firm and time. These variables are the pr actice of the triple-play strategy in firm i in year t, early termination fee in firm i in year t, contract term in firm i in year t, high-speed fixed broadband subscribers per 100 habitants in the servic e area of firm i in year t, personal income per capita in the se rvice area of firm i in ye ar t, population density per 100 habitants in the service area of firm i in year t, fi rm size of firm i in year t, marketing expenses of firm i in year t, and capital expenditures of firm i in year t. Expl anatory variables in Step 2 (i.e., Equation (4-5)) are the variables that change only by time such as the HHI for video market in year t, the HHI for voice market in year t, the HHI for data market in year t, the HHI for platform competition in year t, and the percent penetr ation of SMATV in year t Analysis of variance (ANOVA) analysis This study employed Analysis of Variance ( ANOVA) to exam ine differences in market performance among firms. In addition to prev ious regression analysis, this study further investigates the impact of the triple-play stra tegy by firms. Based on yea r(s) of the triple-play strategy practice, firms are divide d by six for the cable industry ( non-TPS, 1 year of TPS, 2 years of TPS, 3 years of TPS, 4 years of TPS, and 5 years of TPS group), and five for the telephone industry (non-TPS, 1 year of TPS, 2 years of T PS, 3 years of TPS, and 4 years of TPS group). ANOVA analyses were performed with each dependent variable in Model II in the cable and telephone industry, respectively. Next, for the dependent variables s howing statistically

PAGE 87

87 significant results, post hoc analyses were performed to compare every pair of group performance. ANOVA analysis This study employed Analysis of Varian ce (ANOVA) to exam ine which service contributes to total operating revenue the most by industry to test RBV research questions. First, this study divided firms into several groups on th e basis of year(s) of the triple-play (from nontriple-play-strategy (Non-TPS) to 5 years of TPS). Then ANOVA analyses were performed to examine differences in individual service revenue between the years of TPS groups and between cable and telephone industry. Measurement, Data, and Statistical Methods for Model III In Model III, this study exam ines the impact of the triple-play strate gy on market entry in the high-speed fixed broadband market. For statistic al analysis, this study employs a panel data design and uses the fixed effects regression estimation while accounti ng for time-specific features in order to look at th e entry mode in the US high-sp eed fixed broadband market. To do so, this study uses the US datasets at the state level from 2000 to 2007. In this case, the dependent variables are addr essed at the state level. Measurement and Data Sources Table 4-3 shows the variables, their m easurement, and the corresponding data sources for entry in the high-speed fixed broadband market. Th e data were mostly collected from the annual reports of the Federal Communications Commission (FCC) and the US Census Bureau. The dependent variable, entry into the highspeed fixed broadband market, was measured by the number of providers in the high-speed fi xed broadband market. Previous studies have considered the number of providers over a defined time frame to examine market entry (Alexander & Feinberg, 2004; Calzada & Vall etti, 2008; Denni & Gruber, 2007; Seim, 2006).

PAGE 88

88 Researchers sometimes focused on lines by platfo rm as a measure of market entry (Malecki, 2002; Loomis & Swann, 2005). However, lines by platform as a measure relates to network technology rather than the firm-relevant measure. Consequently, this st udy investigates the relationship between the triple-play strategy of the firm and market entry, the number of providers (firms) would be a better indicator than lines by platform. Table 4-3 Variables, measurement, and data sources for Model III, market entry Variable Measurement Data sources DV Entry in the high-speed fixed broadband market Continuous The number of providers of highspeed lines in the high-speed fixed broadband market FCC annual report on high speed services for Internet access (20002008) IV Practice of the Triple-P lay Strategy Continuous The number of subscribers in total triple-play market Parks Associate (2005); IDC (2008); Bharat Book Bureau(2009) Competitive Environment Competition Market competition in data market Continuous HHI in the ISPs market FCC biannual report on video market (2005); Television & Cable Factbook (20042008); FCC ARMIS Database (2007); FCC telecom services (2000-2008); Market Research (2001-2008); Jupiter Research (2000-2008) Platform competition Continuous HHI for different fixed platforms by state FCC annual report on telecom services (2000-2008) ICT Continuous High-speed fixed broadband subscribers per 100 inhabitants by state FCC annual report on telecom services (2000-2008) Market potential Income Continuous Personal income per capita by state US Census Bureau (2000-2008) Population density Continuous Population density by state US Census Bureau (2000-2008) Network effects Continuous Previous years fixed broadband subscribers per 100 inhabitants by state FCC annual report on high speed services for Internet access (19992007) LLU regulation price Continuous The monthly price of a Zone 1 UNE loop by state Gregg s(2001-2006) survey of unbundled network element prices in the US As detailed in the literature review, explan atory variables involving the practice of the triple-play strategy, competiti on, and LLU regulation are propos ed. Information communication technology, market potential, and network effects are control variables. To determine the impact of the triple-play strategy, this study used the number of subscrib ers in total triple-play market.

PAGE 89

89 For competitive environment, this study employed the HHI for market competition as a proxy. The HHI between different fixed-broadband technol ogies for platform competition by state is used as a measure of competition. ICT was meas ured as the penetration of high-speed fixed broadband per 100 habitants by state. Market pote ntial was measured by two variables: personal income per capital and population density per 100 habitants. These variables varied by states. To control for network effects, this study used the previous years high-speed fixed-broadband penetration by state. This was due to Economides and Himmelbergs (1995) implication that current subscription has a positive relationship with previous subscr iption due to network effects. LLU regulation price was measured by the monthl y price of a Zone 1 UNE loop by state. LLU regulation price was determined by different cust omer density zones. The densest zone, called Zone 1 in each state, is the most attractive loop for entrant firms because of its high proportion of business customers (Figueiredo & Edwards, 2005). For this reason, LLU regulation price is commonly measured by the monthly price of a Zone 1 UNE loop (Figueiredo & Edwards, 2007; Lee, 2006). Thus it would be appropriate to use the monthly price of a Zone 1 UNE loop as LLU regulation price. Statistical Methods Regression analyses To exam ine the impact of the triple-play st rategy on market entry in the high-speed fixed broadband market, this study condu cted pooled OLS regression by ta king time-fixed effects, but not state-fixed effects, into accoun t. Despite the nature of panel da ta used in this analysis, it was highly possible that explanatory variables did not vary enough for states across the sample. This study formulated the following regression model.

PAGE 90

90 Yit (ENTRY) = 0 + 1(HHI_Data)t + 2(HHI_Platform)t + 3(Fixed broadband)it + 4(Population Density) + 5(Income)it + 6 (Previous Fixed Broadband)it + 7 (LLU price)it + 1t + it (4-6) where 0 is constant, and 1t represents time-dummies. In the empirical model, the dependent variable Yit is the number of providers of high-speed fixed broadband in state i in year t. Explanatory variables are the HHI for data market in year t, the HHI for platform competition in year t, high-speed fixed broadband subscribers per 100 habitants in state i in year t, personal income pe r capita in state i in year t, population density per 100 habitants in state i in year t, previous high-speed fixed broadband subscribers per 100 habitants in state i in year t, and LLU regulation price in st ate i in year t.

PAGE 91

91 CHAPTER 5 RESULT This chapter provides the results of the data an alysis. First, the results of the practice of the triple-play strategy (Model I) are provided. Next the results of the impact of the triple-play strategy on the triple-play providers market performance (Model II) follow. Lastly, the results of the impact of the strategy on market entry in the high-fixed broadband market (Model III) are provided. For each model, this study first provides bot h the description of the secondary dataset and descriptive statistics. It then shows the results of statistical analyses for each model. The Use of Triple-Play In Model I, this study divides the triple-play providers into tw o indus tries: the cable and telephone industry. It then examines which specifi c factors influence the use of the triple-play strategy in each industry using regr essions. As a part of analysis of generic strategies, this study investigates specific firm charac teristics in terms of firm si ze and cost factors by years of practicing the triple-play stra tegy using multiple discriminant function analysis. This study additionally performed ANOVA to examine how the use of triple -play and industry are jointly associated with marketing expenses and capital expenditures as a part of analysis of generic strategies. Data and Descriptive Statistics for Model I In Model I, the drivers of the use of the triple-play strategy are analyzed. RQ1, RQ2, RQ3, RQ4, and RQ5 address factors c ontributing to the use of the triple-play. This study employed secondary data which consisted of cross-secti on (cross-firm data) and time variance (year from 2000 to 2007).

PAGE 92

92 A total of 196 observations were available for the first time. However, for all empirical models, some observations for a portion of the ex planatory variable were missing. Therefore, the number of observations employed for the real regression analysis was smaller than the number of observations being collected. For this reason, a total of 169 observa tions were available in both industries. For the cable industry, 91 observations were available and for the telephone industry, 78 observations were available. To examine the drivers of the use of the tr iple-play strategy in each industry, 10 different explanatory variables were empl oyed. Table 5-1 provides descriptiv e statistics for Model I of the cable industry, and Table 5-2 for the telephone i ndustry. The mean of the length of the use of triple-play was 15.96 months for the cable industry between 2000 and 2007, while telephone industry exercised the triple-play strategy fo r 8.26 months on average between 2000 and 2007. Table 5-1 Descriptive statistics for the use of triple-play in the cable industry Variables Mean (N=91) Standard Deviation Use of triple-play (month) 15.96 19.55 Competition: Multichannel video programming distributor market 1037.20 121.08 Competition: Long distance telephony market 1686.33 592.22 Competition: Internet service provider market 1189.59 462.38 Competition: Platform competition 4145.87 533.83 SMATV market share 1.41 .34 Infra: High-speed fixed broadband line per 100 inhabitants 29.96 50.07 Population density 147.35 116.82 Income per capita 32329.31 5088.49 Firm size: Basic service subscriber size 3993.14 5894.57 Cost: Advertising & marketing 891.78 1273.15 Cost: Capital expenditure 896.76 1252.61 Table 5-3 shows descriptive st atistics for Model I of both industries. The triple-play strategy has been used for 12.31 months on average in both cable and telephone industries between 2000 and 2007. Note that because these means were calculated considering the pretriple-play period (2000 to 2004), the real history of the triple-play strategy in both industries is

PAGE 93

93 much longer than the means in Table 5-1, 5-2, an d 5-3. This point will be addressed in the discussion section. Table 5-2 Descriptive statistic s for the use of triple-pla y in the telephone industry Variables Mean (N=78) Standard Deviation Use of triple-play (month) 8.26 13.82 Competition: Multichannel video programming distributor market 1044.61 122.05 Competition: Long distance telephony market 1661.13 575.35 Competition: Internet service provider market 1188.65 452.66 Competition: Platform competition 4116.08 548.94 SMATV market share 1.39 .34 Infra: High-speed fixed broadband line per 100 inhabitants 17.65 16.34 Population density 120.35 109.19 Income per capita 32890.10 6056.41 Firm size: Basic service lines 10731.12 17858.14 Cost: Advertising & marketing 3925.24 7285.76 Cost: Capital expenditure 2917.79 5421.02 Table 5-3 Descriptive statis tics for the use of triple-play in both industries Variables Mean (N=169) Standard Deviation Use of triple-play (month) 12.31 17.46 Competition: Multichannel video programming distributor market 1040.66 121.83 Competition: Long distance telephony market 1674.45 582.78 Competition: Internet service provider market 1189.14 456.62 Competition: Platform competition 4131.83 539.59 SMATV market share 1.40 .34 Infra: High-speed fixed broadband line per 100 inhabitants 24.20 35.58 Population density 134.70 113.79 Income per capita 32595.40 5561.31 Firm size: Basic service subscriber/lines 7153.74 13358.31 Cost: Advertising & marketing 2329.51 5307.33 Cost: Capital expenditure 1855.72 3963.49 Regression Analysis of the Use of Triple-Play This study employed two different regression anal yses in order to obtain robustness in the results: A two-step model with firmand time-fixed effects and pooled Ordinary Least Square (OLS) regression with a time-fi xed effect. Pooled OLS regression was employed to show the robustness of results of the two-step model as an alternative method.

PAGE 94

94 Initially, this study transformed several explan atory variables that were positively skewed using a logarithmic function. As a rule of sum, a variable whose z-score of skewness is above 2 needs to be transformed because this variable may generate biased results (Field, 2000). In addition, this study conducted correlation analysis in order to prevent po tential multicollinearity problems. Note that multicollinearity problems arise when explanatory variables are highly correlated. To assess the streng th of correlations, the .80 Pearson correlation criterion was employed. Eleven out of twelve explanatory variables were included except SMATV market share. The variable of SMATV market share show ed perfect collinearity with other competition variables and time-dummies, which need ed to be dropped in the analyses. Two-step model with firmand time-f ixed effects in the cable industry This study employed a two-step fixed effect m odel by including timeand firm-fixed effects at the same time. In the first step, th is study conducted regression with the explanatory variables that varied by firm and time in additi on to timeand firm-dummies representing timeand firm-fixed effects, respectiv ely (see Model (1) in Table 5-4) Next, this study estimated timefixed effects using time-dummies in the previ ous step, and then regressed firm-invariant variables (i.e., competition-related variables in this analysis) on this time-fixed effects estimate. The estimates were provided in Model (2) in Tabl e 5-4. Model (1) and Model (2) in the two-step model were statistically signi ficant at the 1 percent leve l, F(11, 11)=88.05, F(4, 76)=1579.33, respectively. Thus F tests support that firmand tim e-fixed effects are signif icant at the 1 percent level. Table 5-4 provides the results of the regressi on of triple-play use in the cable industry. RQ1a addresses the impact of competition on th e use of the triple-play strategy in the cable industry. The result shows that all market compet ition variables were statistically significant at the 1 percent level. As the video market becomes concentrated, cable companies are more likely

PAGE 95

95 to employ the triple-play strate gy. In other words, decreasing le vel of video market competition is associated with the likelihood of triple-play use. Note that market competition was measured by HHI (Herfindall-Hirschman Index), and thus positive signs of coefficients represent decreasing level of competition (also increasing level of market concentration). As the longdistance telephone (voice) market becomes con centrated (i.e., the market becomes less competitive), cable firms are more likely to use th e triple-play strategy. In addition, a lower level of platform competition is associated with the lik elihood of triple-play strategy use. However, as the level of data market competition (competitio n among ISPs) increases, cable firms are more likely to implement the triple-play strategy. Table 5-4 Regression analysis of the use of triple-play in the cable industry Variable Two-step model with firmand time-fixed effects Pooled OLS with timefixed effects Model (1) Model (2) Coefficient B t Statistics B t B t Video market competition .010 4.67* .037 1.04 Voice market competition .010 36.19* .028 2.74* Data market competition -.001 -4.43* -.007 -1.27 Platform competition .005 13.25* .024 2.57* SMATV penetration Dropped Dropped Fixed broadband ln 2.053 .20 ln 10.474 1.14 Population density .901 1.46 1.951 3.45* Income per capita .005 3.61* .006 3.74* Firm size ln 1.417 .64 -3.103e-4 -1.23 Marketing expenses ln -.626 -.20 ln .340 .12 Capital expenditures ln -.302 -.38 ln -.260 -.30 Constant -274.172 -2.76* -52.199 -14.85* -662.409 -5.18* Time fixed effect Yes Yes Firm fixed effect Yes No Observations 91 81 R square .912 .861 Statistically significant at the 5% level ln Log transformed variable Table 5-5 summarizes the relationship between competition and the use of the triple-play strategy in the cable industry. Cabl e firms are more likely to engage in the triple-play strategy as

PAGE 96

96 the level of market competition (video and voice) and platform competition decreases. On the other hand, they are more likely to use the triple -play strategy as the level of data market competition increases. Table 5-5 The impact of competition on cables triple-play Variable Cables triple-play Video market competition + Voice market competition + Data market competition Platform competition + Significant at the 5 percent level + Positive: Positive sign of coefficient, -: Negative sign of coefficient RQ3a addresses the impact of market potential on the use of the triple-play strategy in the cable industry. The result shows th at one of the market potential variables, income per capita, showed statistical significance at the 5 percent le vel. Cable companies are more likely to use the triple-play strategy when the income elasticity of demand is higher than when the income elasticity is lower. RQ2a addresses the impact of informati on communication technology on the use of the triple-play strategy, which was not statistically significant. None of firm variables including size (RQ4a), marketing expenses (RQ5a), and capit al expenditures (RQ5a) was statistically significant in these analyses. These results were consistent with the pooled OLS analysis (as a validation method) in the later section. Pooled ordinary least square regression with time-fixed effects in the cable industry This study conducted pooled OLS regression by taking tim e-fixed effects, but not firmfixed effects, into account to examine the impact of proposed explanatory variables on the use of the triple-play strategy in the cable industry. As suggested, pooled OLS regression was employed to show the robustness of resu lts of the two-step model as an alternative method. Table 5-4 provides the results of the regre ssion of the use of the triple-play strategy in the cable industry.

PAGE 97

97 The pooled OLS model was statis tically significant at the 1 percent level, F(11, 11)=1282.28, p<.001. For the market variable, voice market competition, platform competition, population density, and income per capita were statistically significant at the 5 percent level. All firm level variables, i.e., firm size and cost related factors, are also statistically significant at the 5 percent level. Two-step model with firmand time-fix ed effects in the telephone industry This study also em ployed a two-step fixed e ffect model by including timeand firm-fixed effects simultaneously. Model (1) and Model (2) in th e two-step model (in Table 5-6) were statistically significant at the 1 percent level, F(10, 10)=4833.10, F(4, 66)=32496.32, respectively. Thus F tests support that firmand tim e-fixed effects are signif icant at the 1 percent level. Table 5-6 provides the results of the regression of the trip le-play use in the telephone industry. RQ1b addresses the impact of competition on the use of the triple-play strategy in the telephone industry. The result show s that all market competition variables were statistically significant at the 1 percent level. As the level of video market competition increases, telephone firms are more likely to employ the triple-play strategy. Also, increasing competition in the longdistance (voice) market is associ ated with the likelihood of using the triple-play strategy. As the level of platform competition increases, telephone firms are more likely to offer the triple-play service as well. However for the data market (the ISP market), decreasing competition is related to the use of the triple-play strategy. Table 5-7 summarizes the relationship between competition and the use of the triple-play strategy in the cable industry. Telephone firms are more likely to engage in the triple-play strategy as the level of market competition (vid eo and voice) and platform competition increases.

PAGE 98

98 On the other hand, they are more likely to us e the strategy as the level of data market competition decreases. Table 5-6 Regression analysis of the use of triple-play in the telephone industry Variable Two-step model with firmand time-fixed effects Pooled OLS with timefixed effects Model (1) Model (2) Coefficient B t Statistics B t B t Video market competition -.027 -45.20* -.048 -4.90* Voice market competition -.005 -96.48* -.014 -3.58* Data market competition .003 28.52* .006 4.58* Platform competition -.012 -108.46* -.022 -4.49* SMATV penetration Dropped Dropped Fixed broadband 1.400 .26 -.328 -1.55 Population density .011 1.18 ln 4.473 .52 Income per capita ln 13.221 2.28* ln 13.914 5.60* Firm size ln 4.212 3.59* ln 3.002 3.54* Marketing expenses ln -5.989 -1.18 ln -8.383 -2.78* Capital expenditures ln 6.683 2.04 ln 6.720 2.59* Constant -176.246 -2.35* 82.975 81* -9.832 -.17 Time fixed effect Yes Yes Firm fixed effect Yes No Observations 78 67 R square .775 .711 Statistically significant at the 5% level ln Log transformed variable Table 5-7 The impact of competition on telcos triple-play Variable Telcos triple-play Video market competition Voice market competition Data market competition + Platform competition Significant at the 5 percent level + Positive: Positive sign of coefficient, -: Negative sign of coefficient RQ3b addresses the impact of ma rket potential on the use of th e triple-play strategy in the telephone industry. The result shows that one of the market potential variables, income per capita in telephone areas, showed statis tical significance at the 5 per cent level. Telephone companies

PAGE 99

99 are more likely to use the triple-play strategy when the income elasticity of demand is higher than when the income elasticity is lower. RQ2b addresses the impact of information communication technology on the use of the triple-play strategy, which was not statistically significant. RQ4b and RQ5b address the relationship between a firm variable and the use of the tripleplay strategy. RQ4b addresses the impact of fi rm size on the use of the triple-play strategy, which was statistically significant at the 5 percen t level. The larger firms are more likely to implement the triple-play strategy than the smaller firms. RQ5b addresses the impact of cost on the use of the triple-pla y strategy, which was not st atistically significant. Pooled ordinary least square regression with time-fixed e ffects in the telephone industry This study conducted pooled OLS regression by taking time-fixed effects, but not firmfixed effects, into account to examine the impact of proposed explanatory variables on the use of the triple-play strategy in the telephone indus try. As suggested, pool ed OLS regression was employed to show the robustness of results of the two-step model as an alternative method. Table 5-6 provides the results of the regression of triple-play use in the telephone industry. The pooled OLS model was statistically significant at the 1 percent level, F(10, 10)=5224.46, p<.001. For the market variable, all market competition variables were statistically significant at the 1 percent level. In additi on, two market potential variables, population density and income per capita in areas telephone firms operate, were st atistically significant at the 5 percent level. On the other hand, none of firm variables such as firm size, marketing expenses, and capital expenditures was statistically significant in this an alysis. These results were similar to the results of the two-step model.

PAGE 100

100 Multiple discriminant function an alyses of triple-play use RQ6 addresses the different st rategic behaviors of the trip le-play service providers by industry from a perspective of generic strate gies. This study conducted multiple discriminant function analyses in order to examine how firm variables (i.e., firm size, marketing expenses, and capital expenditures) play different roles in adoption of the triple -play strategy among firm groups categorizing by the years of the triple-play stra tegy practice by industry (i.e., cable vs. telco). Firms were categorized on the basis of year(s ) of triple-play use. Several cable firms showed a five-year-history of the triple-pla y implementation, which was the longest in the sample, while several telephone firms showed a f our-year-history of the triple-play strategy. As such, this study divided cable firms into five depending on how long they employed the tripleplay strategy from one to five years. For th e telephone industry, this st udy categorized telephone firms into four, from one to four years of practic e of the triple-play stra tegy. Firms that did not use the triple-play strategy in the sample were not included in these analyses. Cable industry: A multiple discriminant function analysis was performed using three firm variables as predictors of membership in five groups with the sample size of 49. Predictors were firm size, marketing expenses, and capital expenditures. Five groups were used in the analysis based on years of the triple-play practice from one year to five years. Table 5-8 illustrates the results of discriminant analysis. The results showed that the 76 percent of discriminant function scores of variances are not explained by group membership, Wilks Lambda = 0.76, and that three discriminant functions were calculated while they were not statis tically significant (the three values of 2 were not statistically significant at the 5 percent level). This may be due to the small sample size in each

PAGE 101

101 group as well as a small number of explanatory variables. The ca nonical correlation was 0.425, which was a moderate relationship between pred ictors and the first discriminant function. Table 5-8 Discriminant function analysis of the use of triple-play in the cable industry Variable 1yr Mean(SD(1)) N=12 2yr Mean(SD) N=12 3yr Mean(SD) N=12 4yr Mean(SD) N=9 5yr Mean(SD) N=4 Wilks Lambda S. D. (3) Coef in Func1 S. D. Coef in Func2 S. D. Coef in Func3 Firm size 4027.21 (5915.70) 4285.37 (6710.78) 4675.98 (7242.61) 5446.13 (8258.95) 4161.22 (6092.07) .99 -4.994 2.717 1.369 Marketing 572.70 (522.64) 976.70 (1400.76) 1143.80 (1673.99) 1478.14 (2116.80) 1244.74 (1299.33) .95 .822 2.651 -.297 Capital 814.10 (1097.87) 850.63 (1232.50) 1097.22 (1746.33) 1343.14 (1915.56) 1151.04 (1628.17) .98 4.429 -5.123 -.110 GrpCent (2) : Func1 -.494 -.348 .184 .535 .769 GrpCent : Func2 -.331 .371 -.011 .057 -.214 GrpCent : Func3 .010 -.023 .002 .069 -.123 Canonical R .425 .260 .050 Wilks from function1 to function 3=.762, (12) = 11.97, p > .05 (1) Standard Deviation (2) GrpCent: Group Centroids for the Discriminant Function 1 (3) Standardized Discriminant Coefficient When each independent variable is consid ered individually, marketing was found to be the most important out of three predictors, Wilks Lambda = 0.95. The 95% of variance in marketing expenses was not explained by gr oup membership. The second most important predictor was capital expenditures (Wilks Lam bda = 0.98), the last was firm size (0.99). Again, due to the small sample size in each group and th e small number of predictors, group differences existed for none of three predictors statistically (F values were not statistically significant at the 5 percent level). However, when all independent variables were considered together, di scriminant loadings of the first function suggested that the best pred ictor for distinguishing between firms with 1 year and 2 year of triple-play practi ce and firms with 3, 4, and 5 year s of triple-play practice was marketing expenses (discriminan t loading = 0.421), while the load ing values of the other two variables did not exceed .30. Note that a loadi ng value above .30 is considered substantive (Hair et al., 1998). Loadings of th e second function showed that marketing expenses discriminate

PAGE 102

102 the most between firms with 4 and 5 years of triple-play experience (loading=0.342). Loadings of the third function showed that firm size discri minate the most between firms with 1 year and firms with 2 years of triple -play experience (loading=0.991). The second best predictor was capital expenditures (loading=0.961) and mark eting expenses discriminate the least (loading=0.842). Telephone industry: A multiple discriminant function analysis was performed using three firm variables as predictors of membership in five groups with the sample size of 30. Predictors were firm size, marketing expenses, and capital expenditures. Four groups were used in the analysis based on years of the triple-play practi ce from 1 to 4 years. Table 5-9 illustrates the results of discriminant analysis. Table 5-9 Discriminant function analysis of the use of triple-play in the telephone industry Variable 1yr Mean(SD(1)) N=12 2yr Mean(SD) N=12 3yr Mean(SD) N=12 4yr Mean(SD) N=9 Wilks Lambda S. D. (3) Coef in Func1 S. D. Coef in Func2 S. D. Coef in Func3 Firm size 10641.38 (15967.96) 13266.47 (21516.73) 13580.52 (20918.77) 29528.82 (23117.29) .923 -4.994 2.717 1.369 Marketing 3808.29 (6885.03) 5046.42 (8601.58) 6150.58 (9704.71) 14987 (11291.37) .866 .822 2.651 -.297 Capital 2462.29 (4820.03) 3215.77 (5833.88) 4790.52 (7940.19) 12903.67 (9714.03) .806 4.429 -5.123 -.110 GrpCent (2) : Func1 -.301 -.263 .134 1.437 GrpCent : Func2 -.044 .132 -.120 .071 GrpCent : Func3 .016 -.008 -.015 .011 Canonical R .482 .108 .015 Wilks from function1 to function 3=.758, (9) = 7.050, p > .05 (1) Standard Deviation (2) GrpCent: Group Centroids for the Discriminant Function 1 (3) Standardized Discriminant Coefficient The results showed that the 75.8 percent of di scriminant function scores of variances are not explained by group membership, Wilks La mbda = 0.758, and that three discriminant functions were calculated while they were not statistically significant (the three values of 2 were not statistically significant at the 5 percent leve l on the bottom of the table). This may be due to the small sample size in each group as well as small number of explanatory variables. The

PAGE 103

103 canonical correlation was 0.482, which was a moderate relationship between predictors and the first discriminant function. When each independent variable is consider ed individually, capital expenditures were found to be the most important out of three predictors, Wilks Lambda = 0.806. The 80.6% of variance in capital expenditures was not expl ained by group membership. The second most important predictor was marketing expenses (Wilk s Lambda = 0.866), and the last was firm size (0.923). Again, due to the small sample size in e ach group and the small number of predictors, group differences existed for none of the three predictors statistically (F values were not statistically significant at the 5 percent level). However, when all independent variables we re considered all together, discriminant loadings of the first function suggested that th e best predictor for dist inguishing between firms with 1 year and 2 years of tr iple-play, and firms with 3 and 4 years was capital expenditures (discriminant loading = 0.0888). The second best predictor was marketing expenses (loading = 0.704) and firm size discriminated the least (loading = 0.505). Loadings of the second function showed that firm size discriminates the most betw een firms with 3 years and firm with 4 years of the triple-play practice (loading = 0.719). The se cond best predictor was marketing expenses (loading = 0.611) and capital expe nditures discriminated the least (loading = 0.448). Loadings of the third function showed that firm size discrimi nate the most between firms with 1 year and firms with 2 years of triple-play experience (loading = 0.478) while the loadi ng values of the other two variables did not exceed .30. Comparison of the cable and telephone industries: For the cable industry, marketing expenses were found to be the most important predictor of group memb ership out of three predictors (firm size, marketing expenses, and capital expenditure s). On the other hand, capital

PAGE 104

104 expenditures were found to be the most impor tant predictor of group membership in the telephone industry. Firm size also plays a role in discriminating all firm groups in the telephone industry. Analysis of Variance (ANOVA) anal yses of the use of triple-play RQ6 addresses the different strategic behavi ors of the triple-play service providers by industry from a perspective of generic strategies. In addition to multiple discriminant function analysis, this study conducted another Analysis of Variance (ANOVA) analyses in order to examine differences in marketing expenses a nd capital expenditures as a share of revenue between cable and telephone industries. The practi ce of the triple-play st rategy is jointly taken into account in these analyses. Basic and test statistics are provided in Table 5-10. Table 5-10 Results of ANOVA of cost differences by industry Dependent variable Source SS df MS F Variable Mean (SD) Yrs of TPS Indu Capital expenditures as a share of revenue Yrs of TPS .012 4 .003 .549 1 Telco .152 (.067) Cable .214 (.074) Industry .036 1 .036 6.47* 2 Telco .150 (.064) Cable .186 (.051) Yrs of TPS industry .004 3 .001 .26 3 Telco .158 (.093) Cable .187 (.054) Error .406 72 .006 4 Telco .156 (.023) Cable .220 (.127) 5 Telco n/a Cable .158 (.039) Marketing expenses as a share of revenue Yrs of TPS .008 4 .002 .159 1 Telco .194 (.074) Cable .286 (.140) Industry .013 1 .103 8.552* 2 Telco .199 (.078) Cable .293 (.117) Yrs of TPS industry .002 3 .001 .065 3 Telco .199 (.083) Cable .283 (.116) Error .847 70 .012 4 Telco .239 (.115) Cable .296 (.116) 5 Telco n/a Cable .314 (.133) Statistically significant at the 5% level For capital expenditure as a share of revenue, i ndustry was statistically significant at the 5 percent level. As seen in Figure 5-1, the cable industry is more likely to make an investment in their platforms than the telephone industry. The cable industry showed ir regular patterns of capital expenditures while the telephone industry has gradually increased the amount of capital

PAGE 105

105 expenditures as they implement the triple-play stra tegy. In this analysis, the joint relationship of years of the triple-pla y practice and industry w ith capital expenditure was not statistically significant. Also the main effect of years of the triple-play practice was not statistically significant. Figure 5-1 Capital expenditures as share of revenue by industry For marketing expenses as a share of revenue, industry was statistically significant at the 5 percent level. As seen in Fi gure 5-2, the cable industry is more likely to spend money for marketing and advertising than the telephone in dustry. The cable industry continuously increases its marketing and advertising costs while the te lephone industry tends to increase marketing expenses after 3 years of the triple-play strategy (d espite of insignificant re sults of the interaction between industry and years of the triple-play prac tice). In this analysis, the joint relationship of years of the triple-pla y practice and industry w ith marketing expenses was not statistically significant. Also, the main effect of years of the triple-play practice was not statistically significant. 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 1yr2yr3yr4yr5yrCapital Expendirues shareYears of TPS Telco Cable

PAGE 106

106 Figure 5-2 Marketing expenses as share of revenue by industry Summary Table 5-11 consolidates the resear ch questions of this study (the use of triple-play) and the results of corresponding research questions. Fo r the cable industry, com petition (RQ1a) and market potential related factors (R Q3a) are associated with the us e of the triple-play strategy. For the telephone industry, competition (RQ1b), marke ting potential (RQ3b), and firm size (RQ5b) are associated with the use of th e triple-play strategy. In addition, there is the different strategic orientation between the cable and telephone industries in the provisi on of the triple-play service. Detailed findings are provided in Table 5-11. 0 0.05 0.1 0.15 0.2 0.25 1yr2yr3yr4yr5yrMarking ShareYears of TPS Telco Cable

PAGE 107

107 Table 5-11 Summary table fo r the use of triple-play Research question Result I. Drivers of the use of trip le-play Impact Detailed finding Cable RQ1a. Does competition influence the use of the tripleplay strategy in the cable industry? Yes 1. Cable firms tend to use triple-play as video market competition becomes less competitive. 2. Cable firms tend to use triple-play as voice market competition becomes less competitive. 3. Cable firms tend to use triple-play as data market competition becomes more competitive. 4. Cable firms tend to use triple-play as platform competition becomes less competitive. RQ2a. Does the development of communications network infrastructure influenc e the use of the triple-play strategy in the cable industry? No RQ3a. Does market potentia l influence the use of the triple-play strategy in the cable industry? Yes 1. Cable firms tend to use triple-play as income per capita in cable firms areas increases. RQ4a. Does the firm size influence the use of the tripleplay strategy in the cable industry? No RQ5a. Does cost influence the use of the triple-play strategy in the cable industry? No Telco RQ1b. Does competition influence the use of the tripleplay strategy in the telephone industry? Yes 1. Telephone firms tend to use triple-play as video market competition becomes more competitive. 2. Telephone firms tend to use triple-play as voice market competition becomes more competitive. 3. Telephone firms tend to use tr iple-play as data market competition becomes less competitive. 4. Telephone firms tend to use triple-play as platform competition becomes more competitive.

PAGE 108

108 Table 5-11 Continued Research question Result I. Drivers of the use of triple-play Impact Details TelcoRQ2b. Does the development of communications network infrastructure influen ce the use of the triple-play strategy in the telephone industry? No RQ3b. Does market potential influence the use of the triple-play strategy in the telephone industry? Yes 1. Telephone firms tend to us e triple-play as income per capita in telephone firm s areas increases. RQ4b. Does the firm size influence the use of the tripleplay strategy in the telephone industry? Yes 1. The larger telephone firms are more likely to use triple-play than smaller firms. RQ5b. Does cost influence the use of the triple-play strategy in the telephone industry? No II. Strategic orientation RQ6. Is there a difference in strategic behaviors of the triple-play service providers by industry in terms of the generic strategies? Yes 1. Cable firms are more likely to spend marketing and capital expenditures than telephone firms.

PAGE 109

109 Market Performance of the Triple-Play Providers In Model II, this study divide s the triple-play providers in to two industries: cable and telephone industry. Then, using regression, this study exam ines the impacts of the triple-play strategy on the triple-play providers market performance, contro lling for other variables. RQ7, RQ8, and RQ9 address the impact of the triple-p lay strategy. Market performance here consists of three categories: subscriber performance, financial performa nce, and growth rates. Each category has several dependent variables. This st udy additionally employed Analysis of Variance (ANOVA) to examine differences in market pe rformance among firms that are categorized by the years of the triple-play practice. This study also conducted ANOVA to test RBV related research questions (i.e., RQ10). Data and Descriptive Statistics for Model II In Model II, this study employed secondary da ta which consisted of cross-section (crossfirm data) and time variance (year from 2000 to 2007). The data were mostly collected from the Federal Communications Commission (FCC), each firms SEC filings, and the US Census Bureau. A total of 196 observations were available for the first time. However for all empirical models, some observations for a portion of the ex planatory variable were missing. Therefore the number of observations employed for the real regression analysis was smaller than the number of observations being collected. The number of obs ervations was not equal across the dependent variables. For this reason, the total number of observations in each analysis is different depending on various dependent variables used in Model II. For the cable industry, around 80 observations were employed and for the te lephone industry, around 75 observations were employed. The exact number of observations used in each analysis is addressed in Table 5-12 and Table 5-13.

PAGE 110

110 To examine the impacts of the triple-play strategy on the triple-play providers market performance, 12 different explanat ory variables (6 main variables and 6 control variables) were employed with 14 dependent variables. Table 5-12 illustrates descriptive statistics for Model II of the cable industry, and Table 513 for the telephone industry. For the cable industry, the mean of vide o subscriber size was 5596.31 between 2000 and 2007. The means of voice and data subscriber size were 280.05 and 1123.11, respectively. In addition, the means of video, voice, and data revenue were 2293.57, 124.67, and 516.44, respectively. The mean of EBIT DA was 1085.37. For the growth rate s, the means of video, voice, and data subscriber growth rate were 28.56, 58.37, and 78.85, respectively. The means of video, voice, and data revenue grow th rate were 36.25, 132.79, and 63.72, respectively. The mean of EBITDA growth rate was 629.58. For the telephone industry, the mean of vi deo subscriber size was 56.88 between 2000 and 2007. The means of voice and data subscriber size were 10326.39 and 1033.25, respectively. In addition, the means of video, voice, and data revenue were 180.79, 6166.03, and 2680.86, respectively. The mean of EBIT DA was 4501.10. For the growth rate s, the means of video, voice, and data subscriber growth rate were 19.45, 42.49, and 63.96, respectively. The means of video, voice, and data revenue grow th rate were 1.06, 25.73, and 74.35, respectively. The mean of EBITDA growth rate was 42.19.

PAGE 111

111 Table 5-12 Descriptive statistics for the ma rket performance in the cable industry DV Mean (SD) N Video sub 5596.31 (8174.20) 91 Voice sub 280.05 (485.20) 82 Data sub 1123.11 (1951.45) 91 Video rev 2293.57 (3194.65) 83 Voice rev 124.67 (212.81) 83 Data rev 516.44 (903.69) 83 EBITDA 1085.39 (2131.90) 90 Variables Mean (S.D.) Mean (S.D.) Mean (S.D.) Mean (S.D.) Mean (S.D.) Mean (S.D.) Mean (S.D.) Practice of triple-play (month) 8.00 (12.44) 8.00 (12.44) 8.00 (12.44) 8.59 (12.80) 8.59 (12.80) 8.59 (12.80) 8.00 (12.44) Early termination fee 66.45 (75.55) 66.45 (75.55) 66.45 (75.55) 72.74 (76.10) 72.74 (76.10) 72.74 (76.10) 66.45 (75.55) Contract term (month) 9.67 (8.15) 9.67 (8.15) 9.67 (8.15) 10.58 (7.93) 10.58 (7.93) 10.58 (7.93) 9.67 (8.15) Competition: Multichannel video programming distributor market 1037.20 (121.08) 1037.22 (121.18) 1037.20 (121.08) Competition: Long distance telephony market 1699.40 (630.48) 1696.59 (628.82) 1699.40 (630.48) Competition: Internet service provider market 1247.82 (489.82) 1248.24 (490.87) 1247.82 (489.82) Platform competition 4250.33 (481.39) 4250.33 (481.39) 4250.33 (481.39) 4252.27 (480.96) 4252.27 (480.96) 4252.27 (480.96) 4250.33 (481.39) SMATV market share 1.41 (.34) 1.41 (.34) 1.41 (.34) High-speed fixed broadband line per 100 inhabitants 23.16 (32.30) 23.16 (32.30) 23.16 (32.30) 3658.75 (5124.34) 3658.75 (5124.34) 3658.75 (5124.34) 23.16 (32.30) Population density 14 8.43 (116.05) 148.43 (116.05) 148.43 (116.05) 156.66 (118.17) 156.66 (118.17) 156.66 (118.17) 148.43 (116.05) Income per capita 30923.23 (4222.32) 30923.23 (4222.32) 30923.23 (4222.32) 31618.18 (3676.05) 31618.18 (3676.05) 31618.18 (3676.05) 30923.23 (4222.32) Total operating revenue 3382.14 (4978.01) 3382.14 (4978.01) 3382.14 (4978.01) 3658.75 (5124.34) 3658.75 (5124.34) 3658.75 (5124.34) 3382.14 (4978.01) Marketing expenses 787.20 (1049.62) 787.20 (1049.62) 787.20 (1049.62) 851.47 (1076.47) 851.47 (1076.47) 851.47 (1076.47) 787.20 (1049.62) Capital expenditures 834.88 (1064.72) 834.88 (1064.72) 834.88 (1064.72) 903.66 (1089.37) 903.66 (1089.37) 903.66 (1089.37) 834.88 (1064.72) DV Mean (SD) N Growth: Video sub 28.56 (85.96) 89 Growth: Voice sub 58.37 (126.43) 88 Growth: Data sub 78.85 (120.33) 87 Growth: Video rev 36.25 (156.57) 78 Growth: Voice rev 132.79 (448.54) 80 Growth: Data rev 6 3.72 (88.42) 79 Growth: EBITDA 629.58 (3931.98) 90 Variables Mean (S.D.) Mean (S.D.) Mean (S.D.) Mean (S.D.) Mean (S.D.) Mean (S.D.) Mean (S.D.) Practice of triple-play (month) 8.20 (12.53) 8.31 (12.58) 8.41 (12.63) 9.22 (13.04) 8.96 (12.94) 9.08 (12.99) 8.10 (12.49) Early termination fee 67.82 (75.98) 68.37 (76.32) 68.28 (76.82) 75.52 (77.60) 74.44 (77.00) 75.51 (77.03) 66.97 (75.88) Contract term (month) 9.68 (8.12) 9.58 (8.12) 9.55 (8.17) 10.74 (7.96) 10.61 (7.95) 10.76 (7.90) 9.56 (8.14) Competition: Multichannel video programming distributor market 1039.92 (121.31) 1044.67 (122.02) 1038.24 (121.48) Competition: Long distance telephony market 1668.89 (618.42) 1652.16 (597.82) 1683.10 (617.04)

PAGE 112

112 Table 5-12 Continued Growth: Video sub Growth: Voice sub Growth: Data sub Growth: Video rev Growth: Voice rev Growth: Data rev Growth: EBITDA Variables Mean (S.D.) Mean (S.D.) Mean (S.D.) Mean (S.D.) Mean (S.D.) Mean (S.D.) Mean (S.D.) SMATV market share 1.40 (.34) 1.38 (.34) 1.40 (.34) High-speed fixed broadband line per 100 inhabitants 23.47 (35.67) 23.38 (35.90) 23.62 (36.07) 18.62 (27.95) 18.19 (27.67) 18.43 (27.80) 23.23 (35.51) Population density 151.14 (116.06) 153.06 (115.54) 125.86 (116.28) 159.97 (120.32) 160.62 (118.68) 160.10 (119.45) 150.28 (115.58) Income per capita 30960.34 (4268.99) 30951.29 (4295.86) 30966.64 (4321.87) 31846.99 (3679.71) 31763.35 (3663.46) 31817.31 (3661.38) 30936.40 (4247.28) Total operating revenue 3462.84 (5014.84) 3502.65 (5031.71) 3471.35 (5060.08) 3772.21 (5253.65) 3756.12 (5199.18) 3802.79 (5221.72) 3420.76 (4997.19) Marketing expenses 805.29 (1056.65) 814.14 (1060.54) 811.48 (1067.23) 876.31 (1101.94) 867.46 (1090.09) 878.35 (1094.06) 795.79 (1053.38) Capital expenditures 854.78 (1070.72) 864.90 (1073.84) 854.87 (1077.20) 908.50 (1101.92) 909.66 (1092.28) 918.91 (1097.37) 844.68 (1067.75) Table 5-13 Descriptive statistics for the mark et performance in the telephone industry DV Mean (SD) N Video sub 56.88 (213.74) 79 Voice sub 10326.39 (17236.44) 79 Data sub 1033.25 (2203.06) 79 Video rev 180.79 (879.91) 77 Voice rev 6166.03 (9543.16) 77 Data rev 2680.86 (4686.76) 77 EBITDA 4501.10 (8884.43) 78 Variables Mean (S.D.) Mean (S.D.) Mean (S.D.) Mean (S.D.) Mean (S.D.) Mean (S.D.) Mean (S.D.) Practice of triple-play (month) 2.47 (6.07) 2.47 (6.07) 2.47 (6.07) 1.81 (4.93) 2.15 (5.50) 2.15 (5.50) 2.47 (6.07) Early termination fee 84.54 (69.92) 84.54 (69.92) 84.54 (69.92) 62.34 (59.55) 82.82 (62.92) 82.82 (62.92) 84.54 (69.92) Contract term (month) 9.71 (8.10) 9.71 (8.10) 9.71 (8.10) 7.70 (7.48) 9.67 (8.15) 9.67 (8.15) 9.71 (8.10) Competition: Multichannel video programming distributor market 1049.32 (122.49) 1055.30 (122.39) 1049.32 (122.49) Competition: Long distance telephony market 1624.67 (582.47) 1625.20 (586.85) 1624.67 (582.47) Competition: Internet service provider market 1271.60 (480.80) 1272.72 (484.34) 1271.60 (480.80) Platform competition 4238.76 (506.92) 4238.76 (506.92) 4238.76 (506.92) 4239.19 (517.15) 4253.80 (495.22) 4253.80 (495.22) 4238.76 (506.92) SMATV market share 1.37 (.34) 1.35 (.34) 1.37 (.34) High-speed fixed broadband line per 100 inhabitants 14.37 (11.54) 14.37 (11.54) 14.37 (11.54) 14.60 (12.04) 14.17 (11.51) 14.17 (11.51) 14.37 (11.54) Population density 11 1.10 (69.49) 111.10 (69.49) 111.10 (69.49) 118.66 (71.44) 110.60 (69.89) 110.60 (69.89) 111.10 (69.49) Income per capita 31273.94 (3389.51) 31273.94 (3389.51) 31273.94 (3389.51) 31376.04 (3586.34) 31202.15 (3362.60) 31202.15 (3362.60) 31273.94 (3389.51) Total operating revenue 13713.12 (23125.44) 13713.12 (23125.44) 13713.12 (23125.44) 6551.32 (14595.34) 12601.67 (21391.83) 12601.67 (21391.83) 13713.12 (23125.44)

PAGE 113

113 Table 5-13 Continued Video sub Voice sub Data sub Video rev Voice rev Data rev EBITDA Variables Mean (S.D.) Mean (S.D.) Mean (S.D.) Mean (S.D.) Mean (S.D.) Mean (S.D.) Mean (S.D.) Marketing expenses 3814.36 (7014.35) 3814.36 (7014.35) 3814.36 (7014.35) 1295.42 (3076.33) 3498.83 (6563.08) 3498.83 (6563.08) 3814.36 (7014.35) Capital expenditures 2613.56 (4799.95) 2613.56 (4799.95) 2613.56 (4799.95) 904.23 (2106.73) 2397.33 (4490.10) 2397.33 (4490.10) 2613.56 (4799.95) DV Mean (SD) N Growth: Video sub 19.45 (96.85) 76 Growth: Voice sub 42.49 (249.33) 76 Growth: Data sub 63.69 (155.98) 76 Growth: Video Rev 1.06 (7.70) 62 Growth: Voice Rev 25.73 (124.28) 72 Growth: Data Rev 74.35 (222.61) 74 Growth: EBITDA 42.19 (210.48) 77 Variables Mean (S.D.) Mean (S.D.) Mean (S.D.) Mean (S.D.) Mean (S.D.) Mean (S.D.) Mean (S.D.) Practice of triple-play (month) 2.51 (6.10) 2.59 (6.18) 2.59 (6.18) 1.81 (4.93) 2.32 (5.68) 2.25 (5.61) 2.55 (6.14) Early termination fee 84.33 (70.43) 85.38 (70.71) 85.38 (70.71) 62.34 (59.55) 84.37 (69.93) 83.59 (69.76) 84.09 (70.94) Contract term (month) 9.49 (7.96) 9.6 (7.99) 9.6 (7.99) 7.70 (7.48) 9.68 (8.08) 9.56 (8.05) 9.45 (8.02) Competition: Multichannel video programming distributor market 1048.62 (123.27) 1055.30 (122.39) 1050.79 (122.91) Competition: Long distance telephony market 1602.25 (560.92) 1577.30 (544.86) 1623.48 (582.71) Competition: Internet service provider market 1280.27 (477.26) 1284.21 (479.34) 1277.30 (474.19) Platform competition 4232.54 (508.12) 4235.99 (514.13) 4235.99 (514.13) 4239.19 (517.15) 4252.25 (506.79) 4241.24 (502.50) 4230.70 (511.84) SMATV market share 1.37 (.34) 1.35 (.34) 1.37 (.34) High-speed fixed broadband line per 100 inhabitants 14.37 (11.62) 14.69 (11.65) 14.69 (11.65) 14.60 (12.04) 14.87 (11.63) 14.49 (11.64) 14.53 (11.64) Population density 109.95 (6 9.37) 108.66 (69.06) 108.66 (69.06) 118.66 (71.44) 109.71 (71.34) 110.69 (70.42) 110.32 (69.83) Income per capita 31270.90 (3414.99) 31305.05 (3438.14) 31305.05 (3438.14) 31376.04 (3586.34) 31370.08 (3422.80) 33928.17 (3392.21) 31322.65 (3414.59) Total operating revenue 13826.20 (23281.03) 14243.43 (23516.29) 14243.43 (23516.29) 6551.32 (14595.34) 12445.96 (21073.25) 13086.64 (21764.78) 14029.25 (23399.58) Marketing expenses 3846.99 (7062.09) 3963.44 (7139.34) 3963.44 (7139.34) 1295.42 (3076.33) 3393.26 (6342.28) 3635.36 (6684.09) 3903.68 (7100.83) Capital expenditures 2638.02 (4831.91) 2711.48 (4887.99) 2711.48 (4887.99) 904.23 (2106.73) 2274.76 (4212.65) 2487.70 (4574.77) 2617.71 (4860.99) Regression Analysis of the Impact of Triple-Pl ay on Market Perfo rmance This study employed a two-step model with firmand time-fixed effects in order to examine the impact of the triple-play strategy on the triple-play providers market performance. Market performance here consists of three cat egories: subscriber performance, financial performance, and growth rates. Each category ha s several dependent variables. Main explanatory

PAGE 114

114 variables are the practice of the triple-play stra tegy, early termination f ee, contract term, and competition variables. Other variables are used for the purpose of control. In the first step, this study conducted regression with the explanatory variables that changed by firm and time in addition to timeand firm-dummies representing timeand firm-fix ed effects, respectively. Next, this study estimated time-fixed effects using ti me-dummies in the previous step, and then regressed firm-invariant variables (i.e., competiti on-related variables in this analysis) on this time-fixed effects estimate. F tests support that firmand time-fixed effects are significant across all dependent variables at the 1 percent level. De tailed statistics are provi ded with the report of the results of each analysis. Initially, this study transformed several variables that were positively skewed using logarithmic function using the benchmark of a valu e of 2 z-score of skew ness. In addition, this study conducted correlation analysis in order to prevent potential multicollinearity problems. To assess the strength of correlati ons, the .80 Pearson correlation criterion was employed. Eleven out of twelve explanatory variab les were included except SMATV ma rket share. The variable of SMATV market share showed perfect collineari ty with other competition variables and timedummies, which needed to be dropped in the analyses. Two-step model with firmand time-fix ed effects in the cable industry Subscriber performance in the cable industry: Subscriber perform ance consists of three dependent variables: video, voice, and data subs criber performance. Th e explanatory variables were regressed on each dependent variable in two steps. The regression model in each step (i.e., Model (1) and Model (2) in Table 5-14) was statistically significant at the 1 percent level, and thus F tests support that firmand time-fixed ef fects are significant at the 1 percent level. Table 5-14 provides the results of the regres sion of the triple-pla y strategy impact on subscriber performance in the cable industry. RQ7 addresses the imp act of triple-play on

PAGE 115

115 subscriber performance. For video subscriber perf ormance, the practice of the triple-play strategy was statistically significant at the 5 percent level. As cable fi rms use the triple-play strategy, their video subscriber performance is more likely to improve. In addition, RQ 8 addresses the impact of switching costs on subscriber performance for triple -play providers. The result shows that one of the switching cost variables, co ntract term, was statistically significant at the 5 percent level. The longer cable firms lock in their customers with a contract, the more they are likely to increase the number of video subscribers. RQ9 addresses the impact of competition on subscriber performa nce for triple-play providers. Platform competition wa s statistically significant at the 1 percent level. Cable firms are more likely to increase the number of video subscribers as platform competition decreases. Video market competition, however, had no stat istically significant impact on subscriber performance. Regarding voice subscriber performance, th e practice of the triple-play strategy was statistically significant at the 5 percent level (RQ7). As cable fi rms use the triple-play strategy, the number of voice subscriber is more likely to increase. However, switching cost variables (i.e., early termination fee and contract term) we re not statistically significant (RQ8). Two competition variables were statistically significant at the 1 percent level (RQ9). Cable firms are more likely to increase the number of voice subs cribers as the long dist ance (voice) market becomes less competitive. Cable firms are more likely to increase the number of voice subscribers as the level of platform competition decreases.

PAGE 116

116 Table 5-14 Results of regression of the impact of triple-play on subscriber performance in the cable industry Variable ln Video subscriber (n=91) Voice subscriber (n=91) Data subscriber (n=91) Model (1) Model (2) Model (1) Model (2) Model (1) Model (2) B(1) t (2) B t B t B t B t B t Practice of triple-play .019 2.74* 15.99 2.65* 4.36 .50 Early termination fee -.1.37e-4 -.21 .95 1.04 1.997 3.57* Contract term .026 2.25* -13.97 -1.87 -10.79 -1.15 Video market competition 1.32e-4 1.65 Voice market competition .22 41.39* Data market competition .44 13.76* Platform competition 4.09e-4 21.28* .18 21.02* -.65 -25.44* SMATV penetration dropped Fixed broadband -.001 -.91 .18 .15 -.68 -.51 Population density -.012 -.76 11.80 .60 -11.01 -.41 Income per capita 3.98e-5 .69 -.02 -.21 -.14 -2.83* Total operating revenue 5.61e-6 17 .18 13.22* .59 53.24* Marketing expenses 7.53e-5 4.34* .23 16.23* ln -67.74 -1.07 Capital expenditures 5.86e-5 .56 -.13 -1.62 ln -23.29 -.63 Constant 7.21 2.40* -1.72 -13.41* -1383.0 -.58 -1340.9 -34.70* 5288.7 1.21 R Square .599 .855 .885 .927 .972 .700 (1) Coefficient B; (2) t statistics Statistically significant at the 5% level ln Log transformed variable

PAGE 117

117 For data subscriber performance, the practice of the triple-play strategy was not statistically significant at the 5 percent level (RQ7 ). However, one of the switching cost variables, early termination fee, was statistically significan t at the 5 percent level (RQ8). As cable firms charge higher early termination fees, the number of data subscriber is more likely to increase. Two competition variables were statistically signifi cant at the 1 percent level (RQ9). Cable firms are more likely to increase the number of data subs cribers as the ISPs (data) market becomes less competitive. However, a high level of platform competition is associated with the enhancement of data subscriber performance. Table 5-15 summarize the signifi cant factors of subscriber performance of triple-play providers in the cable industry. As cable firm s use the triple-play strategy, video and voice subscriber performance is more likely to improve Switching costs also cont ribute to better video and data subscriber performance. In addition, cab le firms are more likely to increase the number of voice and data subscribers as voice market competition and data market competition decrease. Video market competition, however, has no signifi cant impact on video subscriber performance. They are more likely to improve video and voice subscriber perfor mance as the level of platform competition decreases. On the other hand, they are more likely to improve data subscriber performance as the level of platform competition increases. Table 5-15 Significant factors of subscrib er performance in the cable industry Subscriber performance Video subscriber Voice subscriber Data subscriber Variable Triple-play + + Switching costs + + Market competition + + Platform competition + + Significant at the 5 percent level + Positive: Positive sign of coefficient, -: Negative sign of coefficient

PAGE 118

118 Financial performance in the cable industry: Financial performance consists of four dependent variables: video, voice, and data revenue performance, and EBITDA. For the analysis of EBITDA, population density was dropped be cause of multicollinea rity problems. The explanatory variables were regressed on each de pendent variable in two steps. Regression models in each step (i.e., Model (1) and Model (2) in Table 5-16) we re statistically significant at the 1 percent level, and thus F te sts support that firmand time-fixe d effects are significant at the 1 percent level throughout the models. Revenue performance: Table 5-16 provides the results of the regression of the impact of triple-play strategy on financial performance (including revenue performance and profitability) in the cable industry. RQ7 addresses the impact of triple-play on financial performance. For video revenue, the practice of the triple-play strategy was not statistica lly significant at the 5 percent level. In addition, RQ8 addresses the impact of switching costs on financial performance for triple-play providers. The resu lt shows that contract term wa s statistically significant. The longer cable firms lock in their customers with a contract, the more they are likely to improve their video revenue. RQ9 addresses the impact of competition on financial performance for triple-play providers. Two competition variables were statistic ally significant at the 1 percent level. As the video market becomes more competitive, cable firms tend to generate less video revenue. In addition, more platform competition tends to l ead to less revenue in the video market.

PAGE 119

119 Table 5-16 Results of regression of th e impact of triple-play on financial performance in the cable industry Variable ln Video revenue (n=91) Voice revenue (n= 91) Data revenue (n=91) EBITDA (n=91) Model (1) Model (2) Model (1) Model (2) Model (1) Model (2) Model (1) Model (2) B ( 1 ) t ( 2 ) B t B t B t B t B t B t B t Practice of triple-play .004 .63 6.97 2.51* .088 .02 -15.57 -.42 Early termination fee -6.40e-3 -1.19 .54 2.12 .907 3.16* 4.16 .70 Contract term .063 3.28* -9.68 -2.60* -5.53 -1.0 35.03 .44 Video market competition 6.19e-4 6.56* 5.44 17.91* Voice market competition .118 79.27* .859 17.66* Data market competition .314 16.61* -.063 -1.20 Platform competition 2.39e-4 16.67* 125 59.43* -.466 -23.33* .708 14.90* SMATV penetration dropped dropped Fixed broadband .001 .73 -.62 -1.31 -.620 .58 -6.31 -1.12 Population density -.002 -.14 3.66 .47 -10.63 -.70 Income per capita 2.99e-5 .65 -.002 -.06 -.084 -3.0* .082 .23 Total operating revenue ln.082 1.08 .08 14.64* .292 59.77* ln 688.54 .90 Marketing expenses ln.027 1.42 ln 67.93 2.22* ln -106.75 -2.72* .971 6.35* Capital expenditures .0001 2.40* ln -5.28 -.91 ln -13.44 -.82 l n 111.18.50 Constant 3.17 1.08 -1.56 -11.0* -924.86 -.95 -850.68 -109.45*4264.79 1.62 .182 18.27* -7748.6 -.73 -10354 -18.69* R square .678 .613 .913 .982 .981 .735 .334 .824 (1) Coefficient B; (2) t statistics Statistically significant at the 5% level ln Log transformed variable

PAGE 120

120 In terms of voice revenue, the practice of the triple-play strategy was statistically significant at the 5 percent level (RQ7). As cable firms use the triple-p lay strategy, voice revenue is more likely to increase. Contract term was statistically significant (R Q8). Unlike the case of video revenue, the longer cable firms lock in th eir customers with a cont ract, the less they are likely to raise voice revenue. Two competition variables were statistically significant at the 1 percent level (RQ9). As the long distance (voice) market becomes more competitive, cable firms are less likely to generate voice revenue. Also, mo re platform competition tends to lead to less revenue in the voice market. For data revenue, the triple-play practice was not statistically significant at the 5 percent level (RQ7). However, one of the switching cost va riables, early termination fee, was statistically significant at the 5 percent level (RQ8). As cable firms charge higher early termination fees, data revenue is more likely to increase. Two competitio n variables were statistically significant at the 1 percent level (RQ9). As the ISPs (data) market becomes more competitive, cable firms are less likely to increase data revenue. However, more platform competition tends to lead to more revenue in the data market. Profitability: In the case of EBITDA, as a measur e of firms profitability, none of strategic variables (i.e., the triple-play use and switching costs) was statistically significant at the 5 percent level (RQ7 & RQ8). Th e practice of the trip le-play strategy also has a negative sign, which suggests the triple-play strategy might negatively imp act EBITDA despite statistical insignificance. Three competition variables, however, were statistically significant (RQ9). As the video market and the long distance (voice) market become more competitive, firms profitability is less likely to increase. Also, more platform competition tends to improve firms profitability. Competition in the ISPs (data) market was not statistically significant.

PAGE 121

121 Table 5-17 summarizes significan t factors of financial perfor mance in the cable industry. From a perspective of revenue pe rformance, the use of the triple-play strategy tends to increase only voice revenue in the cable indu stry. Switching costs play a role in increasing video and data revenue but deliver an opposite e ffect on voice revenue. Cable firm s are more likely to increase video, voice and data revenue as the level of co mpetition in all three markets decreases. Also, lower level of platform competition is associated with better video and voice revenue while a higher level of platform competition is associated with better data revenue. Note that these results were consistent with the case of subs criber performance in previous analyses. Regarding a firms profitability, measured by EBITDA, none of stra tegic variables (i.e., the use of the triple-play stra tegy and switching costs) had a st atistically significant impact. Nevertheless, firms profitability is more likely to increase as video and voice markets are less competitive. Table 5-17 Significant factors of financial performance in the cable industry Financial performance Revenue performance Profitability Variable Video revenue Voice revenue Data revenue EBITDA Triple-play + Switching costs + + Market competition + + + + (video & voice only) Platform competition + + + Significant at the 5 percent level + Positive: Positive sign of coefficient, -: Negative sign of coefficient Growth rates in the cable industry: This study analyzed the growth rates of subscriber and financial performances. The explanatory variables were regressed on each dependent variable in two steps. The regr ession model in each step (i.e., M odel (1) and Model (2) in Table 5-18) was statistically significant at the 1 percen t level, and thus F test s support that firmand time-fixed effects are significant at the 1 percent level.

PAGE 122

122 Table 5-18 provides the results of the regression of the impact of the triple-play strategy on the growth rates for the cable firms. RQ7 addre sses the triple-play strategy impact on growth rates. Out of all the dependent va riables, the growth rates of voi ce subscribers and the growth of data revenue were associated with the practice of the triple-play strategy at the 5 percent significant level. Specifically, there was a nega tive association between the triple-play strategy and the growth of voice subscribers, while ther e was a positive association between the tripleplay strategy and the growth of data revenue. RQ8 addresses the impact of switching costs on growth rates for triple-play providers. Switching costs variables, either early termina tion fee or contract te rm, were statistically significant for the growth rate of EBITDA and th at of voice revenue. Specifically, there was a positive association between the triple-play strate gy and the growth of EBITDA, while there was a negative association between the triple-play strategy and the growth of voice revenue. RQ9 addresses the impact of competition on gr owth rates for triple-play providers. All competition variables were statistically significant across all dependent variables. In detail, for the growth rates of video subscribers and vide o revenue, increasing vide o market competition was negatively associated with the growth, while increasing platform competition was positively associated with the growth. For the growth rates of voice subscribers and voice revenue, increasing long-distance (voice) market competition and increasing platform competition were positively associated with the growth. For the growth rates of data subscribers and data revenue, increasing ISPs market competition was positively associated with the growth, while increasing platform competition was negatively associated with the growth. For the growth rate of EBITDA, increasing video, voice, and platform competitions were negatively associated with the growth, while increasing ISPs market competition wa s positively associated with the growth.

PAGE 123

123 Table 5-18 Results of regression of the growth rates of market performance in the cable industry Variable Growth: Video subscriber (n=89) Growth: Voice subscriber (n=88) Growth: Data subscriber (n=87) Growth: EBITDA (n=90) Model (1) Model (2) Model (1) Model (2) Model (1) Model (2) Model (1) Model (2) B ( 1 ) t ( 2 ) B t B t B t B t B t B t B t Practice of triple-play .573 .37 -7.72 -2.24* .81 .48 -38.21 -.52 Early termination fee -.225 -1.38 .456 .87 .06 .50 18.21 3.12* Contract term -3.98 -.98 17.73 .69 -6.17 -1.39 -311.93 -1.90 Video market competition .343 6.95* 6.90 5.93* Voice market competition -.27 -24.69* 4.33 23.25* Data market competition -.15 -20.59* -1.90 -9.66* Platform competition -.061 -6.13* -.41 -15.07* .11 13.59* 5.49 30.22* SMATV penetration dropped dropped Fixed broadband ln 9.38 1.58 ln -2.92 -.10 .74 1.63 4.87e-5 1.35 Population density -9.52 -1.76 -11.48 .54 .38 .05 -325.70 -1.46 Income per capita -.009 -.63 -.04 -.88 .002 .16 1.21 1.91 Total operating revenue .012 1.65 .004 .18 ln 50.41 1.0 .41 1.39 Marketing expenses ln -26.69 -.90 ln -5.07 -.07 -.001 -.14 -.316 -1.24 Capital expenditures -.047 -1.88 -.056 -.98 -1.09 -1.23 Constant 1780.7 1.67 -14.99 -.18 2838.4 .78 2438.5 18.09*-237.63 -.18 -448.1 -10.62* 16660.2 .52 -38525 -18.12* R square .293 .628 .283 .875 .397 .568 .278 .971 Variable Growth: Video revenue (n= 78) Growth :Voice revenue (n=91) Growth :Data revenue (n=91) Model (1) Model (2) Model (1) Model (2) Model (1) Model (2) B t B t B t B t B t B t Practice of triple-play .20 .13 -.307 -.04 2.12 2.26* Early termination fee .18 .77 3.01 1.39 .129 .87 Contract term -20.92 -1.79 -34.69 -2.24* -6.79 -2.10 Video market competition .138 3.65* Voice market competition -.493 -34.46* Data market competition -.065 -13.40* Platform competition -.056 -8.59* -.644 -40.27* .083 18.39* SMATV penetration dropped Fixed broadband 2.16e-6 1.25 -9.79e-6 1.61 -4.26e-7 -.54 Population density -2.87 -.50 20.77 .88 3.94 .67 Income per capita -.029 -1.22 -.18 -1.58 -1.69e-4 -.01 Total operating revenue .016 1.16 -.009 -.38 -.010 -1.72 Marketing expenses -.03 -1.56 -.029 .032 -.008 -1.04 Capital expenditures -.001 -.05 -.02 -.15 .071 3.50* Constant 1560.8 1.27 128.63 2.25* 2239.3 .55 4021.5 51.53* -470.38 -.43 -335.38 -14.89* R square .338 .522 .187 .958 .455 .633 (1) Coefficient B; (2) t statistics Statistically significant at the 5% level ln Log transformed variable

PAGE 124

124 Table 5-19 summarizes signifi cant factors of growth in the cable industry. From a perspective of market share (the number of subscribers and revenue ), the use of the triple-play strategy is associated with a decline in the grow th rates of voice subscriber market share and revenue market share. On the other hand, the stra tegy is positively associated with the growth rates in the data subscriber market share and da ta revenue market share for the cable firms. In addition, switching costs are associated with a de cline in the growth rates of voice revenue market share. Table 5-19 Significant factors of growth in the cable industry Growth: Subscriber market shar e Growth: Revenue market share Growth: Profitability Variable Growth: Video Sub Growth: Voice Sub Growth: Data Sub Growth: Video Rev Growth: Voice Rev Growth: Data Rev Growth: EBITDA Triple-play + Switching Costs Market Competition + + + (video & voice) (data) Platform Competition + + + Significant at the 5 percent level + Positive: Positive sign of coefficient, -: Negative sign of coefficient A higher level of video market co mpetition is associated with declines in the growth rates of video subscriber market share and video re venue market share. On the other hand, a higher level of voice and data market competition is po sitively associated with the growth rates of voice and data market share (both subscriber and reve nue). A higher level of pl atform competition is positively associated with the growth rates of video and voice market share (both subscriber and revenue). However a higher level of platform co mpetition is associated with a decline in the growth rates of data market shar e (both subscriber and revenue). From a perspective of firms profitability, th e triple-play strategy ha s no significant impact on the growth rates of firms profitability. Switching costs, however, are positively associated with the growth rates of firms profitability. A higher level of video and voice market

PAGE 125

125 competition is negatively associated with firms profitability. On the other hand, a higher level of data market competition is positively associated with firms profitability. Two-step model with firmand time-fixed effects in the telephone industry Subscriber performance in the telephone industry: Subscriber performance consists of three dependent variables: video, voice, and data subscriber perform ance. The explanatory variables were regressed on each dependent variable in two steps. The regression model in each step (i.e., Model (1) and Model (2 ) in Table 5-20) was statistically significant at the 1 percent level, and thus F tests support th at firmand time-fixed effects are significant at the 1 percent level. Table 5-20 provides the results of the regression of the impact of the triple-play strategy on subscriber performance in the telephone industry. RQ7 addresses the impact of triple-play on subscriber performance. For video subscriber perf ormance, the practice of the triple-play strategy was statistically significa nt at the 5 percent level. As telepho ne firms use the triple-play strategy, video subscriber performance is more likely to improve. In addition, RQ 8 addresses the impact of switching costs on subscriber performance for triple -play providers. One of the switching cost variables, early termination fee, was statistically significant at the 5 percent level. As telephone firms charge higher early termination fees, the nu mber of video subscribers is more likely to increase. RQ9 addresses the impact of competition on subscriber performa nce for triple-play providers. Two market competition variables were statistically significant at the 1 percent level. As the video market becomes more competitive, telephone companies are less likely to increase the number of video subscribers. However, more platform competition tends to lead to more increases in the number of video subscribers.

PAGE 126

126 Table 5-20 Results of regression of the impact of triple-play on subscriber performance in the telephone industry Variable ln Video subscriber (n=79) Voice subscr iber (n=79) Data subscriber (n=79) Model (1) Model (2) Model (1) Model (2) Model (1) Model (2) B ( 1 ) t ( 2 ) B t B t B t B t B t Practice of triple-play .219 10.15* .031 2.43* .044 2.47* Early termination fee .016 2.84* -.007 -2.44* -.003 -.88 Contract term .046 0.74 .075 2.32* .083 2.02 Video market competition 0 0 1 4 1 7 Voice market competition -.002 -30.68* Data market competition -.002 -23.99* .0007 13.88* Platform competition -.0002 -2.84* -.0004 -14.18* SMATV penetration dropped Fixed broadband .016 0.90 ln -1.90 -1.79 ln -.08 -0.15 Population density -.007 -6.34* -.002 -2.48* -.002 -2.65* Income per capita 3.0e-5 3.10* -2.21e-5 -1.22 -8.45e-06 -0.61 Total operating revenue ln -.503 -1.52 ln.357 2.18* ln 1.19 6.53* Marketing expenses ln.840 1.31 ln.039 .13 -.00003 -0.47 Capital expenditures -.0002 -6.51* -9.19e-5 -1.40 -9.0e-5 -3.03* Constant -2.96 -0.75 -1.72 -13.41* 7.29 3.26* 15.15 35.81* -4.31 -3.82 R Square .874 .380 .510 .937 .873 .613 (1) Coefficient B; (2) t statistics Statistically significant at the 5% level ln Log transformed variable

PAGE 127

127 Regarding voice subscriber performance, th e practice of the triple-play strategy was statistically significant at the 5 percent level (RQ7). As tele phone firms use the triple-play strategy, the number of voice subscr iber is more likely to increase. Both switching cost variables (i.e., early termination fee and c ontract term) were statistically significant (RQ8). Two market competition variables were statistically significan t at the 1 percent level (RQ9). As the long distance (voice) market becomes more competitive telephone firms are less likely to increase the number of voice subscribers. Also, telephone firm s are more likely to increase the number of voice subscribers as platform competition decreases (RQ9). For data subscriber performance, the practice of the triple-play strategy was statistically significant at the 5 percent level (RQ7). None of the switching cost related variables was statistically significant at the 5 percent level (R Q8). Two competition variables were statistically significant at the 1 percent level (RQ9). As the ISPs (data) market becomes more competitive, firms are less likely to increase the number of data subscribers. However, a high level of platform competition is associated with better data subscriber performance. Table 5-21 summarizes significant factors of subscriber performance in the telephone industry. As telephone firms use the triple-play st rategy, their subscriber performances of all three services tend to improve. Switching costs related factors al so seem to improve video and data subscriber performance. As video market and data market become more competitive, telephone firms are less likely to increase the number of video and data subscribers. On the other hand, as voice market becomes more competitive, te lephone firms are more likely to increase the number of voice subscribers. In addition, high level of platform competition is negatively associated with better subscriber performance of all three services.

PAGE 128

128 Table 5-21 Significant factors of subscrib er performance in the telephone industry Subscriber performance Video subscriber Voice subscriber Data subscriber Variable Triple-play + + + Switching costs + + (fee) (contract term) + Market competition + + Platform competition Significant at the 5 percent level + Positive: Positive sign of coefficient, -: Negative sign of coefficient Financial performance in the telephone industry: Financial performance consists of four dependent variables: video, voice, a nd data revenue performance and EBITDA. The explanatory variables were regressed on each de pendent variable in two steps. Regression models in each step (i.e., Model (1)s and Model (2)s in Table 522) were statistically significant at the 1 percent level, and thus F tests support that firmand time-fixed effects are significant at the 1 percent level throughout models. Revenue performance: Table 5-22 provides the results of the regression of the impact of the triple-play strategy on financial performan ce in the telephone industry. RQ7 addresses the impact of triple-play on financial performance. For video revenue, the practice of the triple-play strategy was statistically significant at the 5 percent level. In addition, RQ8 addresses the impact of switching costs on financial performance for triple-play provi ders. The switching costs were not statistically significant at the 5 percent level. RQ9 addresses the impact of competition on financial performance for triple-play providers. Two competition variables were statistic ally significant at the 1 percent level. As the video market becomes more competitive, firm s are less likely to increase video revenue. However, as platform competition increases, firm s are more likely to increase video revenue. Regarding voice revenue, the practice of th e triple-play strategy was not statistically significant at the 5 percent level (RQ7). One of switching costs factors, contract term, was

PAGE 129

129 statistically significant (RQ8). Two competition va riables were statistically significant at the 1 percent level (RQ9). As the long distance (voice) market becomes more competitive, firms are less likely to increase voice revenue. Also, mo re platform competition tends to generate less revenue in the voice market. For data revenue, the practice of the triple-pla y strategy was statistically significant at the 5 percent level (RQ7). Switching costs related factor s were not statistically significant (RQ9). Two competition variables were statistically significant at the 1 percent level (RQ9). As the ISPs (data) market becomes more competitive, firms are less likely to increase data revenue. However, a higher level of platform competition is associ ated with increases in data revenue. Profitability: In the case of EBITDA, as a measure of firms profitabi lity, none of the strategic variables was statistically significan t at the 5 percent level (RQ7 & RQ8). Four competition variables were statistically significant (RQ9). As the video market and the long distance (voice) market become more competitive, firms profitability is less likely to increase. However, as the ISPs (data) market becomes more competitive, firms profitability is more likely to increase. In addition, as platform competition in creases, firms profitability is less likely to increase.

PAGE 130

130 Table 5-22 Results of regression of the impact of triple-play on financial pe rformance in the telephone industry Variable ln Video revenue (n=91) Voice revenue (n= 91) Data revenue (n=91) EBITDA (n=91) Model (1) Model (2) Model (1) Model (2) Model (1) Model (2) Model (1) Model (2) B ( 1 ) t ( 2 ) B t B t B t B t B t B t B t Practice of triple-play .17 2.25* .021 1.60 40.77 2.39* 95.24 1.41 Early termination fee .01 .28 -.004 -1.64 .75 .14 4.32 .35 Contract term .14 .95 .79 2.57* -1.59 -.03 -151.22 -1.26 Video market competition 5.85e-4 3.22* 9.66e-5 12.33* 14.24 84.67* Voice market competition 4.24 282.23* Data market competition .802 10.95* -.684 -20.76* Platform competition -.0007 -18.15* 2.53e -4 10.60* .-.685 -10.51* 3.49 116.1* SMATV penetration dropped dropped Fixed broadband -.01 -.20 .009 1.36 ln -1216.6 -1.62 3.82 .10 Population density -.004 -1.26 -.0008 -1.90 -1.28 -1.06 l n 969.61 .69 Income per capita -5.75e-5 -1.35 -1.13e-5 -1.53 .124 1.47 .382 .05 Total operating revenue .458 .61 ln.337 2.71* -74.22 -.32 l n 452.43 .51 Marketing expenses 5.24e-4 1.87 ln.151 .75 .84 2.02 ln.811 .86 Capital expenditures 1.62e-5 .17 2.23e-5 1.29 .18 3.41* l n 414.62 .37 Constant -4.96 -.99 2.385 7.79* 3.39 2.77 -1.29 -11.09* -2511.6 -1.15 3335.8 10.23*-18367 -1.55 -39235 -136.9* R Square .742 .880 .774 .766 .896 .389 .433 .996 (1) Coefficient B; (2) t statistics Statistically significant at the 5% level ln Log transformed variable

PAGE 131

131 Table 5-23 summarizes significan t factors of financial performance for telephone firms. From a perspective of revenue pe rformance, the use of the triple-play strategy tends to lead to higher video and data revenue in the telephone industry. Switchi ng costs related factors also contribute to increase voice revenue. As the leve l of competition in all three markets decreases, telephone firms are more likely to increase vi deo, voice and data revenue. A higher level of platform competition is associated with an increas e in video and data revenue while a lower level of platform competition is associated with an increase in voice revenue. From a perspective of a firms profitability, none of strategic variables (i.e., the use of the triple-play strategy and switching co sts) had a statistically signif icant impact. Finally, as video market and voice market become more competitive, firms profitability is less likely to increase. As data market becomes more competitive, how ever, firms profitability is more likely to increase. Table 5-23 Significant factors of financia l performance in the telephone industry Financial performance Revenue performance Profitability Variable Video revenue Voice revenue Data revenue EBITDA Triple-play + + Switching costs + Market competition + + + + (video & voice) (data) Platform competition + + Significant at the 5 percent level + Positive: Positive sign of coefficient, -: Negative sign of coefficient Growth rates in the telephone industry: This study analyzed the growth rates of subscriber and financial performances. The e xplanatory variables were regressed on each dependent variable in tw o steps. The regression m odel in each step (i.e., Model (1) and Model (2) in Table 5-24) was statistically significant at the 1 percent level, and thus F tests support that firmand time-fixed effects are sign ificant at the 1 percent level.

PAGE 132

132 Table 5-24 provides the results of the regression of the triple-play strategy impact on the growth rates in the telephone industry. RQ7 addre sses the impact of triple -play on growth rates. Out of all the dependent variables, the growth rates of video subscribers and the growth of video revenue were positively associated with the prac tice of the triple-play strategy at the 5 percent significant level. RQ8 addresses the impact of switching costs on growth rates for tr iple-play providers. Switching costs variables, either early termina tion fee or contract te rm, were statistically significant for the growth rate of data subscr ibers and data revenue. Specifically, there was a positive association between the triple-play strategy and the growth rates of these two variables. RQ9 addresses the impact of competition on gr owth rates for triple-play providers. All competition variables were statistically significant across all dependent variables. For the growth rates of video subscribers and video revenue, increasing video market competition was negatively associated with the growth. Increasi ng platform competition was positively associated with the growth rate of video revenue, while increasing platform competition was associated with a decline in the growth rate of video subs cribers. For the growth rates of voice subscribers and voice revenue, increasing long-distance (voice ) market competition and increasing platform competition were positively associated with the gr owth. For the growth rates of data subscribers and data revenue, increasing ISPs market competit ion was positively associated with the growth, while increasing platform competition was associated with a decline in the growth. For the growth rate of EBITDA, increasing video, voice and platform competitions were positively associated with the growth, while increasing ISPs market competition wa s negatively associated with the growth.

PAGE 133

133 Table 5-24 Results of regression of the growth rates of market performance in the tele phone industry Variable Growth: Video subscriber (n=89) Growth: Voice subscriber (n=88) Growth: Data subscriber (n=87) Growth: EBITDA (n=90) Model (1) Model (2) Model (1) Model (2) Model (1) Model (2) Model (1) Model (2) B ( 1 ) t ( 2 ) B t B t B t B t B t B t B t Practice of triple-play 5.26 3.02* -2.62 -.97 3.44 1.49 1.86 .70 Early termination fee .530 .54 .113 .26 -.132 -.95 -.748 -1.07 Contract term -10.77 -1.15 10.57 1.68 9.46 3.06* -4.33 -.50 Video market competition .352 9.75* -.464 -6.17* Voice market competition .208 33.63* -.062 -11.83* Data market competition -.119 -16.18* .134 10.26* Platform competition -.034 -3.96* 211 13.21* .139 7.69* -.079 -5.80* SMATV penetration dropped dropped Fixed broadband ln 43.68 -1.01 ln 164.7 2.06 3.152 1.19 -1.14 -.63 Population density -.208 -.96 .001 .01 .073 .37 .28 1.35 Income per capita -.007 -1.74 .0008 .29 .002 .47 -.002 -.67 Total operating revenue ln 13.10 .41 ln 100.8 1.96 ln -7.32 -.19 116.87 2.44* Marketing expenses .011 .71 16.83 .18 133.92 1.06 -.010 -.33 Capital expenditures 1780.7 1.67 16.20 .33 -45.81 -1.26 .008 .54 Constant 201.57 .83 -14.99 -.18 -1291 -2.13 -1429.8 -20.04*-503.92 -1.01 -555.8 -7.20 -756.49 -2.68* -38525 -18.12* R square .486 .760 .168 .825 .189 .536 .133 .680 Variable Growth: Video revenue (n=78) Growth :Voice revenue (n=91) Growth :Data revenue (n=91) Model (1) Model (2) Model (1) Model (2) Model (1) Model (2) B t B t B t B t B t B t Practice of triple-play 1.152 2.48* -.009 -.01 7.62 1.71 Early termination fee -.090 -.66 -.289 -.83 -.246 -.30 Contract term .598 .48 6.79 1.44 16.16 3.24* Video market competition .035 11.98* Voice market competition .363 19.67* Data market competition -.321 -11.66* Platform competition .003 5.61* .261 14.21* .359 7.98* SMATV penetration dropped Fixed broadband -.003 -.01 233.79 1.88 7.86 1.86 Population density -.008 -.33 .214 1.31 .042 .14 Income per capita -.001 -2.09 .002 .76 .007 1.04 Total operating revenue -2.03 -.71 55.97 2.12 110.14 3.25* Marketing expenses -9.09 -.73 67.73 .79 363.91 2.13 Capital expenditures 3.67 .82 38.84 1.08 -202.15 -2.33* Constant 73.65 1.06 128.63 2.25* -1228.0 -5.15* -2122.9 -23.45* -1887.44 -2.51* -1508.0 -7.82* R square .448 .725 .411 .889 .434 .494 (1) Coefficient B; (2) t statistics Statistically significant at the 5% level ln Log transformed variable

PAGE 134

134 Table 5-25 summarizes significant factors of growth rates for telephone firms. From a perspective growth in market share related measures (i.e., the number of subscribers and revenue), the use of triple-play st rategy tends to lead to higher gr owth rates in the video market for the telephone firms. Switching co sts also tend to increase the gr owth rates of the data market share (both subscribers and revenue). A higher le vel of video and voice market competition is associated with a decline in the growth ra tes of the video and voice market share (both subscribers and revenue). On the other hand, a higher level of data market competition is positively associated with the growth rates of the data market share (both subscribers and revenue). A higher level of platform competition is associated with a decline in the growth rates of the voice and data market share (both subscribers and revenue). From a perspective of growth of firms profitability, the triple-play strategy has no statistically significant impact. Nor do switching co sts related factors. A higher level of video and voice market competition is positively associat ed with firms profitability. However a higher level of data market competition is negativ ely associated with firms profitability. Table 5-25 Significant factors of gr owth in the telephone industry Growth: Subscriber market shar e Growth: Revenue market share Growth: Profitability Variable Growth: Video Sub Growth: Voice Sub Growth: Data Sub Growth: Video Rev Growth: Voice Rev Growth: Data Rev Growth: EBITDA Triple-play + + + Switching Costs + Market Competition + + + + (video & voice) + (data) Platform Competition + + + + + Significant at the 5 percent level + Positive: Positive sign of coefficient, -: Negative sign of coefficient

PAGE 135

135 Analysis of Variance (ANOVA) Results of the Impact of Triple-Play on Market Performance This study employed Analysis of Variance ( ANOVA) to examine differences in market performance among firms. RQ7 addresses the im pact of the triple-play strategy on market performance by firms in addition to previous regr ession analyses. Based on year(s) of the tripleplay practice, firms are divided into six groups for the cable indus try (non-TPS, 1 year of TPS, 2 years of TPS, 3 years of TPS, 4 years of TPS, and 5 years of TPS group), five for the telephone industry (non-TPS, 1 year of TPS, 2 years of T PS, 3 years of TPS, and 4 years of TPS group). ANOVA analyses were performed with each dependent variable in each industry. Next, for the dependent variables showing stat istically significant results, post hoc analyses were performed to see which group mean differences led to signifi cant ANOVA results by comp aring every pair of group means. Cable industry: Table 5-26 illustrates the results of ANOVA of market performance by years of the triple-play strategy in the cable industry, but onl y statistically significant results. Years of the triple-play strategy was statistically significant for four dependent variables at the 5 percent level: voice subscriber, voice revenue, the growth rate of da ta subscriber, and the growth rate of data revenue. For these four variables, post hoc analysis was c onducted using the Tukey method. For voice subscriber, there was a statistically si gnificant mean difference between non-TPS-group and 4 years of TPS group, t=1229.74, p<.05. For voice revenue, there was a statistically significant mean difference between non-TPS-group and 4 years of TPS group, t=502.64, p<.05. For the growth rate of data revenue, there was a statistically significant mean difference between

PAGE 136

136 non-TPS-group and 4 years of TPS group, t=85.86, p<.05. However for the growth of data subscriber, none of the pairs of group means showed statis tically significant results. Table 5-26 Results of ANOVA of market perfor mance by years of triple-play in the cable industry Dependent Variable Source SS df MS F Variable Mean (SD) Yrs of TPS Voice subscriber Yrs of TPS 20747199 5 4149439 5.248* 0 130.6 (299.74) Error 76695998 97 790680 1 297.07 (348.68) 2 577.8 (686.17) (Significant group difference: non vs. 4 yrs)(1) 3 946.11 (1299.86) 4 1388.56 (2027.05) 5 1257.82 (1583.55) Voice revenue Yrs of TPS 3297715 5 659543 4.639* 0 62.71 (148.99) Error 12511098 88 142171 1 136.56 (185.17) 2 231.75 (267.02) (Significant group difference: non vs. 4 yrs) 3 403.68 (515.70) 4 570.77 (821.48) 5 532.84 (656.41) Growth of data subscriber Yrs of TPS 198622 5 39724 3.75* 0 109.46 (144.07) Error 985187 93 10593 1 29.42 (14.79) 2 23.75 (11.70) 3 24.66 (16.46) (Significant group difference: Not detected) 4 6.32 (16.75) 5 11.49 (4.96) Growth of data revenue Yrs of TPS 99712 5 19942 3.43* 0 87.94 (106.76) Error 487770 84 5806 1 29.20 (29.07) 2 28.30 (11.58) 3 21.68 (9.53) (Significant group difference: non vs. 4 yrs) 4 8.75 (29.47) 5 21.04 (21.84) Statistically significant at the 5% level (1) Post hoc test result summary In sum, there are group differences in the voice market share and the growth rates of data market share among TPS groups in the cable indust ry. In particular, firms with 4 years of TPS practice are more likely to have higher voice mark et share than non-TPS-fi rms. Note that early movers (i.e., firms with 4 or 5 years of TPS practice) faced different market environment compared to start-ups (i.e., firms with 1 or 2 y ears of TPS practice), whic h was not controlled in this analysis. In addition, firms with 4 years of TPS practice tend to have lower growth rates of data market share than non-TPS-firms. Howeve r there is no significant difference in firms profitability among TPS firm groups.

PAGE 137

137 Telephone industry: Table 5-27 illustrates th e results of ANOVA of market performance by years of the triple-play stra tegy in the telephone industry, but only statistically significant results. Years of the triple-play strategy was statistically significant for seven dependent variables at the 5 percent level: video subscriber, data subscrib er, video revenue, data revenue, EBITDA, the growth rate of video subscriber and the growth rate of video revenue. Table 5-27 Results of ANOVA of market performance by years of triple-play in the telephone industry Dependent Variable Source SS df MS F Variable Mean (SD) Yrs of TPS Video subscriber Yrs of TPS 13356304 4 3339076.10 31.41* 0 9.91 (65.04) Error 9354614.4 88 106302.44 1 123.11(205.03) 2 320.58 (461.66) (Significant group difference: non vs. 3 yrs; non vs. 4 yrs)(1) 3 623.83 (759.04) 4 1983.67 (496.85) Data subscriber Yrs of TPS 288884960 4 72221240.24 11.48* 0 520.70 (1110.43) Error 553624546 88 6291188.03 1 1810.15 (2803.43) (Significant group difference: non vs. 4 yrs; 1 yr vs. 4 yrs; 2 yr vs. 4 yrs; 3 yr vs. 4 yrs) 2 2746.70 (4131.8) 3 3527.95 (5054.47) 4 9273.33 (6746.56) Video revenue Yrs of TPS 47351186 4 11705177.57 13.20* 0 0 (0.00) Error 61873536 69 966774.01 1 522.52 (1203.63) (Significant group difference: non vs. 4 yrs; 1 yr vs. 4 yrs; 2 yr vs. 4 yrs; 3 yr vs. 4 yrs) 2 880.49 (2016.69) 3 1104.56 (2337.96) 4 6304.00 Data revenue Yrs of TPS 338049460 4 145235738.6 3.27* 0 2193.87 (4165.57) Error 2170169540 84 19479301.2 1 3083.18 (4673.58) (Significant group difference: non vs. 4 yrs; 1 yr vs. 4 yrs; 2 yr vs. 4 yrs; 3 yr vs. 4 yrs) 2 3155.97 (6245.93) 3 4549.35 (8776.48) 4 14888.5 (13411.69) EBITDA Yrs of TPS 1e+009 4 342689741 3.96* 0 3600.85 (8189.87) Error 7e+009 88 86577412 1 5054.45 (8568.44) (Significant group difference: non vs. 4 yrs; 1 yr vs. 4 yrs; 2 yr vs. 4 yrs; 3 yr vs. 4 yrs) 2 6570.44 (10128.15) 3 9205.00 (14099.95) 4 24034.6 (17134.12) Growth of video subscriber Yrs of TPS 366667 4 91666 15.41* 0 1.52 (11.99) Error 517674.4 87 5950 1 10.36 (18.04) (Significant group difference: non vs. 2 yrs; non vs. 3 yrs; 1 yr vs. 2yrs; 1 yr vs. 3 yrs) 2 195.52 (229.4) 3 120.32 (107.57) 4 54.73 (42.795) Growth of video revenue Yrs of TPS 9985.14 4 3268 6.82* 0 0 (0.00) Error 25223.99 88 2529 1 0 (0.00) (Significant group difference: non vs. 2 yrs; non vs. 3 yrs; 1 yr vs. 2 yrs; 1 yr vs. 3 yrs) 2 31.03 (27.33) 3 30.56 (64.41) 4 7.38 Statistically significant at the 5% level (1) Post hoc test result summary

PAGE 138

138 For these seven variables, post hoc analysis was conducted using the Tukey method. For video subscriber, there was a statistically significant mean difference between non-TPS-group and 3 years of TPS group, t=533.15, p<.05, and between non-TPS-group and 4 years of TPS group, t=3221.72, p<.05. For data subscriber, there was a statistica lly significant mean difference between non-TPSgroup and 4 years of TPS group, t=23235.72, p<.05; be tween 1 year of TPS group and 4 years of TPS group, t=22658.39, p<.05; between 2 years of TPS group and 4 years of TPS group, t=21311.03, p<.05; between 3 years of TPS group and 4 years of TPS group, t=19892.59, p<.05. For video revenue, there was a statistically significant mean difference between non-TPSgroup and 4 years of TPS group, t=6304.00, p<.05; be tween 1 year of TPS group and 4 years of TPS group, t=5781.48, p<.05; between 2 years of TPS group and 4 years of TPS group, t=5423.51, p < .05; between 3 years of TPS grou p and 4 years of TPS group, t=5199.44, p<.05. For data revenue, there was a statistically significant mean difference between non-TPSgroup and 4 years of TPS group, t=23235.72, p<.05; be tween 1 year of TPS group and 4 years of TPS group, t=22658.91, p<.05; between 2 years of TPS group and 4 years of TPS group, t=21311.03, p<.05; between 3 years of TPS grou p and 4 years of TPS group, t=19892.59, p<.05. For EBITDA, there was a statistically significant mean difference between non-TPSgroup and 4 years of TPS group, t=33978.61, p<.05; be tween 1 year of TPS group and 4 years of TPS group, t=33631.81, p<.05; between 2 years of TPS group and 4 years of TPS group, t=32371.43, p<.05; between 3 years of TPS group and 4 years of TPS group, t=29634.66, p<.05. For the growth rate of video subscriber, there was a statistically significant mean difference between non-TPS-group and 2 years of T PS group, t=101.38, p<.05; between non-TPS-group

PAGE 139

139 and 3 years of TPS group, t=89.51, p<.05; betwee n 1 year of TPS group and 2 years of TPS group, t=90.43, p<.05; between 1 year of TPS group and 3 years of TPS group, t=78.57, p<.05. For the growth rate of video revenue, there was a statistically significant mean difference between non-TPS-group and 2 years of TPS group, t=31.03, p<.05; between non-TPS-group and 3 years of TPS group, t=30.56, p<.05; between 1 year of TPS group and 2 years of TPS group, t=31.03, p<.05; between 1 year of TPS group and 3 years of TPS group, t=30.56, p<.05. In sum, there are group differences in the video and data market share among TPS groups in the telephone industry. Note that early movers (i.e., firms with 3 or 4 years of TPS practice) faced different market environment compared to start-ups (i.e., firms with 1 or 2 years of TPS practice), which was not controlled in this analysis. In particul ar, firms with 4 years of TPS practice are more likely to have higher video a nd data market share than other TPS groups. In addition, there are significant group differences in the growth rates of video market share among TPS groups. Firms with 2 years of TPS practice s how the highest growth rates of video market share than other TPS groups. From a perspective of firms profitability, firms with 4 years of TPS practice show the highest profitability. Analysis of Variance (ANOVA) Results of the Impact of Triple-Play on Market Performance To address RBV questions (RQ10), this st udy employed Analysis of Variance (ANOVA) to examine which service contributes the mo st to total operating revenue by industry. The dependent variables are shares of three different service revenue (i.e., vi deo, voice, and data) as shares of total revenue in each industry21. The explanatory variables are years of the triple-play practice and industry. Based on year(s) of the trip le-play practice, firms are divided into six 21 Individual service revenue as share of total revenue was calculated as follows. Individual service revenue was divided by total operating revenue in each industry.

PAGE 140

140 groups for the cable industry (non-TPS, 1 year of T PS, 2 years of TPS, 3 years of TPS, 4 years of TPS, and 5 years of TPS group), and five for th e telephone industry (non-TPS, 1 year of TPS, 2 years of TPS, 3 years of TPS, and 4 years of TPS group). ANOVA analyses were performed with two explanatory variables to examine differences in individual service revenue as a share of total revenue between cable a nd telephone industry. Table 5-28 illustrates the results of ANOVA of individual service revenue share. For video revenue as a share of total operating revenue, in dustry was statistically significant at the 5 percent level. For voice revenue as a share of total operating revenue, industry and years of the triple-play practice were statistically significant at the 5 percent level. For all three analyses, interaction effect between years of the triple-play practice and industry wa s not statistically significant. In sum, for video revenue as a share of tota l operating revenue, cable firms are more likely to have higher video revenue share than telephone firms. For voice and data revenue as a share of total operating revenue, te lephone firms are more likely to have higher revenue sh are than cable firms. In addition, there are sign ificant group differences in data revenue as a share of total operating revenue among TPS groups. In both indus tries, firms with 4 years of TPS practice show the highest data revenue share than other TPS firms.

PAGE 141

141 Table 5-28 Results of ANOVA of in dividual service revenue share Dependent Variable Source SS df MS F Variable Mean (SD) Yrs of TPS Indu Video revenue as a share of total revenue 0 Telco 00 (00) Cable .67 (.59) Yrs of TPS .15 5 .03 .29 1 Telco .05 (.03) Cable .54 (.19) Industry 3.14 1 3.14 29.43* 2 Telco .07 (.03) Cable .52 (.18) Yrs of TPS Industry .43 4 .11 1.01 3 Telco .09 (.05) Cable .49(.15) Error 16.74 157 .11 4 Telco .05(n/a) Cable .45 (.15) 5 Telco (n/a) Cable .41 (.12) Voice revenue as a share of total revenue 0 Telco .61 (.28) Yrs of TPS .01 4 .002 .16 1 Cable .09 (.14) Telco .53 (.15) Cable .11 (.13) Industry .01 1 .103 8.55* 2 Telco .55 (.11) Cable .12 (.12) Yrs of TPS Industry .00 3 .001 .07 3 Telco .50 (.13) Cable .14 (.10) Error .85 70 .012 4 Telco .42 (.17) Cable .15 (.08) 5 Telco (n/a) Cable .11 (.12) Data revenue as a share of total revenue 0 Telco .12 (.09) Cable .10 (.07) Yrs of TPS .41 5 .082 15.33* 1 Telco .21 (.07) Cable .16 (.05) Industry .08 1 .077 14.38* 2 Telco .23 (.07) Cable .18 (.05) Yrs of TPS Industry .03 4 .008 1.50 3 Telco .26 (.10) Cable .19 (.05) Error .92 172 .005 4 Telco .30 (.15) Cable .17 (.07) 5 Telco (n/a) Cable .17 (.07) Summary Table 5-29 consolidates the re search questions of this st udy (m arket performance of the triple-play providers) and the results of corres ponding research questions. The practice of the triple-play strategy (RQ7), switching costs (RQ8 ), and competition (RQ9) are associated with subscriber performance of the tr iple-play providers in the cable and telephone industry. Also these factors are associated with revenue performance of the triple-play providers in the cable

PAGE 142

142 and telephone industry. Competition (RQ9) is associat ed with a firms profitability in the cable and telephone industry. From a perspective of gr owth, the practice of th e triple-play strategy (RQ7), switching costs (RQ8), and competition (RQ9) are associated with growth rates of subscriber and revenue market sh are in the cable and telephone i ndustry. There are differences in market performance among three services (i.e ., video, voice, and data) between cable and telephone firms with the practice of triple-play (RQ10) Detailed findings will be provided in Table 5-30.

PAGE 143

143 Table 5-29 Summary table fo r the use of triple-play Research question Result I. Market performance of the triple-p lay providers Impact Detailed finding Cable RQ7. Do triple-play bundles influence tripleplay providers market performance such as subscriber performance, financial performance, and growth rates? Yes 1. Subscriber performance a. As cable firms use the triple-play strategy, their video subscriber performance is more likely to improve. b. As cable firms use the triple-play strategy, their voice subscriber performance is more likely to improve. c. The use of triple-play has no significant impact on data subscriber performance. 2. Financial performance 2-1. Revenue performance a. The use of triple-play has no significant impact on video revenue performance. b. As cable firms use the triple-play strategy, their voice subscriber performance is more likely to improve. c. The use of triple-play has no significant impact on data revenue performance. 2-2. Profitability a. The use of triple-play ha s no significant impact on a firms profitability. 3. Growth a. The use of triple-play is associated with a decline in the growth rates of voice subscrib ers for the cable firms. b. The use of triple-play is pos itively associated with the growth rates of data revenue for the cable firms.

PAGE 144

144 Table 5-29 Continued Research question Result I. Market performance of the triple-p lay providers Impact Detailed finding Cable RQ8. Do switching cost s influence triple-play providers market performance such as subscriber performance, financial performance, and growth rates? Yes 1. Subscriber performance a. The longer cable firms lock in their customers with a contract, the more they are like ly to improve video subscriber performance. b. Switching cost related fact ors have no significant impact on voice subscriber performance. c. As cable firms charge higher early termination fees, they tend to improve data subscriber performance. 2. Financial performance 2-1. Revenue performance a. The longer cable firms lock in their customers with a contract, the more they are li kely to improve video revenue performance. b. The longer cable firms lock in their customers with a contract, the less they are lik ely to improve voice revenue performance. c. As cable firms charge higher early termination fees, they tend to improve data revenue performance. 2-2. Profitability a. Switching cost related fact ors have no significant impact on a firms profitability. 3. Growth a. Contract term, one of switc hing cost factors, is associated with a decline in the growth rates of voice revenue for cable firms. b. Early termination fee, one of switching cost factors, is associated with the growth in a cable firms profitability.

PAGE 145

145 Table 5-29 Continued Research question Result I. Market performance of the triple-p lay providers Impact Detailed finding Cable RQ9. Does competition influence triple-play providers market performance, such as subscriber performance, financial performance, and growth rates? Yes 1. Subscriber performance a. Video market competiti on has no significant impact on video subscriber performance for the triple-play providers in the cable industry. b. A higher level of platfo rm competition is negatively associated with better video subscriber performance. c. A higher level of voice market competition is negatively associated with better voice subscriber performance. d. A higher level of platfo rm competition is negatively associated with better voice subscriber performance. e. A higher level of data market competition is negatively associated with better data subscriber performance. f. A higher level of platform competition is positively associated with better data subscriber performance. 2. Financial performance 2-1. Revenue performance a. A higher level of video market competition is negatively associated with better video revenue performance. b. A higher level of platfo rm competition is negatively associated with better video revenue performance. c. A higher level of voice market competition is negatively associated with better voi ce revenue performance. d. A higher level of platfo rm competition is negatively associated with better voi ce revenue performance. e. A higher level of data market competition is negatively associated with better data revenue performance. f. A higher level of platform competition is positively associated with better data revenue performance.

PAGE 146

146 Table 5-29 Continued Research question Result I. Market performance of the triple-p lay providers Impact Detailed finding Cable RQ9. Does competition influence triple-play providers market performance, such as subscriber performance, financial performance, and growth rates? Yes2-2. Profitability a. A higher level of video market competition is negatively associated with a cable firms profitability. b. A higher level of voice market competition is negatively associated with a cable firms profitability. c. A higher level of plat form competition is negatively associated with a cable firms profitability. 3. Growth a. A higher level of video market competition is negatively associated with the growth rate s of video subscribers and video revenue. b. A higher level of platform competition is positively associated with the growth rate s of video subscribers and video revenue. c. A higher level of voice market competition is positively associated with the growth rate s of voice subscribers and voice revenue. d. A higher level of platform competition is positively associated with the growth rate s of voice subscribers and voice revenue. e. A higher level of data market competition is positively associated with the growth rates of data subscribers and data revenue. f. A higher level of platform competition is negatively associated with the growth rates of data subscribers and data revenue.

PAGE 147

147 Table 5-29 Continued Research question Result I. Market performance of the triple-p lay providers Impact Detailed finding Cable RQ9. Does competition influence triple-play providers market performance, such as subscriber performance, financial performance, and growth rates? Yes3. Growth continued g. A higher level of video and voice market competition is negatively associated with the growth in a cable firms profitability. h. A higher level of data market competition is positively associated with the growth in a cable firms profitability. i. A higher level of plat form competition is negatively associated with the growth rates in a cable firms profitability. Telco RQ7. Do triple-play bundles influence tripleplay providers market performance such as subscriber performance, financial performance, and growth rates? Yes 1. Subscriber performance a. As telephone firms use th e triple-play stra tegy, their video subscriber performance is more likely to improve. b. As telephone firms use the triple-play strategy, their voice subscriber performance is more likely to improve. c. As telephone firms use th e triple-play stra tegy, their data subscriber performance is more likely to improve. 2. Financial performance 2-1. Revenue performance a. As telephone firms use the triple-play stra tegy, their video revenue performance is more likely to improve. b. The use of triple-play ha s no significant impact on voice revenue performance. c. As telephone firms use th e triple-play stra tegy, their data revenue performance is more likely to improve. 2-2. Profitability a. The use of triple-play ha s no significant impact on a firms profitability. 3. Growth a. The use of triple-play is pos itively associated with the growth rates of video subscribers and video revenue for the telephone firms.

PAGE 148

148 Table 5-29 Continued Research question Result I. Market performance of the triple-play providers Impact Details TelcoRQ8. Do switching costs influence triple-play providers market performance such as subscriber performance, financial performance, and growth rates? Yes 1. Subscriber performance a. As telephone firms charge higher early termination fees, they tend to improve video subscriber performance. b. The longer telephone firms lock in their customers with a contract, the more they are likely to improve voice subscriber performance. Also as they charge higher early termination fees, they tend to improve voice subscriber performance. c. Switching cost relate d factors have no significant impact on data subscriber performance. 2. Financial performance 2-1. Revenue performance a. Switching cost relate d factors have no significant impact on video revenue performance. b. The longer telephone firms lock in their customers with a contract, the less they are likely to improve voice revenue performance. c. Switching cost relate d factors have no significant impact on data revenue performance. 2-2. Profitability a. Switching cost relate d factors have no significant impact on a firms profitability. 3. Growth a. Contract term, one of switching cost factors, is associated with the growth rates of data subscribers and revenue for telephone firms.

PAGE 149

149 Table 5-29 Continued Research question Result I. Market performance of the triple-play providers Impact Details Telco RQ9. Does competition influence triple-play providers market performance, such as subscriber performance, financial performance, and growth rates? Yes 1. Subscriber performance a. A higher level of video market competition is negatively associated with better video revenue performance for telephone firms. b. A higher level of platform competition is positively associated with better video subscriber performance. c. A higher level of voice market competition is positively associated with better voice subscriber performance. d. A higher level of platform competition is positively associated with better voice subscriber performance. e. A higher level of data market competition is negatively associated with better data subscriber performance. f. A higher level of platform competition is positively associated with better data subscriber performance. 2. Financial performance 2-1. Revenue performance a. A higher level of video market competition is negatively associated with better video revenue performance. b. A higher level of platform competition is positively associated with better video revenue performance. c. A higher level of voice market competition is negatively associated with better voice revenue performance. d. A higher level of platform competition is negatively associated with better voice revenue performance.

PAGE 150

150 Table 5-29 Continued Research question Result I. Market performance of the triple-play providers Impact Details Telco RQ9. Does competition influence triple-play providers market performance, such as subscriber performance, financial performance, and growth rates? Yes 2-1. Revenue performance continued e. A higher level of data market competition is positively associated with better data revenue performance. f. A higher level of platform competition is negatively associated with better data revenue performance. 2-2. Profitability a. A higher level of video market competition is negatively associated with a firms profitability. b. A higher level of voice market competition is negatively associated with a firms profitability. c. A higher level of data market competition is positively associated with a firms profitability. d. A higher level of platform competition is negatively associated with a firms profitability. 3. Growth a. A higher level of video market competition is negatively associated with the growth rates of video subscribers and video revenue. b. A higher level of platform competition is positively associated with the growth rates of video subscribers and video revenue. c. A higher level of voice market competition is negatively associated with the growth rates of voice subscribers and voice revenue. d. A higher level of platform competition is negatively associated with the growth ra tes of voice subscribers and voice revenue.

PAGE 151

151 Table 5-29 Continued Research question Result I. Market performance of the triple-play providers Impact Details Telco RQ9. Does competition influence triple-play providers market performance, such as subscriber performance, financial performance, and growth rates? Yes 3. Growth continued e. A higher level of data market competition is positively associated with the growth rates of data subscribers and data revenue. f. A higher level of platform competition is negatively associated with the growth rates of data subscribers and data revenue. g. A higher level of video and voice market competition is positively associated with the growth in a cable firms profitability. h. A higher level of data market competition is negatively associated with th e growth in a cable firms profitability. i. A higher level of platform competition is positively associated with the growth rates in a cable firms profitability. II. Resource based-view of strategy RQ10. Are there differences in market performance among three services (i.e., video, voice, and data) between cable and telephone firms with the practice of triple-play Yes 1. For video revenue as a share of total operating revenue, cable firms tend to have higher video share than telephone firms. For voice and data revenue as a share of total operating reve nue, telephone firms tend to have higher revenue shar e than cable firms.

PAGE 152

152 The Impact of Triple-Play on Market Entry In Model III, this study exam ines the impacts of the triple-play strate gy on market entry in the high-speed fixed broadband market, contro lling for other variables. RQ11, RQ12, and RQ13 address the impact of the triple -play strategy on market entry. Th e dependent variable was the number of provider of high-speed fixed lines. Data and Descriptive Statistics for Model III In Model III, this study employed secondary data which consisted of cross-section (crossstate data) and time variance (year from 2000 to 2007) The data were mostly collected from the Federal Communications Commission (F CC), and the US Census Bureau. A total of 273 observations were available to examine the impact of the triple-play strategy on market entry (measured by the number of provi ders of high-speed fi xed lines) in the highspeed fixed broadband market. Several states were eliminated because of missing variables and thus 42 state data were employed in this analysis. Depende nt variables were the number of broadband service provider. A total of 10 different explanatory variables (6 main variables and 4 control variables) were employed. Table 5-30 illustrates descriptive statistics for Model III. The mean of the number of providers of high-speed fixed lines was 34.78 between 2000 and 2007. Table 5-30 Descriptive statistics for market entry in the high-speed fixed broadband market Dependent variable Mean(SD) N Number of providers of high-speed lines 34.78 (25.90) 273 Variables Mean (S.D.) Practice of triple-play: Total TPS subscribers 4619.05 (10842.35) Competitive environment: Internet service provider market 1244.89 (482.42) Platform competition 4521.27 (1158.26) High speed fixed broadband lines per 100 inhabitants 11.10 (8.39) Population density 178.95 (235.34) Income per capita 31480.09 (5084.59) Previous high-speed fixed broadband per 100 inhabitants 7.35 (5.61) LLU regulation price 32.82 (325.66)

PAGE 153

153 Regression Analysis of the Impact of Triple-Play on Market Entry This study employed pooled Ordinary Least Square (OLS) regression with tim e-fixed effect in order to examine the impact of the trip le-play strategy on market entry in the high-speed fixed broadband market. In Model III, this study did not employ state-leve l fixed effects because explanatory variables did not vary enough for states across the sample22. This study conducted correlation analysis in or der to prevent potenti al multicollinearity problems. To assess the strength of correlati ons, the .80 Pearson correlation criterion was employed. All eight explanat ory variables were included to examine the impact of the triple-play strategy on the number of providers. Pooled Ordinary Least Square (OLS) model w ith time-fixed effects This study conducted pooled OLS regression by taking time-fixed effects, but not statefixed effects, into account to examine the impact of proposed explanatory variables on entry of the high-speed fixed line providers in the high speed fixed broadband market. Table 5-31 provides the results of the regres sion of market entry. The pooled OLS model was statistically significant at the 1 percent level, F(12, 41)=7649.37, p<.001 for the number of providers. Thus F test supports that time-fixed effects are significant across all dependent variables at the 1 percent level. 22 This study initially employed both timeand state-fixed effects using timeand state-dummies. However, statefixed dummies were drop ped because explanatory variables did not vary enough for states across the sample. Thus this study reasonably assumed that there is homogeneity among states across the sample.

PAGE 154

154 Table 5-31 Results of regression of market entry Variable Number of providers of high-speed fixed lines (n=273) Coefficient B t statistics Practice of triple-play: Total TPS subscribers .003 4.26* Competitive environment: Internet service provider market .003 .87 Platform competition -.001 -.92 High speed fixed broadband lines per 100 inhabitants -1.27 -2.37* Population density 8.78e-6 1.05 Income per capita -.0009 -.61 Previous high-speed fixed broadband per 100 inhabitants -1.02 -.97 LLU regulation price -.0008 -2.02* Constant -16.01 -.20 R square .763 Statistically significant at the 5% level RQ11 addresses the impact of the triple-pla y strategy on market entry. For the number of providers of high-speed fixed lines, the practice of the triple-play strategy, which was measured by total triple-play subscribers, was statistically significant at the 5 percent level. As firms implement the triple-play strategy (hence as the nu mber of triple-play subscriber increases), the number of providers of high-speed fixed lines is more likely to increase. In other words, market entry is more likely to occur. RQ13 addresses the impact of LLU regulat ion on market entry. Local Loop Unbundling (LLU) regulation price was statis tically significant at the 5 percent level. As LLU regulation price decreases, the number of providers of high-sp eed fixed lines is more likely to increase. RQ12 addresses the impact of competitive envi ronment on market entry. None of variables was statistically significant at the 5 percent level. Table 5-32 summarizes signifi cant factors of market entry in the highspeed fixed broadband market. As firms implement the triple-p lay strategy, the number of providers of highspeed fixed lines tends to increase overtime. In addition, a lower level of LLU regulation price comes into play in increasing market entr y with the triple-pla y strategy. Competitive environment factors have no statistically significant impact on market entry.

PAGE 155

155 Table 5-32 Significant factors of market entr y in the high-speed fixed broadband market Variable Number of providers of high-speed fixed lines Practice of triple-play + Market competition Platform competition LLU regulation Significant at the 5 percent level + Positive: Positive sign of coefficient, -: Negative sign of coefficient Summary Table 5-33 consolidates the research questi ons of this study (m arket entry in the highspeed fixed broadband market) and the results of corresponding research questions. The practice of the triple-play strategy (RQ11) is associated with market entry. LLU regulation (RQ13) is also associated with market entry. However competition (including both competitive environment and platform competition) has no significant imp act on market entry (RQ12). Detailed findings will be provided in Table 5-33. Table 5-33 Summary of significan t factors of market entry Research question Result Market entry Impact Detailed finding RQ11. Do triple-play bundles influence market entry in the broadband market? Yes As firms implement the triple-play strategy, the number of providers of high-speed fixed lines tends to increase but at decreasing rates. RQ12. Does competition influence market entry in the broadband market when the tripleplay strategy is practiced? No RQ13. Does LLU regulation influence market entry in the broadband market when the tripleplay strategy is practiced? Yes A lower level of LLU regulation price comes into play in increasing market entry with the triple -play strategy.

PAGE 156

156 CHAPTER 6 DISCUSSION AND CONCLUSION This study exam ines the potential determinants of the use of the tr iple-play strategy in cable and telephone industries. In addition, this study assesses the impact of the triple-play strategy on market performance of triple-play providers, and its impact on market entry in the high-speed fixed broadband market. This chapter summarizes the empirical results and analyses. Implications and limitations of this study, and sugge stions for future research are also discussed in this chapter. Summary Results and Analyses The Drivers of the Use of Triple-Play This study exam ined market and firm factors of the use of the triple -play strategy in the cable and telephone industries, respectively. For the market factor, thre e market competition factors (i.e., video, voice, and data market competition), platform competition, infrastructure, and market potential factors were assessed. For the firm factor, firm size and cost factors were analyzed. Comparison of the cable and telephone industries Based on descriptive statistics of the use of the triple-play strategy in the cable and telephone industries (see Tables 51 (pg. 91) and 5-2 (pg. 92)), this study describes differences in factors between t hose two sectors. The descriptive statistics indicate that cab le firms began to engage in the triple-play strategy earlier than telephone companies. The mean of the use of the triple-p lay strategy in the cable industry was 32.24 (unit: mo nth) while the mean in the telephone industry was 25.6 (unit: month). It indicates that cable firms generally engage in the triple-play strategy about 6 months ahead of telephone firms.

PAGE 157

157 For the market factor, the means of compe tition factors were similar between industries because these factors were captured at the na tional level. The mean of high-speed fixed broadband lines per 100 inhabitants in cable serv ice areas (29.96) was higher than that in telephone service areas (17.65). Also, population density in cable service areas (147.35) was higher than that in telephone service areas (120.3 5). However, income per capita in cable service areas (32329.31) was lower than that in telephone service areas (32890.10). For the firm factor, the firm size of cable (3993.14 million, unit: USD), measured by total operating revenue, was much smaller than th at of telephone (10731.12 million, unit: USD) on average. However, cable firms generally spent more in marketing and capital expenditures as a share of total operating revenue than tele phone firms, 29.4% vs. 20.8%, and 19.3% vs. 15.4%, respectively. The impact of competition RQ1 addresses the im pact of competition on the use of the triple-pla y strategy. The results of the regression suggested that video, voice, an d data market competition were statistically significant drivers of the use of the triple-play strategy in the cable i ndustry. In addition, the results indicate that video, voice, and data market competition were significant contributors to the use of the triple-pla y strategy in the telephone industries. In addition, plat form competition was a very significant factor in the use of triple-play in both industries. However, how market competition and platform competition motivate the firm to employ the triple-play strategy is different in the two industries. Recent studies suggest that the purpose of the triple-play strategy is 1) to protect firms market share and to reduce consumer turnover (Crampes & Hollander, 2006), 2) to maintain their dominant market posit ions (Bauer, 2006), 3) to offset the loss from competition (Baranes & LeBlank, 2006), and 4) to relax competition through differentiation via bundling (Carbajo et al., 1990; Chen, 1997).

PAGE 158

158 Table 6-1 summarizes the results of the regre ssion of the triple-play strategy use in the cable and telephone industries. RQ1a addresses the impact of competition on the use of the triple-play strategy in the cable industry. In the cable industry, firms are more likely to employ the triple-play strategy as the voice (long-dist ance) market becomes concentrated. Stated differently, firms are more likely to employ the trip le-play strategy as the level of competition in the voice market is relatively low. Market co mpetition was measured by Herfindall-Hirschman Index (HHI), and thus positive signs of coeffici ents represent decreasing levels of competition. The emergence of innovative voice transmissi on technology, Voice over Internet Protocol (VoIP), has enabled cable firms to generate a ne w revenue stream. This VoIP will threaten the traditional telephone fixed-line in the near future, ma inly due to the benefits of VoIP in terms of low cost of service and potential value-added services (Zimmerman, 2007). It is likely that the cable firms might have attempted to incorporat e a competitive, substitutable product into their bundles to compete with the telepho ne companies that provide other triple-play se rvices so to protect their overall market shares. Thus cable firms aggressively intr oduced VoIP within the triple-play service by expecting VoIPs displ acement of traditional fixed telephone lines. Table 6-1 Significant factors of the use of triple-play Explanatory variable Cables triple-play Telcos triple-play Video market competition (MVPDs market) + Voice market competition (Long distance market) + Data market competition (ISPs market) + Platform competition + High-speed fixed broadband lines per 100 inhabitants Population density Income per capita + + Firm size (Total operating revenue) + Marketing expenses Capital expenditures Significant at the 5 percent level + Positive: Positive sign of coefficient, -: Negative sign of coefficient

PAGE 159

159 The results of the regression show that cable firms tend to implement the triple-play strategy as the level of video market competiti on increases in the cable industry. Firms employ the triple-play strategy as the video market beco me concentrated. Considering that cable industry ranked first in terms of the number of subs cribers among all multich annel video programming distributors (MVPD) (FCC, 2009), cable firms might use the trip le-play strategy in order to maintain their dominant market positions in the video market. On the other hand, the result of the regression shows that cable firms tend to implement the triple-play strategy as the level of data mark et competition (competition among ISPs) increases. This implies that cable firms tend to employ the triple-play strategy to protect their market share and to avoid consumer turnover in data markets where new rivals continuously enter. Unlike the data market, cable firms are likely to use the triple-play strategy as platform competition is relatively low. Considering that cable modem controlled two-third of the broadband market in 2007 (FCC, 2007), cable firms tend to employ the trip le-play strategy in an effort to maintain their current market position. RQ1b addresses the impact of competition on the use of the triple-play strategy in the telephone industry. For the telephone industry, the result suggests th at firms are more likely to employ the triple-play strategy as the video market becomes competitive. Competition in the video market increased as a result of the Te lecommunications Act of 1996 and technological convergence, which provided telephone firms with an opportunity to develop new revenue streams using video services. In addition to th e possibility of generating additional revenue, telephone firms engage in the triple-play strategy in order to protect their overall market share as competitive responses to cables triple-play strategy.

PAGE 160

160 The result also suggests that telephone firms use the triple-play strategy when the voice market becomes competitive. The decline of fixed telephone lines has continued since 2001 (FCC, 2008)23. To bundle telephone services with other services, telephone firms may expect to decrease consumer turnover and thus recover losses from competition in the fixed-line long distance market. The result suggests that telephone firms are lik ely to use the triple-play strategy when the ISPs (data) market becomes less competitive. Considering that over a half of the top 10 ISPs are telephone firms in the ISPs market24, the result implies that telephone firms attempt to maintain their current market position through bundling by preventing rivals entry. Lastly, the result suggests that telephone firms implement the triple-play strategy as platform competition increases. This implies that telephone firms tend to employ the triple-play strategy to protect their market share and to avoid consumer turnover as platform competition intensifies. RQ1c compares the impact of competition on the use of the triple-play strategy between the cable and telephone i ndustries. The results suggest that cable firms use the triple-play strategy when the market becomes less competitive and they have a relatively dominant position. It is likely that cable firms see the triple-play st rategy as a device enabling them to maintain their current market position, protect their market sh are, and advance into a new market with an attractive, competitive product. For telephone firms, the results indicate that telephone firms use the triple-play strategy when the market beco mes competitive, which implies that telephone 23 According to FCC (2008), fixed telephone lines declined by 4.4% from 2005 to 2006, 1.4% from 2004 to 2005, 2.9% from 2003 to 2004, 3.3% from 2002 to 2003. 24 In 2008, the top 10 ISPs were SBC_ATT (market share: 15.4 %), Comcast (15.3%), RoadRunner (9%), Verizon (8.8%), America Online (7.7%), EarthL ink (3.1%), Charter (3%), Qwest (2.9 %), Cablevision (2.5%), and United Online (1.5%) (Jupiter Research, 2008).

PAGE 161

161 firms use the triple-play strategy as a mere defe nsive strategy. In other words, telephone firms might launch the triple-play bundle in order to protect against consumer turnover and/or to recover losses from new competition. Overall, all results point to the fact that bot h cable and telephone firms use the triple-play strategy to prevent rivals (eithe r incumbent or potential entrant) from taking their market share. Specifically there is a sense of mutual depende ncy in this process. That is, when a firm contemplates the use of the trip le-play strategy, it would also cons ider whether its rivals would employ the strategy. It happens mostly because if a firm continues to sell services independently while its rival offers the triple-play bundle, the firm will be at a disadva ntage over its rival, and consequently lose market share. The impact of market potential RQ3 addresses the im pact of market potential on the use of the triple -play strategy. Market potential usually drives communications firms to develop new innovations in order to increase their market share (Grubesic & Murray, 2004; Se o, 2008). The results of the regression suggest that market potential factors ar e statistically significant drivers of the use of the triple-play strategy in the cable and telephone industries. The income variable is positively associated with the likelihood of the use of the triple-play strategy in both cable and telephone industries. Cable and telephone companies are more likely to use the triple-play strategy when the income elasticity of demand is higher than when the income elasticity is lower. This result is consistent with the finding by Seo (2008), which revealed that in areas where market demand is higher, a firm is more likely to offer triple-play services. The impact of firm size RQ4 addresses the im pact of firm size on the use of the triple-play strategy. In general, larger firms have sufficient capital resources to expand their business (Dean et al., 1998; Gentry

PAGE 162

162 & Jamison, 2005). As a firm factor, firm size was st atistically significant in the use of the tripleplay strategy in the telephone industry. Larger telephone firms are more likely to implement the triple-play strategy than smaller firms. It points to the fact that larger firms have more resources to venture into non-core product area s. It is also likely that la rger firms have larger network infrastructure, which often leads to a cost advantage for delivering triple-play services. For the use of the triple-play strategy in the cable i ndustry, however, firm size had no statistically significant impact on the use of the triple-play strategy. The impact of other factors RQ2 addresses the im pact of information co mmunication technology on the use of tripleplay in the cable and telephone industry, which wa s not statistically sign ificant. RQ5 addresses the impact of cost on the use of the triple-play strategy, which was not stat istically significant. Strategic orientation RQ6 addresses the differences in strategic behaviors of the triple-play service providers by industry in term s of generic strategies. This stud y analyzes differences in strategic behaviors of the triple-play service providers by industry in terms of the ge neric strategies. First, this study conducted multiple discrimina nt function analyses in order to examine how firm factors (i.e., firm size, marketing expe nses, and capital expend itures) play different roles on groups with 1 year of the triple-play service (TPS), 2 y ears of TPS, 3 years of TPS, 4 years of TPS, and 5 years of TPS. For the cable industry, the result s uggests that there is a significant group difference between firms with 1 and 2 year(s) of TPS and firms with 3, 4, and 5 years of TPS, and marketing expenses were the best predictor for this difference. Moreover, market expense accounts for the group difference between firms with 4 years and firm with 5 years of TPS. Lastly, there was a group differen ce between firms with 1 year and firms with 2

PAGE 163

163 years of TPS. In this case, firm size discrimi nated the most, capital expenditures were next, and marketing expenses the least. Figure 6-1 illust rates how firm variables discriminate groups. Figure 6-1 Discrimination of firms by fi rm variables in the cable industry. These results suggest that marketing becomes more important as the cable firms becomes more vetted in the triple-play strategy. Marketing expenses increase for firms with 1 year of TPS to firms with 2 years of TPS, from firms with 2 years of TPS to firms with 3 and 4 years of TPS, and from firms with 3 and 4 years of TPS to firm s with 5 years of TPS. The results also suggest that there is a difference in firm size and capital expenditures between firm s with 1 year of TPS and firms with 2 years of TPS. In terms of firm size, smaller firms enter the triple-play market later than larger firms, which supports previ ous findings that small firms tend to be less aggressive in offering the triple-play service (Dea n et al., 1998; Gentry & Jamison, 2005). Larger firms are more likely to invest in capital expe nditures in the provision of the triple-play service as well. Regarding the telephone industry, the result shows there is a significant group difference between firms with 1 and 2 year(s) of TPS a nd firms with 3 and 4 y ears of TPS. Capital

PAGE 164

164 expenditures were the best predictor for this diffe rence, marketing expenses were next, and firm size the least. Next, there was a group difference between firms with 3 years and firms with 4 years of TPS. All three firm f actors account for the difference, however the contribution of each factor is different from the previous case of difference (i.e., firms with 1 and 2 years of TPS vs. firms with 3 and 4 years of TPS). In this case, fi rm size was the best predictor for this difference, marketing expenses were next, and capital e xpenditures the least. Lastly, there was a group difference between firms with 1 year and firms with 2 years of TPS, and firm size contributed to this difference. Figure 6-2 illustrates how firm variables discriminate groups. Figure 6-2 Discrimination of firms by firm variables in the telephone industry. These results imply that, for the telephone industry, firm size is the most important factor in executing the triple-play strategy. Specifically, larger firms begin to use the triple-play strategy earlier than sma ller firms. This might occur because firm size is closely related to market power and resources with respect to capital (Faulhaber & Hogendorn, 2000; Koski & Majumdarm, 2002; Leibowitz & Margo lis, 2002). As mentioned earlier, it is a costly endeavor to enter into the video market because of the need to provide the infrastructure and access to content simultaneously. Along with firm size, the results suggest that the earlier firms start the

PAGE 165

165 triple-play strategy, the greater they invest in marketi ng (i.e., marketing expenses) and innovations and infrastructure (i.e., capital expenditures). In summary, there are differences in contributing firm f actors among cable and telephone industries executing the triple-play strategy. Cabl e firms are more likely to focus on how to persuade consumers via marketing activities. Howe ver, telephone firms are heavily dependent on their size in the provision of the triple-play, and firm size is an important determinant of marketing and investment. This might be explai ned by the technical difference between networks of cable and telephone firms in delivering the triple-play service (J anssen & Mendys-Kamphorst, 2008). Whereas cable firms are able to economically carry IP relevant se rvices (especially voice transmission using VoIP) within the triple-play service over their netw ork, telephone firms need higher bandwidth capacity to deliv er video services within the bundle, which requires substantial investment in infrastructure. For this reason, cable firms are more easily able to launch the tripleplay service at a lower cost th an telephone firms. Thus cable fi rms are less dependent upon their size, and rather concentrate on marketing. Yet, telephone firms face substantial costs to upgrade their network infrastructure to deliver the triple-p lay service, which is pos sible more feasible for larger firms at this time. Another statistical analysis, Analysis of Variance (ANOVA), was performed to examine differences in marketing expenses and capital expenditures by industry as we ll as the years of the triple-play practice. This st udy controlled for firm size by dividing marketing expenses and capital expenditures by total operating revenue, and generated two new dependent variables: marketing expenses as a share of revenue and capital expenditures as a share of revenue. The result of ANOVA shows that marketing e xpenses as a share of revenue is different between the cable and telephone indu stries. In general, the cable industry is more likely to invest

PAGE 166

166 in marketing than the telephone industry. In add ition, capital expenditures as a share of revenue are different between the cable and telephone indust ries. Cable firms are more likely to make an investment in their network platfo rms than the telephone industry. The results suggest that cable firms attempt to differentiate themselves with the tripleplay strategy via marketing activities as well as the improvement of their services or network infrastructure with substantia l capital expenditures. Comparativ ely, the telephone firms increase their marketing expenses and capital expenditure s very gradually, and they are less likely to invest in innovations and infrastructure than cable firms. These findings are consistent with previous results of multiple discriminant function analysis. From a regulatory perspective, it is widely supported that regulat ed firms invest in innovation or infrastructure much less than unreg ulated firms (Crandall, 2009; Alesina et al., 2003). The Telecommunications Act of 1996 requi red telephone companies to make their facilities available to all unaffiliated Internet Service Providers (ISPs) in a nondiscriminatory basis. In other words, despite the substantial cost of upgrading infras tructure and innovating technologies, telephone companies should allow other rival ISPs to utilize their network. Consequently telephone firms are reluctant to in vest because they may not achieve expected sales or returns on expenditures in the short-term (Wellenius & Townsend, 2005). The finding here supports the same notion proposed by earlier scholars. Based on the results, this study concludes that cable and telephone firms show different strategic modes in the provision of the triple-play service. Cable firms are more likely to invest in marketing which often promote the uniquene ss of services and co mmunicate to consumers why their services are valuable. In addition, cable firms tend to spend more capital expenditures to upgrade their platforms or to develop their se rvices and/or technol ogies. Such investment

PAGE 167

167 enables cable firms to create value-added servic es, which helps them further their ability to differentiate. In fact, cable firm s offered the triple-play service at significantly discounted price initially to compete with the te lephone companies. However, they have continuously increased the price of the triple-play se rvice by differentiating themselves (Lee, 2009; Lee & Lee, 2008). On the other hand, telephone firms tend to inve st less in marketing, infrastructure, and/or innovative services than cable firms. Histori cally, telephone firms have been regulated as universal service providers rather than as profit maximizers. Th is tradition may force telephone firms to focus on the low cost strategy in the provision of triple-pla y. In addition, regulation influences firms reluctance to invest in networ k infrastructure (Crandall 2009; Alesina et al., 2003). As suggested, telephone firms have to lease their network to other rivals (unaffiliated ISPs), which may result in slow return on invest ment. Consequently, it re moves the incentive to invest in infrastructure (Janssen & Mendys -Kamphorst, 2008; Welle nius & Townsend, 2005). Rather, telephone firms tend to fo rm strategic alliances with direct broadcas t satellite providers to acquire video content, instead of investing in their infrastruc ture to carry high bandwidth IP service (video). For these reasons telephone firms might decide to enter the triple-play market later than cable firms, and their triple-play be haviors might be more reactive or defensive comparatively. Overall, the unique characteristics and envir onmental conditions of industry drive cable and telephone firms to choose different strate gic modes in the provision of the triple-play strategy. Initially, both industries concentrated on price competi tion by offering the triple-play service at a significant discount to attract consumers (Lee, 2009). However, as firms became more experienced, cable firms have gradually focused on differentiation by investing marketing

PAGE 168

168 and innovations while telephone firms implement triple-play by saving marketing expenses and innovations but by focusing on strategic alliances. The Impact of Triple-Play on Market Pe rformance of th e Triple-Play Providers This study examined the impact of the triple -play strategy on market performance of the triple-play providers in the cab le and telephone industries, resp ectively. For the strategy factor, the practice of the triple-play strategy and switching costs were employed. Market competition (video, voice, and data market competition) an d platform competition were assessed jointly with the impact of the triple-play strategy while cont rolling for other factors su ch as infrastructure, market potential factors, firm size and cost factor s. Market performance he re consists of three categories: subscriber performance, financial performance, and growth rates. The impact of the triple-p lay strategy on subscriber and financial performance RQ7 addresses the impact of triple-play on market performance (i.e., subscriber and financial performance) in the cable and telephon e industry. Regarding the cable industry, the results of the regression indicated that the triple -play strategy was statistically significant in increasing video and voice subscribers as well as voice revenue. For the telephone industry, the results of the regression suggested that the triple-play strategy was statistically significant in increasing subscribers of all three services as well as video and data revenue. Regarding EBITDA, as a measure of a firms profitabilit y, the triple-play strategy was not statistically significant in either cable or telephone indus try. Table 6-2 summarizes the results of the regression regarding the impact of the triple-play strategy on mark et performance of the tripleplay providers in the cable and telephone industries.

PAGE 169

169 Table 6-2 Significant factors of market performance Industry Dependent variable Explanatory variable Triple-play Early termination fee Contract term Video market competition Voice market competition Data market competition Platform competition Cable Video subscriber + + + Voice subscriber + + + Data subscriber + + Video revenue + + + Voice revenue + + + Data revenue + + EBITDA + + + Growth: Video subscriber + Growth: Voice subscriber Growth: Data subscriber + Growth: Video revenue + Growth: Voice revenue Growth: Data revenue + + Growth: EBITDA + + + + Telco Video subscriber + + + Voice subscriber + + Data subscriber + + + Video revenue + + Voice revenue + + + Data revenue + + EBITDA + + + Growth: Video subscriber + + Growth: Voice subscriber + + Growth: Data subscriber + + Growth: Video revenue + + + Growth: Voice revenue + + Growth: Data revenue + + Growth: EBITDA + Significant at the 5 percent level + Positive: Positive sign of coefficient; -: Negative sign of coefficient

PAGE 170

170 Overall, it appears that the triple-play st rategy helps enhance ove rall subscriber and revenue performance in the cable and telephon e industries. Bundling typically reduces in heterogeneous consumers valuations (Adams & Yellen, 1976; Salinger, 1995), which can attract different segments of consumer s who have different valuations for products (Schmalensee, 1984; Stremersch & Tellis, 2002). Consequently, mark et share-type performance of bundle providers can be improved in that bundle providers can extr act consumers surplus (Adams & Yellen, 1976; Bakos & Brynjolfsson, 1999; Stremersch & Tellis, 2002). Moreover, the results suggest that cable a nd telephone firms succeed in introducing a new IP service (i.e., telephony for cable and video se rvice for telco) using the triple-p lay strategy. Cable firms successfully enhance voice performance (both subscriber and revenue performance), and telephone firms improve their video performan ce in the number of subscribers and revenues. This happens because if a firm bundles existi ng technology or services with a new one, it can reduce the degree of consumers uncertainty abou t new technology or services (Eppen et al., 1991). However, the triple-play strategy is not likel y to enhance a firms profitability. Although the triple-play strategy helps cable and telephone fi rms increase in the number of subscribers and revenues, adding new services wi thin the triple-play bundle incurs new marketing expenses. So while they expand the cable or telecommunications market, th e strategy might not enhance profitability in the short run. This study additionally employed Analysis of Variance (ANOVA) to examine differences in market performance among firms on the basis of years of the triple-p lay practice (non-TPS, 1 year of TPS, 2 years of TPS, 3 years of TPS, 4 years of TPS, and 5 years of TPS group). The results of ANOVA suggest that th ere are significant group differen ces in voice subscribers and

PAGE 171

171 voice revenue in the cable industry. In particular, firms with 4 years of TPS practice are likely to have higher voice market share than non-TPS-group. It is plausible that there might be a time threshold of relatively better performance. In case of voice performance in the number of subscribers and revenue in the ca ble industry, firms investing in the triple-play strategy for 4 years show better performa nce than non-TPS-firms. With respect to the telephone industry, the re sults of ANOVA indica te that there were statistically significant group differences in video and data subscriber performance, video and data revenue, and EBITDA. In particular, firms w ith 4 years of TPS practice are likely to have higher video and data market share (subscribers and revenue) than other TPS groups (non-TPS, 1 year of TPS, 2 years of TPS, and 3 years of T PS practice, respectively). For firms profitability, firms with 4 years of TPS pract ice achieved the highest profitability compared to other TPS groups (non-TPS, 1 year of TPS, 2 years of TPS, and 3 years of TPS practice, respectively). As the case in the cable industry, it is possible that there is a time thresh old of relatively better performance. In the case of video and data ma rket share in the tele phone industry, firms using the triple-play strategy for 4 years show be tter performance than other TPS firms. Overall, the results point to the fact that there might be a ti me threshold of above average or better performance. In other words, the trip le-play strategy is rela tively successful in enhancing subscriber and financial performance when firms have engaged in the strategy for at least 4 years (in the sample used in this st udy), though the impact of the strategy varies by services. In addition, the strategy seems to be e ffective in improving the financial performance of non-core, newer products such as voice for cable and video for telephone firms. Lastly the strategy has impacted the firms profitability differently between TPS firms in the telephone industry.

PAGE 172

172 The impact of the triple -play strategy on grow th RQ7 addresses the impact of triple-play on market perfor mance (growth rates) in the cable and telephone industry. In te rms of growth, the results of the regression suggest that the triple-play strategy is associated with deceler ated growth rates of voice subscribers while contribute to the growth rate of da ta revenue in the cable industry. It is likely that the triple-play strategy impacts the growth of non-co re services for cable firms, esp ecially the data market share. For the telephone industry, the triple-play strategy positively relates to th e growth rates of both video and data revenue. It suggests that the triple -play strategy is likely to accelerate the growth of both non-core and new services for telephone firms. The results of ANOVA suggest that there were group differences in the growth rates of data subscribers and revenue in the cable indust ry. In particular, firms with 4 years of TPS practice tend to have lower grow th rates of data market share than non-TPS-group. It implies that there might be a threshold of the growth of mark et share in the cable industry. The triple-play strategy initially accelerates the growth of data ma rket share to a certain stage; however after the stage, the growth of data mark et share begins to decline. For the telephone industry, there are group differe nces in the growth rates of video market share among TPS groups. Firms with 2 years of T PS practice tend to have the highest growth rates of video market share than other TPS groups (non-TPS-group, 1 years of TPS group, and 3 years of TPS group). It is possible that there is also a threshold of growth for the video market in the telephone industry. The triple-p lay strategy promotes the growth of video market share for 2 years; however, after the 2 years, the positive imp act of the strategy might eventually slow down. This study found that triple-play generally e nhances subscriber and financial performance for 4 years in both cable and telephone industries. Interestingly, cable firms investing in tripleplay for 5 years, however, show lower market sh are of all three services than other TPS groups.

PAGE 173

173 Firms with 5 years of TPS also show a remarkable reduction of growth rate s of all three services. This pattern cannot be detected in the telephone industry, possibly because the longest history of the triple-play strategy in this industry was 4 ye ars in the sample. Figures 6-3 and 6-4 depict how subscriber and revenue have changed by years of the triple-play strategy in the telephone industry and in the cable industry, respectively. The results suggest that as an increasing number of communica tion firms offers triple-play services, competition with the triple-play bundle be tween cable and telephone firms intensifies, and thus it may start to negativ ely influence their market performance and the growth may begin to level off. Recent studies contend that bundl ing does not always have a positive impact on a sellers profit. This is true especially when two firms compete in a multi-product market with the bundle (Matutues & Regibeau, 1992). In this case, the bundled products will lead to homogenization of customers in that the cons umers preferences for each product will be averaged out due to the nature of the bundle (R eisinger, 2004). Therefore, bundling, by nature, can generate new types of competition, which would result in a reduction in market performance (Fay & MacKie-Mason, 2001). The findings are consistent with the fi ndings by Bughin and Mendonca (2007), which suggest that the triple-play stra tegy helps increase revenue by aggregating different consumer segments, although this impact cannot fully compen sate for losses from competition incurred the triple-play bundles. Tardiff (2007) claimed th at increased competition among the triple-play services will erode the firms market share and di minish their profits in the long run. The results in this study seem to support the proposed notion.

PAGE 174

174 Figure 6-3 Subscriber and revenue by years of triple-play in the cable industry Figure 6-4 Subscriber and revenue by years of triple-play in the telephone industry In sum, the triple-play strategy promotes the growth of data market in the cable industry while contributes to the growth of both non-core services (i.e., vi deo and data) for the telcos. In addition, there might be diminishing returns for pa rticular service market performance after the triple-play providers reach a certa in stage of market share. For example, the triple-play strategy accelerates the growth of data market share of cable firms for 3 years; however after the threshold, the strategy begins to lose its imp act. For the telephone industry, the triple-play strategy accelerates the growth of video market share for 2 years; howev er after the threshold, 0 5000 10000 15000 20000 25000 30000 35000 0yr 1yr2yr3yr4yr Video Sub Voice Sub Data Sub 0 5000 10000 15000 20000 25000 0yr 1yr2yr3yr4yr Video Rev Voice Rev Data Rev 0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000 0yr 1yr2yr3yr4yr5yr Video sub Voice Sub Data Sub 0 500 1000 1500 2000 2500 3000 3500 4000 0yr 1yr2yr3yr4yr5yr Vdieo Rev Voice Rev Data Rev

PAGE 175

175 the impact of the strategy begins to level off. La stly for the cable industry, the triple-play strategy might become less effective for all three services after 4 years of practices. The impact of switching costs RQ8 addresses the im pact of switching costs on market performance for the triple-play providers. The results of the regression showed that switching costs, either early termination fee or the term of contract, were positively associated with increases in video and data subscribers as well as video and data revenue in the cable industry. For the telephon e industry, the results showed that switching costs are asso ciated with increases in video and data subscribers as well as voice revenue. In addition, switching costs contri bute to the growth of data subscribers and revenue in the telephone industry. The results suggest that switching costs, jo intly with the tripleplay strategy, are likely to enhance subscriber and financia l performance in the cable and telephone industry. In addition, accelerating return s on the data market share in the telephone industry are also associated with these switching costs. In general, switching costs help lock in consumers in order to avoid consumer turnover and guarantee future profits (Eppen et al., 1991; Shapiro & Varian, 1999; Shy, 2001), and thereby so ften competition (Gans, 2000; Klemperer, 1995). The results in this study seem to s upport the theories of switching costs. The result of the regression, on the other ha nd, showed that switching costs were unlikely to increase voice revenue in the cable industry. This suggests that switching costs associated with the triple-play strategy may not effectively lock in voice consumers or not soften competition. From a perspective of a firms profitability, switching costs associated with the triple-play strategy generally have no significa nt impact in either the cable or telephone industry. However switching costs do increase the gr owth of profitability among th e cable firms. This phenomenon may relate to the introductory pr icing strategy. That is, cable firm s often set a lower initial price point to compete with other triple-play of tele phone firms (Lee, 2009), a nd then gradually raise

PAGE 176

176 the price of the triple-play pack age after they lock in the subs cribers (Lee & Lee, 2008). Such introductory pricing strategy may contribute to the speedy increases of cable firms profits (Cabral et al., 1999). In sum, switching costs associated with th e triple-play strategy, generally enhance subscriber and financial performance in th e cable and telephone industries by protecting consumer turnover. However, switching costs ar e less likely to increa se voice revenue in the cable industry. For the growth rate s, switching costs increase the grow th of data market share in the telephone industry. Lastly switching costs were positively associated with the growth of profitability only in th e cable industry. The impact of competition RQ9 addresses the im pact of competition on market performance for the triple-play providers. The results of the regression indicate that market competition had statistically significant impacts on market share (i.e., the numbe r of subscribers and re venue) of triple-play providers in the cable industry except for vi deo market competition. For the telephone industry, video, voice, and data market competition had significant impacts on market share (i.e., the number of subscribers and revenue) of all three services. In other words, as competition intensifies in the market, the triple-play service providers tend to lose their market share. This finding supports the findings of previous studi es, which suggest that increased market competition negatively influenced communications firms market share in the long-distance market (Kaserman & Mayo, 2002), in the local telecommunications market (Woroch, 2002), and in the broadband market (Shy, 2001). However market competition seems to play a different role in the grow th of market share. For the cable industry, the results of the regres sion show that higher le vels of voice and data market competition accelerate th e growth of voice and data mark et share while intensive video

PAGE 177

177 market competition decelerates th e growth of video market share. For the telephone industry, the results indicate that a higher level of data mark et competition is positively associated with the growth of data market share while intensive vi deo and voice market comp etition decelerates the growth of video and voice market share. In sum, the results here suggest that a higher level of market competition is negatively associated with increases in market share of all three services in both industries, and that competition is often asso ciated with the growth of market share of specific services (either positively or negatively) in each industry. The results of the regression indicate that platform competition influenced market share as well. For the cable industry, a lower level of platform competition was positively associated with increases in video and voice subscribers a nd revenue while a higher level of platform competition was positively associated with an increase in data subscribers and revenue. For the telephone industry, a higher level of platform co mpetition is positively associated with increases in video and data subscribers and revenue. In addition, a higher level of pl atform competition is positively associated with increases in voice subscribers while a lower level of platform competition is positively associated with increases in voice revenue. As suggested by others, platform competition drives the speedy development of new IP technologies and services, and these IP technolo gies and services might threaten traditional services (Zimmerman, 2007). The tr iple-play service in each indus try includes at least one IP relevant service that is delivered over the netw ork platforms (i.e., VoIP for cable, and video for telco). Cable and telephone firms thereby encounter competition within components of tripleplay services. In particular, cable firms VoIP ha s the potential to canniba lize the market share of fixed-line telephony service of telephone co mpanies (Ferguson, 2002; Grubesic & Murray, 2004; Loomies et al., 2005; Shim & Oh, 2006) while telephone firms new video entertainment such as

PAGE 178

178 IP TV and/or satellite video entertainment mi ght take away multichannel video distributors market share (Bauer, 2006; Crampes & Hollander, 2006; Loomies et al., 2005). The degree of competition intensity in each service market is dependent on the stage of development of the rivals technologies and networks which is influenced by platfo rm competition. Thus platform competition has incurred different impacts on video and voice service performance in each industry. Furthermore, the result shows th at platform competition is associated with an increase in data service performance in both industries. This might happen b ecause platform competition is most likely to impact the broadband deployment (Lee, 2006), which leads to an increase in overall market size (DotEcon & Criterion Economics, 2003). Increased market size gives firms an opportunity to generate new demand and ne w revenue stream by forcing firms to offer improved quality of services and diverse serv ice choices (Denni & Gruber, 2007; Zimmerman, 2007). Hence a higher level of platform competition is positively associated with increases in data service subscriber and revenue in the cable and telephone industries. From a perspective of a firms profitability, the results of the regr ession show that video and voice market competition and platform compet ition had statistically significant impacts in the cable and telephone i ndustry. Specifically, cable and telep hone firms are likely to enhance their profitability as the video and voice markets become less competitive. It is also true when platform competition becomes less competitive. Th ese findings support the findings of previous studies, which suggest that increased market comp etition had a negative impact on firms profits by eroding incumbent firms market share (Zimmerman, 2007). In sum, market competition generally decrea ses market share of video, voice, and data services in the cable and telephone industry. Al so market competition has differently (either

PAGE 179

179 positively or negatively) influenced the growth of market share of specific services in each industry. In addition, platform competition has incurred differe nt impacts on video and voice service performance in each industry. Lastly, lowe r levels of market competition and platform competition are positively associated with the improvement of a firm s profitability in the cable and telephone industry. Resource-Based View (RBV) of Strategy The RBV approach focuses on an individual fi rm s unique capabilities and their impact on the firms business strategy in the market ANOVA analyses were performed with two explanatory variables (i.e., years of the triple-play practice and industry) to examine differences in individual service revenue as a share of total revenue to test RBV research questions (i.e., RQ10). The results of ANOVA suggest that there wa s a difference in video revenue as a share of total operating revenue between th e cable and telephone industries. In particular, the mean of video revenue share in the cable industry was higher than the mean of video in the telephone industry. There was a significant difference in voice revenue as a share of total operating revenue between the cable and telephone industries as well. In this case, the mean of voice revenue share in the telephone industry was higher than the mean of voice in the cab le industry. For data revenue as a share of total operating revenue, th ere were differences between industries as well as the years of the triple-play strategy. The mean of data reve nue share in the telephone industry was higher than the mean of data in the cable indu stry. This was also a di fference in data revenue share by the years of the triple-play strategy implementation. Th e mean of data revenue share of the firms with 3 years of TPS practice was highe r than that of the other TPS groups. Figure 6-5 consolidates three service revenues as a share of to tal revenue by industry. As seen in Figure 6-5, in th e cable industry, video revenue took the highest proportion of total operating revenue, data reve nue share came next, and voice revenue recorded the lowest

PAGE 180

180 proportion. In the telephone industry, however, voi ce revenue took the highest proportion of total operating revenue, data revenue share was second, and video revenue was the lowest. Figure 6-5 Individual servi ce revenue as a share of total revenue by industry A long history of operating traditional servi ces in the cable and telephone industries enables firms to accumulate resources, experience s, and/or knowledge in managing their core services. Such resource, knowledge, or experience leads to competitive advantages over firms new rivals due to their inimitable natures (Die rickx & Cool, 1989). Conse quently, such resources lead to superior performance when used strategically (Barney, 1991; Maijoor & Van Witteloostuijin, 1996). Thus cable firms are lik ely to have a competitive advantage in the provision of video service while telephone firms have a competit ive advantage in the provision of voice service. The result here suggests that cable and telephone firms have acquired superior performance in their core serv ice i.e., cable for video service and telco for voice service, compared with other two services. As firms engage in the triple-play strategy, it is likely that their core service revenue as a share of total revenue might d ecline overtime while new service revenue as a share of total 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0yr 1yr2yr3yr4yr5yr Data share Voice share Video share 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.90yr 1yr2yr3yr4yr Data share Voice share Video share

PAGE 181

181 revenue increases. Though this trend has not been detected statistically, the triple-play strategy seems to have a positive impact on the deploymen t of new services within the bundle. The Impact of Triple-Play on Market Entr y in the High-Speed Fixed Broadband Market RQ11 addresses the im pact of the triple-pla y strategy on market entry in the high-speed fixed broadband market. This study examined the impact of the triple-p lay strategy on entry in the high-speed fixed broadband market, controlli ng simultaneously for othe r exogenous variables. Data market competition, platform compe tition, and local loop unbundling regulation were assessed with the impact of the triple-play stra tegy. The dependent variable was the number of providers of high-speed fixed lines. The impact of the tr iple-play strategy The results of the regression i ndicate that the practice of triple-play strategy is positively associated with increasing market entry in the high-speed fixed broadband market. As the diffusion rate of broadband increases, market size increases, and demand for new services grows (Denni & Gruber, 2007). In add ition, technological convergence allows firms to expand their businesses into new communications service areas by developing innovative technologies and services that attract consumers (Bauer, 2006) or by achieving efficiency in production and distribution processes in terms of time and costs (Antonelli, 1997). Due to the development of communication technologies, firms ar e able to increase their market share and raise profits (Denni & Gruber, 2007). As the triple-play service attracts various segments of consumers, the incumbent firms increasingly engage in the triple-play strategy via their networks. New entrants also begin to use the triple-play strategy. For example, broadband service providers that build their facili ties-based networks start to offer the triple-play bundle over a single network, and the number of these providers cont inues to increase (FCC, 2009). Moreover, utility companies provide (or plan to offer) triple-play services via strategic alliances with broadband

PAGE 182

182 over power line providers, leadi ng to an increase in market entr ance (FCC, 2009). In that light, the practice of triple-play strategy is positively associated with entry in the high-speed fixed broadband market. Nevertheless, as firms implement the triple-play strategy (hence as the number of tripleplay subscriber increases), the growth rate of th e number of providers of high-speed fixed lines is likely to level off. As seen in Figure 6-6, market entry accelerates as firms implement the tripleplay strategy, but at a decreasing rate after 2 years of the triple -play strategy. This implies that there might be a threshold of market entry as sociated with the practice of triple-play. Figure 6-6 Market entry by year s of the triple-play strategy From a regulatory perspective, it has been widely discussed whether the triple-play bundling deters entry. Pernet (2006) asserted the possibility of us ing the triple-play strategy to leverage the dominant position of communications firm in a se rvice into a new market. Peitz (2008) claimed that bundling can e ffectively block entry. Bauer (2005 ) suggested that triple-play providers are able to prevent single service pr oviders from entering the market. If the bundled price is not predatory and bundling practices are based on competitive responses to rivals bundling behavior, bundling may not lead to fore closure in the market (Kuhn, 2004). If bundling 0 10 20 30 40 50 60 70 80 01234 Market Entry 0 0.2 0.4 0.6 0.8 1 1.2 1234 Growth rates of market entry

PAGE 183

183 practices help a firm to leverage its market power into another market, it would generate anticompetitive effects, which may require certain policy tools. The findings of this study i ndicate that triple-play bundli ng could encourage an initial increase of new entries to the market, contrary to the previous notion th at it might actually deter entry. However, this phenomenon might be shortl ived as the triple-play service providers consolidate their market positions and lock in the customers. The impact of the LLU regulation RQ13 addresses the im pact of Local Loop U nbundling (LLU) regulation on market entry. The results of the regression indi cate that LLU regulation price significantly influences entry in the high-speed fixed broadband market. Specifica lly as LLU regulation price decreases, the number of providers of highspeed fixed lines increases. This finding is consistent with previous st udies of LLU regulation, which support that 1) unbundling stimulates the deployment of advanced technology in the wi reline sector (Bauer, 2005), and 2) unbundling increases market entry in the broadband market because it enables potential entrants to provide high bandwidth services without investment in infrastructure (Baranes, 2005). When the positive relationship be tween the practice of tr iple-play and market entry is taken into account, increasing market entry and the triple-play service will give consumers more benefits in terms of price and diverse service options. Implications Theoretical Implications One of the main goals of this study is to examine the impact of competition on the use of the triple-play strategy. It is proposed that a main driver of the triple-play strategy is the need to mitigate competition (Crampes & Hollander, 2006) Using the US firm dataset from 2000 to 2007, this study examines the impact of market competition and platform competition on the use

PAGE 184

184 of the triple-play strategy. This study found that competition plays a significant role in the use of triple-play in the cable and telephone industry. Specifically, bot h cable and telephone firms use the triple-play strategy to prevent rivals (either incumbent or potential entrant) from encroaching their market shares, to maintain their domina nt market positions, and to soften competition. There is also a sense of mutual dependency in this process of applying such a strategy. This finding is consistent with Bughin and Mendonca s (2007) findings that intensive market competition compels communications firms to offer triple-play bundles. This study assesses the applicab ility of Porters generic stra tegy in categorizing the tripleplay strategy by the cable and telephone indus tries. Porter (1980) proposed three types of strategies: overall cost leadersh ip strategy, differentiation, and fo cus. In particular, this study tests differences in market expense and capital expenditures between two i ndustries. In general, cable firms are more likely to invest in mark eting to promote their services and capital expenditures to upgrade their network platforms a nd to develop new services and technologies while telephone firms tend to be less inclined to invest in marketing and generally slower in upgrading infrastructure. According to Porter, the differentiation strategy centers on perceived uniqueness of services, which needs to invest marketing activities while the cost leadership strategy requires cost minimization in all areas. Therefore it is likely that cable firms tend to focus more on a differentiation strategy while te lephone firms employ a cost leadership strategy in the provision of the triple-play service. This study empirically supports bundling theories in the co mmunications sector. Bundling refers to the practice of selling products in a single package (Stremersch & Tellis, 2002). Bundling typically reduces in heteroge neous markets, which enhances market performance (Adams & Yellen, 1976; Salinger, 1995). However, the bundled product averages

PAGE 185

185 out the consumers preferences for each product, which leads to price competition and thereby negatively influences market performance (Mat utues & Regibeau, 1992; Reisinger, 2004). Using the US firm dataset from 2000 to 2007, this study ex amines the impact of the triple-play strategy on market performance (subscriber and financial performance (market share and profitability), and their growth) of the triple-play service providers. The result indicates that the triple-play stra tegy generally enhances market shares in the cable and telephone industry. The results also show that the triple-play strategy did not enhance firms profitability in the cable and telephone in dustry probably due to new marketing expenses associated with new services. In addition, this st udy suggests that there mi ght be a time threshold of better performance among TPS firms. For exampl e, the triple-play strategy is relatively successful in enhancing data market share when fi rms have engaged in the strategy for at least 4 years in the cable and telephone industry. The strategy significantly increases telephone firms profitability after 4 years of the practice of triple-play. In consideration of the growth rates, this study found that the strategy accelera tes the growth of market shar e only to a certain point. For example, the triple-play strategy accelerates the gr owth of data market share for 3 years only. For the telephone industry, the triple-pla y strategy accelerates the growth of video market share for 2 years only. Lastly the positive impact of the trip le-play strategy on growth rates might level off after 4 years of practices in the cable industry mostly likely due to the competition from telcos competitive triple-play strategy. This study also tests the role of switching costs in the provision of th e triple-play service. The result indicates that switchi ng costs generally cont ribute to improve market share, jointly with the triple-play st rategy. Thus this finding empirically supports switching costs theories,

PAGE 186

186 which assert that switching costs enable a firm to obtain a reliable consumer base during the term of the contract, which may guarantee future profits in that (Shapiro & Varian, 1999). The resource-based view of strateg y, emphasizing the individual firms unique capabilities and their impact on the firms business strategy in the market (Chan-Olmsted, 2006), is assessed with respect to the triple-play strateg y. This study examines differences in individual service performance between the cable and telep hone industries. The result indicates that cable and telephone firms demonstrate th e best performance in their co re service compared to other two services because they have resources that ar e unique and inimitable in their core service. This study partially validates bundling theories with respect to market entry, which is the role of bundling in deterring new entry in the market (Peitz, 2008). Using the US firm dataset from 2000 to 2007, this study examines the impact of the triple-play bundle on entry in the highspeed fixed broadband market. The result indica tes that the triple-pla y bundle is positively associated with market entry but at decreasing rates. It implies that triple-play bundling could encourage an initial increase of new entries to the market, contrary to the previous notion that it might actually deter entry. As triple-play service providers attract various segments of consumers by providing new communications services and technologies, an increasing number of firms enters the market by engaging in triple-play. Howe ver, this phenomenon might be shortlived as the triple-play service providers consolidate thei r market positions and lo ck in the customers. From a regulatory perspective, this study investigates how local loop unbundling regulation influences market en try. LLU regulation refers to th e process of requiring incumbent operators to open, wholly or in part, the last mile of their telecommunications networks to competitors (OECD, 2003). The result indicates that LLU regulation positively influences market entry in consideration of the triple -play strategy. It happens perhaps because LLU

PAGE 187

187 regulation encourages competitive firms to enter the market and to create new revenue streams with the provision of new high bandwidth IP se rvices with lower cost of broadband network connections. This finding is consistent with prior studies that LLU regulat ion stimulates market entry in the broadband market (Baranes, 2005; Bauer, 2005). Strategic Implications This study investigates how ma rket and firm factors influenc e the use of the triple-play strategy in the cable and telephone industries. Th e results indicate that cable and telephone firms show distinctive strategic behaviors on the basis upon their resources, capabilities, and surroundings. Cable firms are more likely to enga ge in the triple-play strategy in order to maintain their current position and/or to protect their market share from competition incurred by telcos triple-play. They tend to focus on ma rketing and innovation, and make efforts to differentiate themselves. In addition, they engage d in the triple-play st rategy ahead of telephone firms. On the other hand, telephone firms employ the triple-play strategy mostly to defend their market share as a result of the competition from cables triple-play. They are less likely to make investment in marketing and/or innovation, and tend to remain in a cost leadership position. This result implies that cable firms might con tinuously invest and e nhance their services and network platforms by capturing the new technologys benefits (i.e., cost reduction, service varieties, and good quality of services), and cons equently they might have more opportunities for product differentiation, which may lead to more strategic options to commercialize new IP services in a converging market. On the other hand, it also implies that telephone firms might continuously use triple-play as a response to co mpetition from cables triple-play by minimizing their investments in new services and network platforms. In addition to cost minimization behaviors, larger telephone firms might enjoy scale economies in distri buting the triple-play service over their network on the ba sis of a relatively stab le base of customers. As a result, it

PAGE 188

188 might delay the commercialization of new technology in the telephone in dustry in a converging market. The result of this study indicates that the triple-play strategy, in general, successfully increase the market share of video, voice, and data service in the cab le and telephone industry. The triple-play bundle may aggregate consumers heterogeneous preferences for individual services, which may result in the enhancement of market share-type performance in both industries. Furthermore, if a firm creates multip le triple-play bundles (that aggregate different segments of consumers) to satisfy multiple clusters of consumers, the triple-play strategy might be used as a profit maximizing tool. The result of this study shows that the trip le-play strategy has no si gnificant impact on a firms profitability in the cable and telephone indust ry. It might be due to the cost related issues. In the provision of the triple-p lay service, new market expens es and new capital expenditures would likely incur, which would reduce profitability in a short run. However the result of this study also found that the triple-p lay strategy might be effective in contributing to a firms profitability after it reaches a ce rtain stage. For example, the st rategy had a significant impact on the telcos after they practiced the strategy for 4 years. In sum, triple-play providers might experience initial temporary loss in the provision of the triple-play; however, they might begin to enhance their profitability as th ey reach a certain point. The result of this study indicates that th ere might be a time threshold of improved performance in the provision of the triple-play se rvice. In other words, market performance of the triple-play providers is likely to be enhanced after a certain point; however after the threshold, it is likely to be reduced. It may be most likely due to competition from other triple-play service providers. Hence triple-play service providers would generally have to move toward

PAGE 189

189 differentiation to avoid price competition. In other words, the next stage of successful triple-play strategy would be to provide value-added services and improve quality of services on the basis of their unique capabilities and resources (OECD, 2006). Firms are able to use the triple-play strategy to protect their market share or to maintain their dominant market positions (Bauer, 2006). Triple-play providers have a competitive advantage over single service provi ders in that they can use cro ss-branding and cross-promotions (ITU, 2006). Bundling also offers more opportunitie s for differentiation (C arbajo et al., 1990). As a result single service provide rs would also need to provide differentiated services and/or lower prices, or to target niche market to avoid direct competition with the triple-play providers. The result of this study shows that switching costs, jointly with the triple-play strategy, generally enhance market share of triple-play providers in the cable and telephone industries. The result indicates that switching costs acceler ate the growth of prof itability in the cable industry. The results imply that switching costs successfully locks in cons umers and thus prevent consumer turnover, which lead to enhanced market share and profitability growth. Therefore, the creation of switching costs is an important strategic option for the successful provision of the triple-play strategy. Policy Implications This study examines the impact of the triple-play bundle on market entry. With the emergence of the triple-play strategy, the potential of triple-play to deter entry has been debated (Peitz, 2008; Pernet, 2006). The results suggest that the triple-play st rategy is positively associated with market entry but at decreasing rate s. It appears that despite initial increases in market entry due to the impact of the triple-play bundle, it may play a role in preventing entry to some degree in the future when the practice of triple-play bundling becomes more prevalent. Zimmerman (2007) noted that the multiple-play stra tegy may change the current market structure

PAGE 190

190 to duopoly in the communications market in the future. In that light, wh ether the triple-play strategy can be employed in orde r to deliberately deter rivals entry should be continuously evaluated. In general, if the bundled price is not predatory an d bundling practices are based on competitive responses to rivals bundling behavi or, bundling may not have any negative impact on market entry (Kuhn, 2004). However, if bundling pr actice helps a firm to leverage its market power into another market, it would generate an ticompetitive effects, which may require certain policy intervention. This study also investigates the role of local loop unbundli ng (LLU) in market entry in the high-speed fixed broadband market by consider ing simultaneously the impact of the tripleplay strategy. The result of this study shows a positive relationship between unbundling and market entry. LLU regulation enables small entrants to connect with established network infrastructure and to provide high bandwidth se rvices without investment in infrastructure. Consequently, LLU regulation cont ributes to an increase in mark et entry (Baranes, 2005), and thus the regulation positively in fluences the deployment of advanced networks (Bauer, 2005). That is, LLU regulation creates an environment where not only facility-based firms but also small competitive firms could use the network infrastructure to deliver high bandwidth IP services and the triple-play service, and conseque ntly, firms, especially small competitive firms, may generate new revenue streams by using unbundl ed local loops for entry into the high-speed fixed broadband market. Such opportunities might be increased with the deployment of broadband networks, and thus cons umers might benefit from the diverse IP services as well as the triple-play services.

PAGE 191

191 Limitations and Suggestions for Future Research This study has limitations that are mainly associated with data availa bility and the use of a single proxy in most cases. Due to the paucity of data, more divers e explanatory variables cannot be included through all m odels utilized in this study. For the use of the triple-play model, if more data related to firms resources are available from 2000 to 2007, richer analyses of strategic be haviors could be possibl e both at the industryand firm-level. For the impact of triple-play on market performance model, this study employs individual service subscriber and financial data rather than the triple-play service subscriber and financial data as a whole because the triple-play se rvice data were not availa ble. If these kinds of data were accessible, the analyses of impacts of the triple-play strategy could be further refined. In addition, if pricing data of each service by fi rms were available, this study could empirically examine the direct impact of the triple-play competition between the two industries on their market performance. Throughout the two models, this study captures competition at the national level though in reality firm-to-firm competition varies by local areas, because of data availability. Although all firms are influenced by competition at the nationa l level, if direct competition between local cable and local telephone firms coul d be captured at the local level, more refined analyses would be possible. Moreover, to measure each variable, this study employs a single proxy for most variables due to the lack of data. The use of a single proxy for a variable might reduce the level of validity. If multiple measurements for these variables we re possible, more improved research for the analyses of the triple-pla y strategy could be achieve d by enhancing validity. Although this study attempts to capture the re levant variable overtime using longitudinal analyses, the history of practice of the triple-play strategy is relatively short, which may be too

PAGE 192

192 soon to detect the exact impact of the triple-play strategy. Th us continuous observations and analyses would be required by tracking all the changes. For future research, it would be informative to determine the optimal multiple play bundle from a consumers perspective. There is no st andard bundle in the curre nt market. The question can begin by answering whether the market needs a standard bundle (i.e., optimal bundle). To do so, information about demand patterns is necessar y. For example, if demand is too varied and distinctive across services, a firm may rais e its market share by implementing optional bundling, which refers to a bundl ing method that consumer can generate the bundle that meets their needs (Ben-Akiva & Gershenfel, 1988; Stre mersch & Tellis, 2002). If demand is either positively or negatively correlated, a standard bu ndle could increase a firms profit (Adams & Yellen, 1976; Salinger, 1995; Schm alensee, 1984). In this case, the optimal bundle can be found using multivariate statistical methods such as conjoint analysis. Pricing is also an interesting area to be e xplored. Depending on how the price of the tripleplay bundle is set, benefits to consumers and supp liers would be different. It is widely supported that in a market where competition becomes intense and consumers valuation is negatively correlated, mixed bundling strategy can be a better pure bundling strategy (Stremersch & Tellis, 2002). Applying theories behind pricing strategies to the triple-play bundle, optimal pricing strategy could be found from a firms perspec tive as well as a consum ers perspective. It would be valuable that how marketing expenses and capital expend itures are related to the elastic ratio of customer. This analysis woul d provide information that how a dollar change in expenses impacts a change of the number of consumer, which he lps establish a long-term budget plan for a firm.

PAGE 193

193 Technological convergence is a main driver of the triple-play st rategy, which causes structural changes in the telecommunications mark et. At this point, how to define markets for services and how to design policy and regula tion should be discussed. New services have emerged due to technological conve rgence, and these services curr ently compete with traditional services. If these new and traditional services an d markets are defined differently, the regulation imposing these services can be also different (e.g., fixed telephony vs. Vo IP). Consequently, it may prevent fair competition. Thus it would be valuable to study how a fair competitive environment can be built by clarifying definitions of services and markets in the convergence era. As technological convergence is rapidly achieved in the te lecommunications market, the quadruple-play strategy is ready to emerge. The quadruple-play strategy could be an outcome of perfect convergence (all vide o, fixed telephony, data, and wirele ss services delivered over a single network). Thus there are many research ag endas related to the quadruple play such as market definitions of services and players, re levant regulations and policies, the patterns of mergers and acquisitions, anti-trust issues relate d to bundling, and the role of switching cost in deployment of the quadruple play.

PAGE 194

194 LIST OF REFERENCES Aaker, D. A. (1984). How to select a business strategy. California Management Review, 26 (3), 167-175. Achrol, R., & Kotler, P. (1999). Ma rketing in the network econom y. Journal of Marketing, 63 146-164. Adams, W., & Yellen, J. (1976). Commod ity bundling and the burden of monopoly. Quarterly Journal of Economics, 91 475-298. Ahn, H., & Litman, B. R. (1997). Vertical inte gration and consumer welfare in the cable industry. Journal of Broadcasting & Electronic Media, 41 453-477. Albarran, A. B. (1996). Media economics: Understanding ma rkets, industries and concepts Ames, IA: Iowa State Press. Alesina, A., Ardagna, S., Nicoletti, G ., & Schiantarelli, F. (March 2003). Regulation and investment. NBER working paper series, 9560. Alexander, D., & Feinberg, R. (2004). En try in local telecommunication markets. Review of Industrial Organization, 25 107-127. Ambler, T., Kokkinaki, F., & Puntoni, S. (2004) Assessing marketing performance: Reasons for metric selection. Journal of Marketing Management, 20 (3/4), 475-498. Antonelli, C. (1997). A regulatory regime for innovation in the comm unications industries. Telecommunications Policy, 21 (1), 35-45. Aron, D. J. and Burnstein, D. E. (2003). Broadba nd adoption in the United States: An empirical analysis. In A. L. Shampine (Ed.) Down to the wire: Studies in the diffusion and regulation of telecommunications technologies (pp. 119-138), Nova Science Publishers, Haupaugge, NY. Bakos, Y., & Brynjolfsson, E. (2000). Bundlin g information goods: Pricing, profits, and efficiency. Management Science, 45 (12), 1613-1630. Baranes, E., & Bourreau, M. (2005). An economists guide to local loop unbundling. Communications & Strategies, 57 (1), 13-31. Baranes, E., & Le Blanc, G. (2006). Bundlin g strategies in ICT sector: An introduction. Communications & Strategies 63, 13-18. Barney, J. (1991). Firm resources and sustained competitive advantage. Journal of Management, 17(1), 99-120. Bauer, J. M. (2005). Bundling, di fferentiation, alliances and mergers: Convergence strategies in US communication market. Communications & Strategies, 60 (4), 59-83.

PAGE 195

195 Bauer, J. M. (2005). Unbundling policy in the United States. Communications & Strategies, 57(2), 59-82. Beam, R. (2006). Quantitative methods in medi a management and economics. In A. Albarran, S. Chan-Olmsted, & M. Wirth (Eds.) Handbook of Media Management and Economics (pp. 523-552). Mahwah, NJ: Lawrence Erlbaum. Bell, M., Silver, M., & Stray, S. (2001). Why growth rates do differ. Journal of Economic Studies, 9 (2), 51-67. Ben-Akiva, M., & Gershenfel, S. (1988). Multi -featured products and services: Analyzing pricing and bundling strategies. Journal of Forecasting 17, 175196. Bharat Book Bureau (January 2009). Quad play and service bundli ng strategy: Market analysis and forecasts 2009 Navi Mumbai, India. Bourreau, M., & Dogan, P. (2004). Service-based vs facility-based competition in local access networks. Information Economics and Policy, 16 287-306. Bresnahan, T., & Reiss, P. (1990). Entry in monopoly markets. Review of Economic Studies, 57, 531-553. Bresnahan, T., & Reiss, P. (1991). Entry an d competition in concentrated markets. Journal of Policial Economy, 99 (5), 977-1009. Bughin, J., & Mendonca, P. (2007). Convergence and triple play bundling: an empirical assessment for European telecommunications. Communications & Strategies, 68 121139. Burnett, R. (1992). The implications of ownershi p changes on concentration and diversity in the phonogram industry. Communication Research, 19, 749-769. Burton, M., Kaserman, D., Mayo, J. (1999). Modeli ng entry and barriers to entry: A test of alternative specifications, Antitrust Bulletin, Summer 387-420. Cabral, L. M., Salant, D. J., & Woroch, G. A. (1999). Monopoly pricing with network externalities. International Journal of Industrial Organization, 17 199-214. Calzada, J., & Valletti, T. (2008). Netw ork competition and entry deterrence. The Economic Journal, 118, 1223-1244. Carbajo, J., de Maza, D., & Seidmann, D. ( 1990). A strategic motivation for commodity bundling. The Journal of Industrial Economics, 38 (3), 283-298. Carlton, D., & Waldman, M. (2002) The strategic use of tying to preserve and create market power in evolving industry. Rand Journal of Economics, 33, 194-220.

PAGE 196

196 Chan-Olmsted, S.M. (2006). Competitive Strategy for Media Firms: Strategic and Brand Management in Changing Media Markets Mahwah, NJ: Lawrence Erlbaum & Associates, 13-37. Chen, Y. (1997). Equilibrium Product Bundling, Journal of Business 70(1), 85. Choi, J., & Stefanadis, C. (2001). Tying, i nvestment, and the dynamic leverage theory. Rand Journal of Economics, 32 52-71. Church, J. & Gandal, N. (2005). Platform comp etition in telecommunications. In Cave, M., Majumdar, S. & Vogelsang I. (ed.), The handbook of telecommunications vol. 2: Technology evolution and the Internet ( pp.118-153). Amsterdam, The Netherlands: Elsevier science. Cisco Systems (2005). Time Warner cable succeeds with innovative Triple play voice services. Retrieved February 5, 2008 from http://www.cisco.com/en/US/s olutions/collateral/ns341/ ns427/ns506/ns550/net_customer_profile0900aecd8026c5b8.pdf Clarke, R., Hassett, K., Ivanova, Z., & Kotlikoff, L. (2004). Assessing the economic gains from telecomm competition NBER Working Paper Series 10482. . Clement, M. T. (2004). Direct and indirect network eff ects: Are they equivalent? International Journal of Industrial Organization, 22 633-645. Comin, D., & Mulani, S. (2005). A theory of growth and volatility at the aggregate and firm level NBER working paper series, 11503. Constantiou, I. D., & Kautz, K. (2008). Ec onomic factors and diffusion of IP telephony: Empirical evidence from an advanced market. Telecommunications Policy 32, 197-211. Cooper, M., & Kimmelman, G. (1999). The digital divide confr onts the Telecommunications Act of 1996: Economic reality versus public policy Mimeo. Crampes, C., & Hollander, A. (2006). Triple play time. Communication and Strategies, 63, 5171. Crandall, R. (February 2009). Broadband policy: Investment and regulation Mimeo. Crandall, R. W., Ingraham, A. & Singer, H. (2004). Do unbundling policies discourage CLEC facilities-based investment? Topics in Economic Analysis & Policy, 4 (1), Article 1. Crawford, G. S. (2008). The discriminatory in centives to bundle in the cable television industry. Quantitative Marketing and Economics, 6(1), 41-78.

PAGE 197

197 Dean, T. J., Brown, R. L., & Bamford, C. E. (1998). Differences in large and small firm responses to environmental cont ext: strategic implications fro m a comparative analysis of business formulations. Strategic Management Journal, 19(8), 709-728. Denni, M., & Gruber, H. (2007). The diffusion of broadband telecommunications in the U.S.: the role of different forms of competition. Communications & Strategies, 68 139-158. Dierickx, I., & Cool, K. (1989). A sset stock accumulation and the sustainability of competitive advantage. Management Science, 35 (12), 1504-1511. Distaso, W., Lupi, P., & Manenti, F. (2006) Platform competition and broadband uptake: Theory and empirical evidence from the European Union. Information Economics and Policy, 18, 87 106. DotEcon & Criterion Economics (2003). Competition in broadband provision and its implication for regulatory polic y: A report for the Brussels Round Table. Retrieved July 7, 2006, from http://www.dotecon.com/publications/BRTfull15 10-03.pdf Economides, N. & Himmelberg, C. (1995). Critical mass and network size with application to the US fax market. Discussion paper EC-95-11, Ster n School of Business, New York University. Economides, N. (1996a). Network externalities, complementarities, and invitation to enter. European Journal of Political Economy, 12 211-233. Economides, N. (1996b). The economics of networks. International Journal of Industrial Organization, 14 673-699. Emmons, W. M., & Prager, R. A. (1997). The e ffects of market structure and ownership on prices and service offering in the U.S. cable television industry. Rand Journal of Economics, 28(4), 732-750. Eppen, G. D., Hanson, W. A., & Martin, R. K. (1991). Bundling New products, new markets, low risk. Sloan Management Review, 32 (4), 7-14. Faulhaber, G. R. (2003). Policy-induced comp etition: the telecommuni cations experiments. Information Economics and Policy, 15 73-97. Faulhaber, G. R., & Hogendorn, C. (2000) The market structure of broadband telecommunications. The Journal of Industrial Economics, 48 (3), 305-329. Fay, S. A. & MacKie-Mason, J. K. (July, 2001). Competition between firms that bundle information goods SSRN working paper series, 975733. Federal Communications Commission (August, 2008 ) Trends in Telephone Service. Washington, DC: FCC.

PAGE 198

198 Federal Communications Commission (February 10, 2006). Video competition Washington, DC: FCC. Federal Communications Commission (February, 2009). Video competition Washington, DC: FCC. Federal Communications Commission (May 8, 2007a ) Telephone subscribership in the United States. Washington, DC: FCC. Federal Communications Commission (October 2007b). High-speed services for Internet Access: Status as of December 31, 2006. Washington, DC: FCC. Ferguson, C. (2002). The US broadband problem. Retrieved September 22, 2008 from http:// www.brookings.edu/~/media/Files/rc/p apers/2002/07technology_ferguson/pb105.pdf Field, A. (2000). Discovering statistics using SPSS for Windows. London, UK: Sage. Figueiredo, R. J., & Edwards, G. (2007). Does private money buy public policy? Campaign contributions and regulatory out comes in telecommunications. Journal of Economics and Management Strategy 16 (3), 547-576. Ford, G. S., & Jackson, J. D. (1997). Horizontal Concentration and Verti cal Integration in the Cable Television Industry. Review of Industrial Organization, 12 501 Ford, G. S., Koutsky, T. M., & Spiwak, L. J. (June 2005). Competition after unbundling: Entry, industry structure, and convergence. Phoenix Center Policy Working Paper 21. Foros, O., & Hansen, B. (2001). Competiti on and compatibility among Internet service providers. Information Economics and Policy, 13, 411-425. Frieden, R. (2005a). Unbundling the local loop: A cost/b enefit analysis for developing nations. Info 7(6), 3-15. Gabel, D., & Huang, G. (January, 2003). Promoting innovation: Impact of local competition and regulation on deployment of advanced telecommunications services for business. Paper presented at the conference of the Amer ican Federation of E volutionary Economics (AFEE), Washington, D.C. Gabel, R., & Gabel, D. (September, 1997). The application of cost data in the telecommunications industry Presented at the Telecomm unications Policy Research Conference, Arlington, VA. Gans, J. (2000). Network competition and consumer churn. Information Economics and Policy, 12, 97-109. Gentry, R., & Jamison, M. (May 2005). Market behavior and survival of entrepreneurial firms: how do rookies compete? Public Utility Research Center Working Paper.

PAGE 199

199 Goddard, J., Tavakoli, M., & Wilson, J. O. (2009). Sources of variation in firm profitability and growth. Journal of Business Research, 62 495-508. Gotz, G. (2002). Sunk costs, windows of prof it opportunities, and the dynamics of entry. International Journal of Industrial Organization, 20 1409-1436. Grubesic, T. H., & Murray, A. T. (2004). Waiting for broadband: Local competition and the spatial distribution of advanced telecomm unication services in the United States. Growth and Change, 35(2), 139-165. Hair, J. F., Anderson, R. E., Bl ack, B., & Tatham, R. L. (1998). Multivariate data analysis Upper Saddle River, NJ : Pearson Education. Hause, J. C., & DuRietz, G. (1984). Entry, i ndustry growth and the microdynamics of industry supply. Journal of Political Economy, 92 (4), 733-757. Hausman, J. A. (2001). Regulation by TSLR IC: Economic effects on investment and innovation. In J. Sidak, C. Engel, & G. Knieps (Eds.), Competition and regulation in telecommunications: Examining Germany and America Boston, MA: Kluwer Academic. Hazlett, T. W. (2006). Cable television. In Ca ve, M., Majumdar, S. & Vogelsang I. (ed.), The handbook of telecommunications vol. 2: Structure, regulation and competition ( pp.191240). Amsterdam, The Nether lands: Elsevier science. Higgins, J. (2004, July). A di fferent kind of price war. Broadcasting & Cable 134, 14. Hitt, M. A., Ireland, R. D., & Hoskisson, R. E. (2005). Strategic Management: Competitiveness and globalization Versailles, KY: South-Western. Hovenkamp, H. (1990). Antitrust analysis of market power, with some thoughts about regulated industries. In J.R. Allison and D.L. Thomas (Eds.) Telecommunications Deregulation: Market Powe r and Cost Allocation issues Westport, CT: Quorum Book. Hubbard, R. G., & Lehr, W. H. (1998). Improving local exchange competition: Regulatory crossroads New York, NY: Columbia University. International Telecommunication Union (2003). Promoting broadband: Background paper for workshop on promoting broadband. Retrieved November 11, 2005 from http://www.itu.int/osg/spu/ni/promotebr oadband/PB03-PromotingBroadband.pdf International Telecommunication Union (2006). Multiple play. Retrieved May 1, 2008, from http://www.itu.int/WORLD2006/forum/multiple_play.html Janssen, M. & Mendys-Kamphorst, E. (2008). Triple play: How do we secure future benefits? Telecommunications Policy, 32 735-743. Jedidi, K., Mela, C., & Gupta, S. (1999). Managing advertising and promotion for long-run profitability. Marketing Science, 18 (1), 1-22.

PAGE 200

200 Karikari, J. A., Brown, S. M., & Abramowitz, A. D. (2003) Subscriptions for direct broadcast satellite and cable TV in th e US: an empirical analysis. Information Economics and Policy, 15, 1-15. Kaserman, D. L., & Mayo, J. W. (2002). Competi tion in the long-distance markets. In Cave, M., Majumdar, S. & Vogelsang I. (ed.), The handbook of telecommunications vol. 1: Structure, regulation and competition (pp.509-561). Amsterdam, The Netherlands: Elsevier science. Katz, M. L. & Shapiro, C. (1985). Network ex ternalities, competition, and compatibility. American Economic Review, 75 424-440. Katz, M. L. & Shapiro, C. (1992). Product in troduction with netw ork externalities. The Journal of Industrial Economics, 40 (1), 55-84. Keller, K. (2003). Building, measuring, and managing brand equity Upper Saddle River, NJ: Prentice-Hall. Kenney, P. (2003). A guide to econometrics. Cambridge, Massachusetts; MIT Press. Kerlinger, F. & Pedhazur, E. (1973). Multiple regression in behavioral research New York, NY: Holt, Rinehart Winston. Kim, H. & Kwon, N. (2003). The advantage of network size in acquiring new subscribers: a conditional logit analys is of the Korean m obile telephony market. Information Economics and Policy, 15 17-33. Kim, J. (2002). Product differentiation and netw ork externality: a comment on Economides: Network externalities, complementar ities, and invitations to enter. European Journal of Political Economy, 18, 397-399. Kim, M., & Kim, H. (2007). Is there early take-off phenomenon in diffusion of IP-based telecommunications services? Omega, 35 727-739. Kim, R. (May 28, 2007). Triple play not enough? Say hey to quad play telecoms adding cell service to TV, Internet, landlines Retrieved November 2, 2008 from http://www.sfgate.com/cgi-bin/artic le.cgi?f=/c/a/2007/05/28/BUG0LQ1F1I1.DTL Klemperer, P. (1995). Competition when consumer s have switching costs: an overview with applications to industrial organization, macroeconomics and international trade. Review of Economic Studies, 62 (4), 515-540. Koski, H. A., & Majumdarm S. K. (2002). Pa ragons of virtue? Competitor entry and the strategies of incumbents in the U.S. local telecommunications industry. Information Economics and Policy, 14 453-480. Koski, H., & Kretschmer, T. (2005). Entry, standards and competition: Firm strategies and the diffusion of mobile telephony. Review of Industrial Organization, 26 89-113.

PAGE 201

201 Krauss, J. (2006). Market drivers for tr iple play. Retrieved March 14, 2008, from http://www.keymile.com/media/internet/pre ss_events/text_press/TecPreview__Triple_Pl ay_Market_Drivers.pdf Laffont, J., & Tirole, J. (2001). Competition in telecommunications Cambridge, Massachusetts: MIT Press. Landers, D., & Chan-Olmsted, S. (April 2002). Assessing the changing network television market: A resource-based analysis of broadcast television networks. Paper presented to the Broadcast Education Association, Las Vegas, NV. Lee, S. (2006). Broadband deployment in the United States: Examining the impacts of the platform competition. The International Journal on Media Management 8(4), 173-181. Lee, S. M. (2009). The Triple-play Bundle Strategy of Cable and Telephone Companies in the Current U.S. Telecommunications Market. The International Journal on Media Management,11(2), 61-71. Lee, S. W. (2006). Broadband deployment in th e United States: Examining the impacts of the platform competition. The International Journal on Media Management, 8 (4), 173-181. Lee, S. W., & Lee, S. M. (2008). Multiple Play Strategy in Global Telecommunication Markets: An Empirical Analysis International Journal of Mobile Marketing, 3 (2), 44-53. Lee, S., & Marcu, I. M. (2007c, August). Fixed and mobile broadband deployment: An empirical analysis of adoption factors. Paper presented to the 35th Research Conference on Communication, Information and Internet Policy Arlington, VA. Lee, S., Brown, J. (2008). Examining broadband adoption factors: An empirical analysis between countries. Info: the Journal of Policy, Regulation, and Strategy for Telecommunication, Information, and Media, 10 (1), 2539. Lee, S., Chan-Olmsted, S., & Ho, H. (2008). The emergence of mobile virtual network operators (MVNOs): An examination of the business strategy in the global MVNO market. The International Journal on Media Management, 10 10-21. Li, S. (2004). Market competition and the medi a performance of Taiwans cable television industry. Journal of Media Economics, 17 (4), 279-294. Li, S. (2007). Market competition and the me dia performance: Reexamining the media performance of the cable te levision industry in Taiwan. Journal of Media Economics, 20(3), 189-210. Liao, C., & Tauman, Y. (2002). The ro le of bundling of price competition. International Journal of Industrial Organization, 20 365-389. Liebowitz, S. J. & Margolis, S. E. (1994). Network externality: An uncommon tragedy. Journal of Economic Perspective 8(2). 133-150.

PAGE 202

202 Liebowitz, S. J., & Margolis, S. E. (2002). Networ k Effects. In Cave, M. E., Majumdar, S. K., & Vogelsang, I. (Eds.) Handbook of Telecommunications Economics: Structure, regulation and competition, Vol 1 (pp. 75-96). Armsterdam, The Netheralnds: Elsevier Science. Litman, B. (1979). The television networ ks, competition and program diversity. Journal of Broadcasting, 23 393-409. Lommies, D., & Swann, C. (2005). Intermoda l competition in local telecommunications markets. Information Economics and Policy, 17 97-113. Maijoor, S., San Witteloostuijin, A. (1996). An em pirical test of the resource-based theory: Strategic regulation in the Dutch audit industry. Strategic Management Journal, 17 (1), 549-469. Malecki, E. (2002). Local competition in tel ecommunications in United States: Supporting conditions, policies, and impacts. The Annals of Regional Science, 36 437-454. Matutes, C., & Regibeau, P. (1992). Compatib ility and bundling of complementary goods in a duopoly, The Journal of Industrial Economics, 40 (1), 37-54. Michalakelis, C., Varoutas, D., & Sphicopoul os, T. (2008). Diffusion models of mobile telephony in Greece. Telecommunications Policy, 32 234-245. Mitchell, M. F., & Skrzypacz, A. (2006). Netw ork externalities and long-run market shares. Economic Theory, 29, 621-648. Moorman, C., & Rust, R. (1999). The role of marketing. Journal of Marketing, 63 180-197. Nalebuff, B. (2004). Bundling as an entry barrier. Quarterly Journal of Economic 119, 159187. National Cable and Telecommunications Association. Statistics: Cable Broadband Availability Retrieved November 2, 2007 from http://www.ncta.com/Statistic/ Statistic/CableBroadbandAvailability.aspx Net Insight (2006). Net Insight interim re port. Retrieved on January 13 2009 from http://netinsight.se/pdf/060503_Interim_Report_Q1_ENG.pdf Nortel (2005). This is the way video service doubles your average revenue per user Retrieved November 18, 2007 from http://www.nortel.com/solutions/brdbnd accss/collateral/nn112240.pdf Nuechterlein, J. E., & Weiser, P. J. (2005). Digital crossroad: American Telecommunications policy in the Internet age Cambridge, MA : MIT Press. Oren, S. S., & Smith, S. A. (1981). Critical mass and tariff structure in electronic communications markets. The Bell Journal of Economics, 12 (2), 467-487.

PAGE 203

203 Organization for Economic Co-ope ration and Development (2006). Multiple play: Pricing and policy trends. Paris: OECD. Organization for Economic Co-operation and Development (2007). Mobile multiple play: New service pricing and policy implications Paris: OECD. Retrieved October 29, 2007 from http://www.oecd.org/dataoecd/18/19/38057443.pdf Ovum (2008). Measuring the effectiveness of bundles: A European perspective. Boston, MA: Ovum. Ozanich (2006). Media finance and valuation. In A. Albarran, S. Chan-Olmsted, & M. Wirth (Eds.) Handbook of Media Management and Economics (pp. 601-621). Mahwah, NJ: Lawrence Erlbaum. Papandrea, F., Stoeckl, N., & Daly, A. (2003). Bundling in the Australian telecommunications industry The Australian Economic Review, 36 (1), 41-54. Parks Associates (April, 2006). The Business of Bundled Services: Consumers, Models, and Uptake. Retrieved July 28, 2008 from http://www .parksassociates.com/research/reports/ tocs/2006/bundled_services.htm Peitz, M. (2008). Bundling may blockade entry. International Journal of Industrial Organization, 26, 41-58. Pernet, S. (2006). Bundles and range stra tegies: The case of telecom operators. Communications & Strategies, 63 19-31. Pindyck, R. S. (2003). Mandatory unbundling and irreversible in vestment in telecom netwoks. MIT Sloan Working Paper No. 4452-03. Porter, M.E. (1980). Competitive Strategy New York, NY: Free Press Porter, M.E. (1991). Toward a dynamic theory of strategy. Strategic Management Journal, 12, 95-117. Reda, R. (2004). Intelligence in communicati on systems evolution, trends and business opportunities. Intelligence in Communication System, 3283 1-15. Reisinger, M. (2004). The effects of product bundling in duopoly Discussion paper 2004-26, University of Munich. Retrieved September 7, 2007 from equb.ub.unimuenchen.de/archive/00000477. Rohlfs, J. (1974). A theory of interdependent demand for a communications service. The Bell Journal of Economics and Management Science, 5 (1), 16-37. Roller, L., & Waverman, L. (2001). Teleco mmunications infrastructure and economic development: A simultaneous approach. American Economic Review 91 (4), 909-923.

PAGE 204

204 Salinger, M. (1995). A graphical analysis of bundling. Journal of Business, 68 (1), 85-98. Savage, S., & Wirth, M. (2005). Price, Program ming and Potential Competition in US Cable Television Markets. Journal of Regulatory Economics, 27 (1), 25. Schmalensee, R. (1984). Gaussian demand and commodity bundling. Journal of Business, 57 211-230. Schumpeter, J. (1950). Capitalism, socialism, and democracy. New York, NY: Harper Publishing Company. Seggie, S., Cavusgil, E., & Phelan, S. (2007). M easurement of return on marketing investment: A conceptual framework and the future of marketing metrics. Industrial Marketing Management, 36, 834-841. Seim, K. (2006). An empirical model of firm entry with endogenous product-type choices. The Rand Journal of Economics, 37 (3), 619-640. Seo, S. (2007). Triple-play competition in the US telecommunications industry: Exploring cable operators adoption pattern of triple-bundled services. International Journal on Media Management 9(1), 1-8. Seo, S. (2008). Convergence and diversification in cable industry: An empi rical analysis of the deployment pattern of cable telephony in US local tel ecommunications markets. Telecommunications Policy 32, 145-155. Shapiro, C. & Varian, H. R. (1999). Information Rules. Boston, MA: Harvard Business School Press. Sherman, B. L. (1995). Telecommunications management: Broadcasting/cable and the new technologies. New York, NY: McGraw-Hill. Shim S., & Oh, J. (2006). Service bundling a nd the role of access charge in the broadband Internet service market. Communications and Strategies, 63 73-91. Shy, O. (2001). The economics of network industries. Cambridge, United Kingdom: Cambridge University Press. Soubeyran, A., Suzumura, K., & Weber, S. (2007). Competition, regulation and welfare in the presence of network externalities. The Japanese Economic Review, 58 (1), 47-70. Stremersch, S. & Tellis, G. (2002). Strategic bun dling of products and prices: A new synthesis for marketing. Journal of Marketing, 66, 55-72. Tardiff, T. J., (2007). Changes in industr y structure and technological convergence: implications for competition policy and regulation in telecommunications. IEEP, 4, 109 133.

PAGE 205

205 Time Warner (2007). Time Warner cable news Retrieved March 29, 2008, from http://www.timewarnercable.com/InvestorRe lations/PressReleases/TWCPressReleaseDet ail.ashx?PRID=2111&MarketID=0 Time Warner Cable (2008). Time Warner Cable reports Retrieved on January 13 2009 from http://www.timewarnercable.com/MediaLibrar y/1/1/Investor%20Relations/documents/20 08_Q2/2Q08_TWCEarningsRelease_08%2005%2008.pdf TNS telecom (2005). Consumers indifferent about telecom service providers. Retrieved July 23, 2008 from http://tnstelecoms.com/press-10-27-05-html Trend Communications (2006). Triple play business development. Retrieved November 30, 2007, from http://www.trendcomms.com UT Starcom (2006). Realizing the triple-play value proposition Retrieved Feburary 2, 2008 from http://www.utstarcom.co m/Document_Library/0640.pdf Vagliasindia, M., Izzet Guney, I., Taubman, C. (2006). Fixed and mobile competition in transition economies.Telecommunications Policy, 30 349367 Varian, H. (2003). Intermediate microeconomics: A modern approach New York: Norton. Whinston, M. (1990). Tying, fo reclosure, and exclusion. The American Economics Review, 80(4), 837-859. Wildman, S. (2006). Paradigms and analytical frameworks in the modern economics and media economics. In A. Albarran, S. Chan-Olmsted, & M. Wirth (Eds.) Handbook of Media Management and Economics (pp. 67-90). Mahwah, NJ: Lawrence Erlbaum. Woroch, G. A. (2006). Local network competition. In Cave, M., Majumdar, S. & Vogelsang I. (ed.), The handbook of telecommunications vol. 1: Structure, regulation and competition (pp.642-716). Amsterdam, The Netherlands: Elsevier science. Yaffee, R. (2003). A prime for panel data analysis Retrieved October 8, 2008 from http://www.nyu.edu/its/pubs/connec t/fall03/yaffee_primer.html Yankee Group (2003). Telco prepare for video: A new era of competition Retrieved May 1, 2008 from http://www.tripleplay.tv/pdf/Yank ee_Group _Telcos Prepare_For_Video.pdf Yankee Group (December 2004). Telco prepare for video: A new era of competition Retrieved February 5, 2008 from http://www.trip leplay.tv/pdf/Yankee_Group _Telcos Prepare_For_Video.pdf Zimmerman, P. R. (2007). Recent developm ent of US wireline telecommunications. Telecommunications Policy, 31 419-437. Zollo, M., & Winter, S. G. (2002). Deliber ate learning and the evolution of dynamic capabilities. Organization Science, 13 (3), 339-351.

PAGE 206

206 Zolnierek, J., Eisner, J., & Burt on, E. (2001). An empirical exam ination of entry patterns in local telephone markets. Journal of Regulatory Economics, 19, 145-149.

PAGE 207

207 BIOGRAPHICAL SKETCH Seonm i Lee received his bachelors degree in mass communication at Ewha Womans University in Seoul, Korea. She also earned a ma sters degree in tel ecommunication at Indiana University in Bloomington, Indi ana. Lee conducted research in the area of telecommunication policy, media economics, media management, and new media technology and regulation. Her work has been published in journals such as International Journal on Media Management, and International Journal of Mobile Marketing.